Articles by alphabetic order
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
 Ā Ī Ñ Ś Ū Ö Ō
1 2 3 4 5 6 7 8 9 0


The Fundamental Wisdom of the Middle Way

From Chinese Buddhist Encyclopedia
Jump to navigation Jump to search
Da1b47.jpg



Mūlamadhyamakakārikā


The Fundamental Wisdom of the Middle Way


The Fundamental Wisdom of the Middle Way

Nagrjuna’s Mūlamadhyamakakārikā


TRANSLATION AND COMMENTARY BY


JAY L. GARFIELD

[Madhyamakakrik. English & Sanskrit]

The fundamental wisdom of the middle way: Nagrjuna’s Mūlamadhyamakakārikā / Translation and commentary by Jay L. Garfield,



I dedicate this work,

with profound gratitude

and respect,to the Most Ven. Professor Samdhong Rinpoche:

scholar, educator, statesman, public servant

and shining exemplar of monastic life.


Preface


This is a translation of the Tibetan text of Mūlamadhyamakakārikā. It is perhaps an odd idea to translate a Tibetan translation of a Sanskrit text and to retranslate a text of which there are four extant English versions. My reasons for doing so are these: First, I am not satisfied with any of the other English versions. Every translation, this one included, of any text embodies an interpretation, and my interpretation differs in various respects from those of my predecessors in this endeavor. This is to be expected. As Tuck (1990) has correctly observed, Nagrjuna, like any philosopher from a distant cultural context, is always read against an interpretive backdrop provided by the philosophical presuppositions of the interpreter, and by previous readings of

Nagrjuna. So I claim no special privileged position vis à vis Streng (1967), Inada (1970), Sprung (1979), or Kalupahana (1986)—only a different position, one that I hope will prove useful in bringing Mūlamadhyamakakārikā into contemporary philosophical discourse. I, like any translator/interpreter must acknowledge that there is simply no fact of the matter about the correct rendering of any important and genuinely interesting text. Interpretations, and with them, translations, will continue to evolve as our understanding of the text evolves and as our interpretive horizon changes. Matters are even more complex and indeterminate when

the translation crosses centuries, traditions and languages, and sets of philosophical assumptions that are quite distant from one another, as is the case in the present project. So each of the available versions of the text embodies a reading. Inada reads Nagarjuna from the standpoint of the Zen tradition, and his translation reflects that reading; Kalupahana reads Nagarjuna as a Theravada commentator on the Kaccynagotta-stra, and his translation reflects that reading, as well as his view about the affinities between James’s pragmatism and Theravada Buddhism. Sprung adopts Murti’s Kantian interpretation of Madhyamika, and his translation reflects that interpretation. Streng reads the text as primarily concerned with religious phenomenology. There is no translation of this text into English, and no commentary on it, that specifically reflects an Indo-Tibetan Prasangika-Mūlamadhyamakakārikā interpretation. Inasmuch as this is my own preferred way to read Nagrjuna, and the reading dominant in Tibetan and highly influential in Japanese and Chinese discussions of Mūlamadhyamakakārikā, I believe that it is important to fill this lacuna in the English bibliography.

Having argued that all translation involves some interpretation and, hence, that there is always some distance between an original text and a translation, however good and canonical that translation may be, it follows that Mlamadhyamakakàrik and dBu-ma rtsaba shes-rab differ, however close they may be and however canonically the latter is treated. Since dBu-ma rtsa-ba shes-rab is the text read by and commented on by generations of Tibetan philosophers, I think that it is important that an English translation of this very text be available to the Western philosophical public. This text is hence worthy in its own right of translation inasmuch as it is the proper subject of the Tibetan philosophical literature I and others find so deep and fascinating.

This is not a critical scholarly edition of the text. It is not philological in intent; nor is it a discussion of the commentarial literature on Nagarjuna’s text. There is indeed a need for such a book, but that need will have to be filled by someone else. This is rather meant to be a presentation of a philosophical text to philosophers, and not an edition of the text for Buddhologists. If philosophers and students who read my book thereby gain an entrance into Nagarjuna’s philosophy and see Mūlamadhyamakakārikā, as interpreted herein, as a text worthy of study and discussion, this work will have served its purpose. Since my intended audience is not Buddhologists, per se, but Western philosophers who are interested in Buddhist philosophy, I have tried to balance standard renderings of Buddhist terminology with more perspicuous contemporary philosophical language. I am not sure that I have always made the right decisions or that I have found the middle path between the extremes of Buddhological orthodoxy and Western revisionism. But that is the aim.

I am also striving for that elusive middle path between two other extremes in translation: I am trying on the one hand to avoid the unreadable literalism of translations that strive to provide a verbatim report of the words used the original, regardless of whether that results in a comprehensible English text. But there is on the other hand the extreme represented by a translation written in lucid English prose purporting to be what the original author would have written had he been a twentieth-century philosopher writing in English, or one that, in an attempt to convey what the text really means on some particular interpretation, is in fact not a translation of the original text, but a completely new book, bearing only a distant relation to the original. This hopelessly mixes the tasks of translation on the one hand and� critical commentary on the other. Of course, as I have noted above, these tasks are intertwined. But there is the fault of allowing the translation to become so mixed with the commentary that one no longer has a grip on, for example, what is Nagarjuna and what is Garfield. After all, although the text is interpreted in being translated, this text should still come out in translation as a text which could be interpreted in the ways that others have read it. Because the original does indeed justify competing interpretations. That is one of the things that makes it such an important philosophical work.

Acknowledgments


Thanks are already due to many who have helped at different stages of this project: Thanks to Bob Thurman and David Sloss for first introducing me to Buddhist philosophy and then for encouraging me to wade deeper. Thanks to David Kalupahana, Steve Odin, Kenneth Inada, and Guy Newland, as well as to David Karnos, Joel Aubel, Dick Garner, and William Herbrechtsmeier for many hours of valuable and enjoyable discussion of this text at the National Endowment for the Humanities Summer institute on Nagarjuna in Hawaii. And thanks to the NEH for the grant support that enabled my participation in that institute. I am especially grateful to Guy Newland for many subsequent conversations, useful suggestions, encouragement, and a critical reading of my work. Thanks to Janet Gyatso for countless hours of profitable and enjoyable philosophical conversation and for many useful and detailed criticisms and suggestions on this and other related work. Thanks to the Ven. Geshe Lobzang Tsetan for starting me in Tibetan, for much useful philosophical interchange, for teaching me an immense amount about Madhyamika, and for his close criticism of this text; to Georges Dreyfus (Geshe Sengye Samdup) for much useful advice and discussion; and to Joshua and Dianne Cutler and the Tibetan Buddhist Learning Center of North America for hospitality. I also thank John Dunne for detailed comments on several chapters of an earlier draft of this translation.

I am grateful to the Indo-American Foundation, the Council for the International Exchange of Scholars, and the Smithsonian Institution for an Indo-American Fellowship in 1990-91. During that time, as a Visiting Senior Research Scholar at the Central Institute of Higher Tibetan Studies, I began work on this project. I owe an enormous debt of gratitude to The Most Ven. Prof. Samdhong Rinpoche and his staff for hosting me and my family at the Central Institute of Higher Tibetan Studies and to Rinpoche himself for his generous personal help. I thank the Ven. Geshe Ngawang Sherab for all of his kind logistical help at Santarakshita Library and for friendship and philosophical interchange. Thanks also to the Ven. Lobzang Norbu Shastri and the Ven. Acarya Ngawang Samten for extensive conversations from which I learned much and for useful comments on this work and to Karma for Tibetan lessons.

I am deeply grateful to the Ven. Prof. Geshe Yeshes Thap-Khas for reading dBu-ma rtsa-ba shes-rab and related texts with me and for giving me his invaluable oral commentary on these texts during that year and on many subsequent occasions. Nobody has taught me more about Madhyamika philosophy, and it is hard� to imagine a more patient, generous, and incisive scholar and teacher. Without his lucid teachings, and without Geshe-la’s enormous patience, I could never have approached this text with any degree of success. While he would not agree with everything I say, my own reading of this text is enormously influenced by his. Special thanks to Sri Yeshi Tashi Shastri for his translation and transcription assistance during many of these sessions and for an enormous amount of cheerful and generous general research assistance, including a great deal of careful proofreading and detailed comments on this translation.

During that year and in subsequent years I also benefited greatly from my visits to the Institute of Buddhist Dialectics. I am deeply grateful to the Ven. Prof. Geshe Lobzang Gyatso for his hospitality and for his teaching. In our many conversations and from his writings I have learned a great deal, and this project certainly reflects his influence. Without his patient advice on interpretative and expository details and without his vigorous critique of many of my ideas it would have been impossible to produce this commentary. I thank the Ven. Sherab Gyatso for his tireless and invaluable translation and assistance during that time. The Ven. Sherab Gyatso, The Ven. Graham Woodhouse, the Ven. Tenzin Dechen, and the Ven. Huen have given much to me in many hours of philosophical interchange through translation help and through their hospitality and friendship. Mr. Phillipe Goldin has also offered many helpful suggestions on the translation and commentary. I also thank the Ven. Khamtrul Rinpoche, the Ven. Geshe Yeshe Topden (Gen Drup-Thop) and Gen Lam-Rim-pa for their teachings and Acarya Nyima Tshering for his introduction and translation on those occasions. Special thanks to Nyima Penthog for improving my Tibetan.

I thank His Holiness the Dalai Lama for his encouragement and for valuable discussion of some difficult interpretative issues.

I am also very grateful to friends and colleagues at Drepung Loseling Monastic College. My visit there was extremely enjoyable and also philosophically fruitful. Thanks to the Ven. Geshe Dak-pa Toepgyal and the Ven. Thupten Dorjee for arranging everything and for talking with me about this and other work. I am very grateful to the Ven. Geshe Namgyal Wangchen for detailed comments and encouragement on this work and for useful discussions about Madhyamika, translation, the task of presenting Buddhist philosophical texts to the West, and other topics.

My acknowledgment of help in India would not be complete without acknowledging the gracious hospitality and assistance in living of Sri N. N. Rai, Sri Arun Kumar Rai, Sri A. R. Singh, and their families in Sarnath; the hospitality of Kunzom Topden Martam and his family in Sikkim—it was the Martam house in which the writing actually got started; and Dr. L. S. Suri of the American Institute of Indian Studies in New Delhi, whose administrative efficiency kept everything moving smoothly.

I am deeply grateful to four friends who read a complete draft of this work and provided honest, searching, sometimes scathing criticism. What more could one ask from colleagues and friends? Many of their suggestions are incorporated in the book as it now stands, and much of whatever is good in it is due to their enormous contributions. Sometimes I have disagreed with each of them. And whatever errors remain are certainly my own. So thanks especially to the Ven. Gareth Sparham, the Ven. Sherab Gyatso, Guy Newland, and Jane Braaten for copious corrections and criticism and for extensive productive discussion. Thanks also to Prof. Alan Sponberg for useful comments on an earlier draft and to Janet Gyatso, Graham Parkes, and Georges Dreyfus for reading and commenting on the penultimate� draft.

Another group of colleagues to whom I owe thanks are those who kept faith. This may require some explanation. I discovered when I—a Western, analytically trained philosopher of mind—began to work on Buddhist philosophy that many in philosophy and cognitive science took this as evidence of some kind of insanity, or at least as an abandonment of philosophy, per se. This is not the place to speculate on the origins or nature of the stigma attaching in some parts of our profession to Asian philosophy. But it is a sad fact to be noted and to be rectified. In any case, I therefore owe special thanks to those who went out of their way to support this work and to let me know that they took it and me seriously. I thank especially my friend and colleague Meredith Michaels for constant support, advice, and encouragement. And I thank Murray Kiteley, John Connolly, Nalini Bhushan, Kathryn Addelson, Elizabeth Spellman, Frederique Marglin, Lee Bowie, Tom Wartenburg, Vere Chappell, Gareth Matthews, and John Robison, as well as Dan Lloyd, Steve Horst, and Joe Rouse. Thanks under this head also go to many of my nonphilosopher colleagues in the Hampshire College Cultural Studies program. I single out Mary Russo, Joan Landes, Susan Douglas, Jeffery Wallen, Norman Holland, and L. Brown Kennedy.

I also gratefully acknowledge the support of several Hewlett-Mellon faculty development grants from Hampshire College and thank the deans of the college for supporting this work so generously. I am also grateful for the support of this project and of related projects involving academic exchange between the American and Tibetan academic communities from President Greg Prince of Hampshire College. Thanks also to Ms. Ruth Hammen and Ms. Leni Bowen for regular logistical support, to Mr. Andrew Janiak for his extensive assistance and editorial suggestions in the final stages of manuscript preparation, and to Mr. Shua Garfield and Mr. Jeremy Mage for additional assistance in manuscript preparation and proofreading. Thanks as well to many groups of students in “Convention, Knowledge and Existence: European and Indo-Tibetan Perspectives” for putting up with and helping me to refine my presentation of this text and for my students in Buddhist Philosophy at Mount Holyoke College for working through an earlier draft of this text.

Portions of the translations of and commentaries on Chapters I, II, XIII, and XXIV appeared in Philosophy East and West in Garfield (1990) and (1994). I thank the editors for permission to use that material here. The Tibetan edition of the text is from dGe ’dun grub, dBu ma rtsa shes rtsa ’grel bzhugs (Commentary on Mūlamadhyamakakārikā), Ge Lugs Pa Students’ Welfare Publishing, Central Institute of Higher Tibetan Studies, Sarnath, 1987.

I am more grateful than I could ever express to my family for accompanying me to India for one year, for enduring my absence when I have been in India alone, and for enduring my preoccupation with this and related philosophical projects. I am especially grateful to Blaine Garson, who has shouldered far more than her fair share of parenting and other household responsibilities. Every stage of this project is dependent upon her help, sacrifice, and support.

I hope that I haven’t forgotten anybody.


Contents


Part One

The Text of Mūlamadhyamakakārikā

 
Dedicatory Verses
I Examination of Conditions
II Examination of Motion
III Examination of the Senses
IV Examination of the Aggregates
V Examination of Elements
VI Examination of Desire and the Desirous
VII Examination of the Conditioned
VIII Examination of the Agent and Action
IX Examination of the Prior Entity
X Examination of Fire and Fuel
XI Examination of the Initial and Final Limits
XII Examination of Suffering
XIII Examination of Compounded Phenomena
XIV Examination of Connection
XV Examination of Essence
XVI Examination of Bondage
XVII Examination of Actions and Their Fruits
XVIII Examination of Self and Entities
XIX Examination of Time
XX Examination of Combination
XXI Examination of Becoming and Destruction
XXII Examination of the Tathgata
XXIII Examination of Errors
XXIV Examination of the Four Noble Truths
XXV Examination of Nirvna
XXVI Examination of The Twelve Links
XXVII Examination of Views


Part Two

The Text and Commentary

 Introduction to the Commentary

 Dedicatory Verses
I Examination of Conditions
II Examination of Motion
III Examination of the Senses
IV Examination of the Aggregates
V Examination of Elements
VI Examination of Desire and the Desirous
VII Examination of the Conditioned
VIII Examination of the Agent and Action
IX Examination of the Prior Entity
X Examination of Fire and Fuel
XI Examination of the Initial and Final Limits
XII Examination of Suffering
XIII Examination of Compounded Phenomena
XIV Examination of Connection
XV Examination of Essence
XVI Examination of Bondage
XVII Examination of Actions and Their Fruits
XVIII Examination of Self and Entities
XIX Examination of Time
XX Examination of Combination
XXI Examination of Becoming and Destruction
XXII Examination of the Tathgata
XXIII Examination of Errors
XXIV Examination of the Four Noble Truths
XXV Examination of Nirvna
XXVI Examination of The Twelve Links
XXVII Examination of Views

 References

 Index



PART ONE

The Text of Mlamadhyamakakrik



Dedicatory Verses


I prostrate to the Perfect Buddha,
The best of teachers, who taught that
Whatever is dependently arisen is
Unceasing, unborn,
Unannihilated, not permanent,
Not coming, not going,
Without distinction, without identity,
And free from conceptual construction.


Chapter I

Examination of Conditions


1. Neither from itself nor from another,
Nor from both,
Nor without a cause,
Does anything whatever, anywhere arise.
2. There are four conditions: efficient condition;
Percept-object condition; immediate condition;
Dominant condition, just so.
There is no fifth condition.
3. The essence of entities
Is not present in the conditions, etc ….
If there is no essence,
There can be no otherness-essence.
4. Power to act does not have conditions.
There is no power to act without conditions.
There are no conditions without power to act.
Nor do any have the power to act.
5. These give rise to those,
So these are called conditions.
As long as those do not come from these.
Why are these not non-conditions?
6. For neither an existent nor a non-existent thing
Is a condition appropriate.
If a thing is non-existent, how could it have a condition?
If a thing is already existent, what would a condition do?
7. When neither existents nor
Non-existents nor existent non-existents are established,
How could one propose a “productive cause?”
If there were one, it would be pointless.
8. An existent entity (mental episode)
Has no object.
Since a mental episode is without an object,
How could there be any percept-condition?
9. Since things are not arisen.
Cessation is not acceptable.
Therefore, an immediate condition is not reasonable.
If something has ceased, how could it be a condition?
10. If things did not exist
Without essence,
The phrase, “When this exists so this will be,”
Would not be acceptable.
11. In the several or united conditions
The effect cannot be found.
How could something not in the conditions
Come from the conditions?
12. However, if a nonexistent effect
Arises from these conditions,
Why does it not arise
From non-conditions?
13. If the effect’s essence is the conditions,
But the conditions don’t have their own essence,
How could an effect whose essence is the conditions
Come from something that is essenceless?
14. Therefore, neither with conditions as their essence,
Nor with non-conditions as their essence are there any effects.
If there are no such effects,
How could conditions or non-conditions be evident?


Chapter II

Examination of Motion


1. What has been moved is not moving.
What has not been moved is not moving.
Apart from what has been moved and what has not been moved,
Movement cannot be conceived.
2. Where there is change, there is motion.
Since there is change in the moving,
And not in the moved or not-moved,
Motion is in that which is moving.
3. How would it be acceptable
For motion to be in the mover?
When it is not moving, it is not acceptable
To call it a mover.
4. For whomever there is motion in the mover,
There could be non-motion
Evident in the mover.
But having motion follows from being a mover.
5. If motion is in the mover.
There would have to be a twofold motion:
One in virtue of which it is a mover,
And one in virtue of which it moves.
6. If there were a twofold motion.
The subject of that motion would be twofold.
For without a subject of motion,
There cannot be motion.

7. If without a mover
It would not be correct to say that there is motion,
Then if there were no motion,
How could there be a mover?
8. Inasmuch as a real mover does not move,
And a non-mover does not move,
Apart from a mover and a non-mover,
What third thing could move?

9. When without motion,
It is unacceptable to call something a mover,
How will it be acceptable
To say that a mover moves?

10. For him from whose perspective a mover moves,
There would be the consequence that
Without motion there could be a mover.
Because a mover moves.

11. If a mover were to move,
There would be a twofold motion:
One in virtue of which he is a mover,
And one in virtue of which the mover moves.

12. Motion does not begin in what has moved,
Nor does it begin in what has not moved.
Nor does it begin in what is moving.
In what, then, does motion begin?

13. Prior to the beginning of motion,
There is no beginning of motion in
The going or in the gone.
How could there be motion in the not-gone?

14. Since the beginning of motion
Cannot be conceived in any way.
What gone thing, what going thing,
And what non-going thing can be posited?

15. Just as a moving thing is not stationary,
A non-moving thing is not stationary.
Apart from the moving and the non-moving,
What third thing is stationary?

16. If without motion
It is not appropriate to posit a mover,
How could it be appropriate to say
That a moving thing is stationary?

17. One does not halt from moving,
Nor from having moved or not having moved.
Motion and coming to rest
And starting to move are similar.

18. That motion just is the mover itself
Is not correct.
Nor is it correct that
They are completely different.

19. It would follow from
The identity of mover and motion
That agent and action
Are identical.

20. It would follow from
A real distinction between motion and mover
That there could be a mover without motion
And motion without a mover.

21. When neither in identity
Nor in difference
Can they be established,
How can these two be established at all?

22. The motion by means of which a mover is manifest
Cannot be the motion by means of which he moves.
He does not exist before that motion,
So what and where is the thing that moves?

23. A mover does not carry out a different motion
From that by means of which he is manifest as a mover.
Moreover, in one mover
A twofold motion is unacceptable.

24. A really existent mover
Doesn’t move in any of the three ways.
A non-existent mover
Doesn’t move in any of the three ways.

25. Neither an entity nor a non-entity
Moves in any of the three ways.
So motion, mover and
And route are non-existent.


Chapter III

Examination of the Senses


1. Seeing, hearing, smelling,
Tasting, touching, and mind
Are the six sense faculties.
Their spheres are the visible objects, etc….

2. That very seeing does not see
Itself at all.
How can something that cannot see itself
See another?

3. The example of fire
Cannot elucidate seeing.
Along with the moved and not-moved and motion
That has been answered.

4. When there is not even the slightest
Nonseeing seer,
How could it makes sense to say
That seeing sees?

5. Seeing itself does not see.
Nonseeing itself does not see.
Through seeing itself
The clear analysis of the seer is understood.

6. Without detachment from vision there is no seer.
Nor is there a seer detached from it.
If there is no seer
How can there be seeing or the seen?

7. Just as the birth of a son is said to occur
In dependence on the mother and father,
So consciousness is said to arise
In dependence on the eye and material form.

8. From the nonexistence of seeing and the seen it follows that
The other four faculties of knowledge do not exist.
And all the aggregates, etc.,
Are the same way.

9. Like the seen, the heard, the smelled,
The tasted, and the touched,
The hearer, sound, etc.,
And consciousness should be understood.


Chapter IV

Examination of the Aggregates


1. Apart from the cause of form,
Form cannot be conceived.
Apart from form,
The cause of form is not seen.

2. If apart from the cause of form, there were form,
Form would be without cause.
But nowhere is there an effect
Without a cause.

3. If apart from form
There were a cause of form,
It would be a cause without an effect.
But there are no causes without effects.

4. When form exists,
A cause of the arising of form is not tenable.
When form is non-existent,
A cause of the arising of form is not tenable.

5. Form itself without a cause
Is not possible or tenable.
Therefore, think about form, but
Do not construct theories about form.

6. The assertion that the effect and cause are similar
Is not acceptable.
The assertion that they are not similar
Is also not acceptable.

7. Feelings, discriminations, and dispositions
And consciousness and all such things
Should be thought of
In the same way as material form.

8. When an analysis is made through emptiness,
If someone were to offer a reply,
That reply will fail, since it will presuppose
Exactly what is to be proven.

9. When an explanation is made through emptiness,
Whoever would find fault with it
Will find no fault, since the criticism will presuppose
Exactly what is to be proven.


Chapter V

Examination of Elements


1. Prior to a characteristic of space
There is not the slightest space.
If it arose prior to the characteristic
Then it would, absurdly, arise without a characteristic
.
2. A thing without a characteristic
Has never existed.
If nothing lacks a characteristic,
Where do characteristics come to be?

3. Neither in the uncharacterized nor in the characterized
Does a characteristic arise.
Nor does it arise
In something different from these two.

4. If characteristics do not appear,
Then it is not tenable to posit the characterized object.
If the characterized object is not posited,
There will be no characteristic either.

5. From this it follows that there is no characterized
And no existing characteristic.
Nor is there any entity
Other than the characterized and the characteristic.

6. If there is no existent thing,
Of what will there be nonexistence?
Apart from existent and nonexistent things
Who knows existence and nonexistence?

7. Therefore, space is not an entity.
It is not a nonentity.
Not characterized, not without character.
The same is true of the other five elements.

8. Fools and reificationists who perceive
The existence and nonexistence
Of objects
Do not see the paci�erefication of objectification.


Chapter VI

Examination of Desire and the Desirous


1. If prior to desire
And without desire there were a desirous one,
Desire would depend on him.
Desire would exist when there is a desirous one.

2. Were there no desirous one, moreover,
Where would desire occur?
Whether or not desire or the desirous one exist.
The analysis would be the same.

3. Desire and the desirous one
Cannot arise together.
In that case, desire and the desirous one
Would not be mutually contingent.

4. In identity there is no simultaneity.
A thing is not simultaneous with itself.
But if there is difference,
Then how would there be simultaneity?

5. If in identity there were simultaneity,
Then it could occur without association.
If in difference there were simultaneity,
It could occur without association.

6. If in difference there were simultaneity,
How could desire and the desirous one,
Being different, be established?
If they were, they would be simultaneous.

7. If desire and the desirous one
Are established as different,
Then why would you think
That they are simultaneous?

8. Since difference is not established,
If you assert that they are simultaneous,
Since they are established as simultaneous,
Do you also assert that they are different?

9. Since nothing different has been established,
If one is asserting simultaneity,
Which different thing
Do you want to say is simultaneous?

10. Thus desire and the desirous one
Cannot be established as simultaneous or not simultaneous.
So, like desire, nothing whatever
Can be established either as simultaneous or as nonsimultaneous.


Chapter VII

Examination of the Conditioned


1. If arising were produced,
Then it would also have the three characteristics.
If arising is not produced,
How could the characteristics of the produced exist?

2. If the three, arising, etc., are separate,
They cannot function as the characteristics of the produced.
But how could they be joined
In one thing simultaneously?

3. If arising, abiding, and ceasing
Have characteristics other than those of the produced,
There would be an infinite regress.
If they don’t, they would not be produced.

4. The arising of arising only gives rise
To the basic arising.
The arising of the basic arising
Gives rise to arising.

5. If, as you say, the arising of arising
Gives rise to the basic arising,
How, according to you, does this,
Not arisen from the basic arising, give rise to that?

6. If, as you say, that which is arisen from basic arising
Gives rise to the basis,
How does that nonarisen basis
Give rise to it?

7. If this nonarisen
Could give rise to that,
Then, as you wish,
It will give rise to that which is arising.

8. Just as a butterlamp
Illuminates itself as well as others,
So arising gives rise to itself
And to other arisen things.

9. In the butterlamp and its place,
There is no darkness.
What then does the butterlamp illuminate?
For illumination is the clearing of darkness.

10. If the arising butterlamp
Does not reach darkness,
How could that arising butterlamp
Have cleared the darkness?

11. If the illumination of darkness occurs
Without the butterlamp reaching darkness,
All of the darkness in the world
Should be illuminated.

12. If, when it is illuminated,
The butterlamp illuminates itself and others,
Darkness should, without a doubt,
Conceal itself and others.

13. How could this arising, being nonarisen,
Give rise to itself?
And if it is arisen from another,
Having arisen, what is the need for another arising?

14. The arisen, the nonarisen, and that which is arising
Do not arise in any way at all.
Thus they should be understood
Just like the gone, the not-gone, and the going.

15. When there is arising but not yet
That which is arising,
How can we say that that which is arising
Depends on this arising?

16. Whatever is dependently arisen,
Such a thing is essentially peaceful.
Therefore that which is arising and arising itself
Are themselves peaceful.

17. If a nonarisen entity
Anywhere exists,
That entity would have to arise.
But if it were nonexistent, what could arise?

18. If this arising
Gave rise to that which is arising,
By means of what arising
Does that arising arise?

19. If another arising gives rise to this one,
There would be an infinite regress.
If something nonarisen is arisen,
Then all things could arise in this way.

20. Neither an existent nor a nonexistent
Can be properly said to arise.
As it is taught before with
“For neither an existent nor a nonexistent.”

21. The arising of a ceasing thing
Is not tenable.
But to say that it is not ceasing
Is not tenable for anything.

22. A static existent does not endure.
A nonstatic existent does not endure.
Stasis� Csti does not endure.
What nonarisen can endure?

23. The endurance of a ceasing entity
Is not tenable.
But to say that it is not ceasing
Is not tenable for anything.

24. Inasmuch as the nature of all things
Is aging and death,
Without aging and death,
What existents can endure?

25. Stasis cannot endure through itself
Or through another stasis.
Just as arising cannot arise from itself
Or from another arising.

26. The ceasing of what has ceased does not happen.
What has not yet ceased does not cease.
Nor does that which is ceasing.
What nonarisen can cease?

27. The cessation of what is static
Is not tenable.
Nor is the cessation of
Something not static tenable.

28. Being static does not cease
Through being static itself.
Nor does being static cease
Through another instance of being static.

29. When the arising of any entity
Is not tenable.
Then the cessation of any entity
Is not tenable.

30. For an existent thing
Cessation is not tenable.
A single thing being an entity and
A nonentity is not tenable.

31. Moreover, for a� Creo nonentity,
Cessation would be untenable.
Just as a second beheading
Cannot be performed.

32. Cessation does not cease by means of itself.
Nor does it cease by means of another.
Just as arising cannot arise from itself
Or from another arising.

33. Since arising, ceasing, and abiding
Are not established, there are no compounded things.
If all compounded things are unestablished,
How could the uncompounded be established?

34. Like a dream, like an illusion,
Like a city of Gandharvas,
So have arising, abiding,
And ceasing been explained.


Chapter VIII

Examination of the Agent and Action


1. This existent agent
Does not perform an existent action.
Nor does some nonexistent agent
Perform some nonexistent action.

2. An existent entity has no activity.
There would also be action without an agent.
An existent entity has no activity.
There would also be agent without action.
3. If a nonexistent agent
Were to perform a nonexistent action,
Then the action would be without a cause
And the agent would be without a cause.

4. Without a cause, the effect and
Its cause will not occur.
Without this, activity and
Agent and action are not possible.

5. If activity, etc., are not possible,
Entities and nonentities are not possible.
If there are neither entities nor nonentities,
Effects cannot arise from them.

6. If there are no effects, liberation and
Paths to higher realms will not exist.
So all of activity
Would be without purpose.

7. An existent and nonexistent agent
Does not perform an existent and nonexistent action.
Existence and nonexistence cannot pertain to the same thing.
For how could they exist together?
8. An actual agent
Does not perform a nonactual action.
Nor by a nonactual one is an actual one performed.
From this, all of those errors would follow.
9. An existent agent
Does not perform an action that
Is unreal or both real and unreal
As we have already agreed.
10. A nonexistent agent
Does not perform an action that
Is unreal or both real and unreal
As we have already agreed.
11. An existent and nonexistent agent
does not perform an action that
Is unreal or both real and unreal
As we have agreed.
12. Action depends upon the agent.
The agent itself depends on action.
One cannot see any way
To establish them differently.
13. From this elimination of� Klim agent and action,
One should elucidate appropriation in the same way.
Through action and agent
All remaining things should be understood.


Chapter IX

Examination of the Prior Entity


1. Since sight and hearing, etc., and
Feeling, etc., exist,
He who has and uses them
Must exist prior to those, some say.
2. If there were no existent thing,
How could seeing, etc., arise?
It follows from this that prior to this,
there is an existent thing.
3. How is an entity existing prior to
Seeing, hearing, etc., and
The felt, etc.,
Itself known?
4. If it can abide
Without the seen, etc.,
Then, without a doubt,
They can abide without it.
5. Someone is disclosed by something.
Something is disclosed by someone.
Without something how can someone exist?
Without someone how can something exist?
6. While prior to all of seeing, etc.,
That prior entity doesn’t exist,
Through seeing, etc., by another one,
That other one becomes disclosed.
7. If prior to all of seeing, etc.,
No prior entity exists,
How could an entity prior
To each seeing exist?
8. If the seer itself is t� Nhe hearer itself,
And the feeler itself, at different times,
Prior to each of these he would have to arise.
But this makes no sense.
9. If the seer itself is distinct,
The hearer is distinct and the feeler is distinct,
Then when there is a seer there would also be a hearer,
And there would have to be many selves.
10. Seeing and hearing, etc.,
And feeling, etc.,
And that from which these are arisen:
There is no existent there.
11. Seeing and hearing, etc.,
And feeling, etc.,
If that to which they belong does not exist,
they themselves do not exist.
12. For whomever prior to.
Simultaneous with, or after seeing, etc., there is nothing,
For such a one, assertions like “it exists” or “it does not exist”—
Such conceptions will cease.


Chapter X

Examination of Fire and Fuel


1. If fuel were fire
Then agent and action would be one.
If fire were different from fuel,
Then it could arise without fuel.
2. It would be forever aflame;
Flames could be ignited without a cause.
Its beginning would be meaningless.
In that case, it would be without any action.
3. Since it would not depend on another
Ignition would be without a cause.
If it were eternally in flames,
Starting it would be meaningless.
4. So, if one thinks that
That which is burning is the fuel,
If it is just this,
How is this fuel being burned?
5. If they are different, and if one not yet connected isn’t connected,
The not yet burned will not be burned.
They will not cease. If they do not cease
Then it will persist with its own characteristic.
6. Just as a man and a woman
Connect to one another as man and woman,
So if fire were different from fuel,
Fire and fuel would have to be fit for connection.
7. And, if fire and fuel
Preclude each other,
Then fire being different from fuel,
It must still be asserted that they connect.
8. If fire depends on fuel,
And fuel depends on fire,
On what are fire and fuel established as dependent?
Which one is established first?
9. If fire depends on fuel,
It would be the establishment of an established fire.
And the fuel could be fuel
Without any fire.
10. If that on which an entity depends
Is established on the basis
Of the entity depending on it,
What is established in dependence on what?
11. What entity is established through dependence?
If it is not established, then how could it depend?
However, if it is established merely through dependence,
� [gn=That dependence makes no sense.
12. Fire is not dependent upon fuel.
Fire is not independent of fuel.
Fuel is not dependent upon fuel.
Fuel is not independent of fire.
13. Fire does not come from something else,
Nor is fire in fuel itself.
Moreover, fire and the rest are just like
The moved, the not-moved, and the goer.
14. Fuel is not fire.
Fire does not arise from anything different from fuel.
Fire does not possess fuel.
Fuel is not in fire, nor vice versa.
15. Through discussion of fire and fuel,
The self and the aggregates, the pot and cloth
All together,
Without remainder have been explained.
16. I do not think that
Those who teach that the self
Is the same as or different from the entities
Understand the meaning of the doctrine.


Chapter XI

Examination of the Initial and Final Limits


1. When asked about the beginning,
The Great Sage said that nothing is known of it.
Cyclic existence is without end and beginning.
So there is no beginning or end.
2. Where there is no beginning or end,
How could there be a middle?
It follows that thinking about this in terms of
Prior, posterior, and simultaneous is not appropriate.
3. If birth came first,
And then old age and death,
Then birth would be ageless and deathless,
And a deathless one would be born.
4. If birth were to come after,
And old age and death first,
How could there be a causeless aging and death
Of one not born?
5. Birth and age and death
Cannot occur at one time.
Then what is being born would be dying
And both would occur without cause.
6. When the series of the prior, simultaneous, and posterior
Is not possible,
Why are you led to posit
This birth, aging, and death?
7. Not only is cyclic existence itself without beginning,
No existent has a beginning:
Neither cause and effect;
Nor character and characterized …
8. Nor feeling and the feeler;
Whatever there is:
All entities
Are without beginning.


Chapter XII

Examination of Suffering


1. Some say suffering is self-produced,
Or produced from another or from both.
Or that it arises without a cause.
It is not the kind of thing to be produced.
2. If suffering came from itself,
Then it would not arise dependently.
For those aggregates
Arise in dependence on these aggregates.
3. If those were different from these,
Or if these were different from those,
Suffering could arise from another.
These would arise from those others.
4. If suffering were caused by a person himself,
Then who is that person
By whom suffering is caused
Who exists distinct from suffering?
5. If suffering comes from another person,
Then who is that person
When suffering is given by another—
Who exists distinct from suffering?
6. If another person causes suffering,
Who is that other one
Who bestowed that suffering,
Distinct from suffering?
7. When self-caused is not established,
How could suffering be caused by another?
Whoever caused the suffering of another
Must have caused his own suffering.
8. No suffering is self-caused.
Nothing causes itself.
If another is not self-made,
How could suffering be caused by another?
9. If suffering were caused by each,
Suffering could be caused by both.
Not caused by self or by other,
How could suffering be uncaused?
10. Not only does suffering not exist
In any of the fourfold ways:
No external entity exists
In any of the fourfold ways.


Chapter XIII

Examination of Compounded Phenomena


1. The Victorious Conqueror has said that whatever
Is deceptive is false.
Compounded phenomena are all deceptive.
Therefore they are all false.
2. If whatever is deceptive is false,
What deceives?
The Victorious Conqueror has said about this
That emptiness is completely true.
3. All things lack entitihood,
Since change is perceived.
There is nothing without entity
Because all things have emptiness.
4. If there is no entitihood,
What changes?
If there were entity,
How could it be correct that something changes?
5. A thing itself does not change.
Something different does not change.
Because a young man doesn’t grow old,
And because and an old man doesn’t grow old either.
6. If a thing itself changed,
Milk itself would be curd.
Or curd would have come to be
An entity different from milk.
7. If there were even a trifle nonempty,
Emptiness itself would be but a trifle.
But not even a trifle is nonempty.
How could emptiness be an entity?
8. The victorious ones have said
That emptiness is the relinquishing of all views.
For whomever emptiness is a view,
That one will accomplish nothing.


Chapter XIV

Examination of Connection


1. The seen, seeing, and the seer:
These three—pairwise or
All together—
Do not connect to one another.
2. Similarly desire, the desirous one, the object of desire,
And the remaining afflictions
And the remaining sources of perception
Are understood in this threefold way.
3. Since different things connect to one another,
But in seeing, etc.,
There is no difference,
They cannot connect.
4. Not only in seeing, etc.,
Is there no such difference:
When one thing and another are simultaneous,
It is also not tenable that there is difference.
5. A different thing depends on a different thing for its difference.
Without a different thing, a different thing wouldn’t be different.
It is not tenable for that which depends on something else
To be different from it.
6. If a different thing were different from a different thing,
Without a different thing, a different thing could exist.
But without that different thing, that different thing does not exist.
It follows that it doesn’t exist.
7. Difference is not in a different thing.
Nor is it in a nondifferent thing.
If difference does not exist,
Neither different nor identical things exist.
8. That does not connect to itself.
Nor do different things connect to one another.
Neither connection nor
Connected nor connector exist.


Chapter XV

Examination of Essence


1. Essence arising from
Causes and conditions makes no sense.
If essence came from causes and conditions,
Then it would be fabricated.
2. How could it be appropriate
For fabricated essence to come to be?
Essence itself is not artificial
And does not depend on another.
3. If there is no essence,
How can there be difference in entities?
The essence of difference in entities
Is what is called the entity of difference.
4. Without having essence or otherness-essence,
How can there be entities?
If there are essences and entities
Entities are established.
5. If the entity is not established,
A nonentity is not established.
An entity that has become different.
Is a nonentity, people say.
6. Those who see essence and essential difference
And entities and nonentities,
They do not see
The truth taught by the Buddha.
7. The Victorious One, through knowledge
Of reality and unreality,
In the Discourse to Katyyna,
Refuted both “it is” and “it is not.”
8. If existence were through essence,
Then there would be no nonexistence.
A change in essence
Could never be tenable.
9. If there is no essence,
What could become other?
If there is essence,
What could become other?
10. To say “it is” is to grasp for permanence.
To say “it is not” is to adopt the view of nihilism.
Therefore a wise person
Does not say “exists” or “does not exist.”
11. “Whatever exists through its essence
Cannot be nonexistent” is eternalism.
“It existed before but doesn’t now”
Entails the error of nihilism.


Chapter XVI

Examination of Bondage


1. If compounded phenomena transmigrate,
They do not transmigrate as permanent.
If they are impermanent they do not transmigrate.
The same approach applies to sentient beings.
2. If someone transmigrates,
Then if, when sought in the fivefold way
In the aggregates and in the sense spheres and in the elements,
He is not there, what transmigrates?
3. If one transmigrates from grasping to grasping, then
One would be nonexistent.
Neither existent nor grasping,
Who could this transmigrator be?
4. How could compounded phenomena pass into nirva?
That would not be tenable.
How could a sentient being pass into nirva?
That would not be tenable.
5. All compounded phenomena, as arising and ceasing things,
Are not bound and not released.
For this reason a sentient being
Is not bound, not released.
6. If grasping were bondage,
Then the one who is grasping would not be bound.
But one who is not grasping is not bound.
In what circumstances will one be bound?
7. If prior to binding
There is a bound one,
There would be bondage, but there isn’t.
The rest has been explained by the gone, the not-gone, and the goer.
8. Whoever is bound is not released.
Whoever is not bound does not get released.
If a bound one were being released,
Bondage and release would occur simultaneously.
9. “I, without grasping, will pass beyond sorrow,
And I will attain nirva,” one says.
Whoever grasps like this
Has a great grasping.
10. When you can’t bring about nirva,
Nor the purification of cyclic existence,
What is cyclic existence,
And what is the nirva you examine?


Chapter XVII

Examination of Actions and Their Fruits


1. Self-restraint and benefiting others
With a compassionate mind is the Dharma.
This is the seed for
Fruits in this and future lives.
2. The Unsurpassed Sage has said
That actions are either intention or intentional.
The varieties of these actions
Have been announced in many ways.
3. Of these, what is called “intention
Is mental desire.
What is called “intentional”
Comprises the physical and verbal.
4. Speech and action and all
Kinds of unabandoned and abandoned actions,
And resolve
As well as …
5. Virtuous and nonvirtuous actions
Derived from pleasure,
As well as intention and morality:
These seven are the kinds of action.
6. If until the time of ripening
Action had to remain in place, it would have to be permanent.
If it has ceased, then having ceased,
How will a fruit arise?
7. As for a continuum, such as the sprout,
It comes from a seed.
From that arises the fruit. Without a seed,
It would not come into being.
8. Since from the seed comes the continuum,
and from the continuum comes the fruit,
The seed precedes the fruit.
Therefore there is neither nonexistence nor permanence.
9. So, in a mental continuum,
From a preceding intention
A consequent mental state arises.
Without this, it would not arise.
10. Since from the intention comes the continuum,
And from the continuum the fruit arises,
Action precedes the fruit.
Therefore there is neither nonexistence nor permanence.
11. The ten pure paths of action
Are the method of realizing the Dharma.
These fruits of the Dharma in this and other lives
Are the five pleasures.
12. If such an analysis were advanced,
There would be many great errors.
Therefore, this analysis
Is not tenable here.
13. I will then explain what is tenable here:
The analysis propounded by all
Buddhas, self-conquerors
And disciples according to which …
14. Action is like an uncancelled promissory note
And like a debt.
Of the realms it is fourfold.
Moreover, its nature is neutral.
15. By abandoning, that is not abandoned.
Abandonment occurs through meditation.
Therefore, through the nonexpired,
The fruit of action arises.
16. If abandonment occurred through abandoning, and
If action were destroyed through transformation,
The destruction of action, etc.,
And other errors would arise.
17. From all these actions in a realm,
Whether similar or dissimilar,
At the moment of birth
Only one will arise.
18. In this visible world,
All actions of the two kinds,
Each comprising action and the unexpired separately,
Will remain while ripening.
19. That fruit, if extinction or death
Occurs, ceases.
Regarding this, a distinction between the stainless
And the stained is drawn.
20. Emptiness and nonannihilation;
Cyclic existence and nonpermanence:
That action is nonexpiring
Is taught by the Buddha.
21. Because action does not arise,
It is seen to be without essence.
Because it is not arisen,
It follows that it is nonexpiring.
22. If action had an essence,
It would, without doubt, be eternal.
Action would be uncreated.
Because there can be no creation of what is eternal.
23. If an action were uncreated,
Fear would arise of encountering something not done.
And the error of not preserving
One’s vows would arise.
24. All conventions would then
Be contradicted, without doubt.
It would be impossible to draw a distinction
Between virtue and evil.
25. Whatever is mature would mature
Time and time again.
If there were essence, this would follow,
Because action would remain in place.
26. While this action has affliction as its nature
This affliction is not real in itself.
If affliction is not in itself,
How can action be real in itself?
27. Action and affliction
Are taught to be the conditions that produce bodies.
If action and affliction
Are empty, what would one say about bodies?
28. Obstructed by ignorance,
And consumed by passion, the experiencer
Is neither different from the agent
Nor identical with it.
29. Since this action
Is not arisen from a condition,
Nor arisen causelessly,
It follows that there is no agent.
30. If there is no action and agent,
Where could the fruit of action be?
Without a fruit,
Where is there an experiencer?
31. Just as the teacher, by magic,
Makes a magical illusion, and
By that illusion
Another illusion is created,
32. In that way are an agent and his action:
The agent is like the illusion.
The action
Is like the illusion’s illusion.
33. Afflictions, actions, bodies,
Agents and fruits are
Like a city of Gandharvas and
Like a mirage or a dream.



 
Chapter XVIII

Examination of Self and Entities


1. If the self were the aggregates,
It would have arising and ceasing (as properties).
If it were different from the aggregates,
It would not have the characteristics of the aggregates.
2. If there were no self,
Where would the self’s (properties) be?
From the pacification of the self and what belongs to it,
One abstains from grasping onto “I” and “mine”.
3. One who does not grasp onto “I” and “mine,”
That one does not exist.
One who does not grasp onto “I” and “mine,”
He does not perceive.
4. When views of “I” and “mine” are extinguished,
Whether with respect to the internal or external,
The appropriator ceases.
This having ceased, birth ceases.
5. Action and misery having ceased, there is nirva.
Action and misery come from conceptual thought.
This comes from mental fabrication.
Fabrication ceases through emptiness.
6. That there is a self has been taught,
And the doctrine of no-self,
By the buddhas, as well as the
Doctrine of neither self nor nonself.
7. What language expresses is nonexistent.
The sphere of thought is nonexistent.
Unarisen and unceased, like nirva
Is the nature of things.
8. Everything is real and is not real,
Both real and not real,
Neither real nor not real.
This is Lord Buddha’s teaching.
9. Not dependent on another, peaceful and
Not fabricated by mental fabrication,
Not thought, without distinctions,
That is the character of reality (that-ness).
10. Whatever comes into being dependent on another
Is not identical to that thing.
Nor is it different from it.
Therefore it is neither nonexistent in time nor permanent.
11. By the buddhas, patrons of the world,
This immortal truth is taught:
Without identity, without distinction;
Not nonexistent in time, not permanent.
12. When the fully enlightened ones do not appear,
And when the disciples have disappeared,
The wisdom of the self-enlightened ones
Will arise completely without a teacher.


Chapter XIX

Examination of Time


1. If the present and the future
Depend on the past,
Then the present and the future
Would have existed in the past.
2. If the present and the future
Did not exist there,
How could the present and the future
Be dependent upon it?
3. If they are not dependent upon the past,
Neither of the two would be established.
Therefore neither the present
Nor the future would exist.
4. By the same method,
The other two divisions—past and future,
Upper, lower, middle, etc.,
Unity, etc., should be understood.
5. A nonstatic time is not grasped.
Nothing one could grasp as
Stationary time exists.
If time is not grasped, how is it known?
6. If time depends on an entity,
Then without an entity how could time exist?
There is no existent entity.
So how can time exist?


Chapter XX

Examination of Combination


1. If, arising from the combination of
Causes and conditions,
The effect is in the combination,
How could it arise from the combination?
2. If, arising from the combination of
Causes and conditions,
The effect is not in the combination,
How could it arise from the combination?
3. If the effect is in the combination
Of causes and conditions,
Then it should be grasped in the combination.
But it is not grasped in the combination.
4. If the effect is not in the combination
Of causes and conditions,
Then actual causes and conditions
Would be like noncauses and nonconditions.
5. If the cause, in having its effect,
Ceased to have its causal status,
There would be two kinds of cause:
With and without causal status.
6. If the cause, not yet ��having
Produced its effect, ceased,
Then having arisen from a ceased cause,
The effect would be without a cause.
7. If the effect were to arise
Simultaneously with the collection,
Then the produced and the producer
Would arise simultaneously.
8. If the effect were to arise
Prior to the combination,
Then, without causes and conditions,
The effect would arise causelessly.
9. If, the cause having ceased, the effect
Were a complete transformation of the cause,
Then a previously arisen cause
Would arise again.
10. How can a cause, having ceased and dissolved,
Give rise to a produced effect?
How can a cause joined with its effect produce it
If they persist together?
11. Moreover, if not joined with its cause,
What effect can be made to arise?
Neither seen nor unseen by causes
Are effects produced.
12. There is never a simultaneous connection
Of a past effect
With a past, a nonarisen,
Or an arisen cause.
13. There is never a simultaneous connection
Of an arisen effect
With a past, a nonarisen,
Or an arisen cause.
14. There is never a simultaneous connection
Of a nonarisen effect
With a past, a nonarisen,
Or an arisen cause.
15. Without connecting,
How can a cause produce an effect?
Where there is connection,
How can a cause produce an effect?
16. If the cause is empty of an effect,
How can it produce an effect?
If the cause is not empty of an effect,
How can it produce an effect?
17. A nonempty effect does not arise.
The nonempty would not cease.
This nonempty would be
The nonceased and the nonarisen.
18. How can the empty arise?
How can the empty cease?
The empty will hence also
Be the nonceased and nonarisen.
19. For cause and effect to be identical
Is not tenable.
For cause and effect to be different
Is not tenable.
20. If cause and effect were identical,
Produced and producer would be identical.
If cause and effect were different,
Cause and non-cause would be alike.
21. If an effect had entitihood,
What could have caused it to arise?
If an effect had no entitihood,
What could have caused it to arise?
22. If something is not producing an effect,
It is not tenable to attribute causality.
If it is not tenable to attribute causality,
Then of what will the effect be?
23. If the combination
Of causes and conditions
Is not self-produced,
How does it produce an effect?
24. Therefore, not made by combination,
And not without a combination can the effect arise.
If there is no effect,
Where can there be a combination of conditions?


Chapter XXI

Examination of Becoming and Destruction


1. Destruction does not occur without becoming.
It does not occur together with it.
Becoming does not occur without destruction.
It does not occur together with it.
2. How could there be destruction
Without becoming?
How could there be death without birth?
There is no destruction without becoming.
3. How could destruction and becoming
Occur simultaneously?
Death and birth
Do not occur simultaneously.
4. How could there be becoming
Without destruction?
For impermanence
Is never absent from entities.
5. How could destruction
And becoming occur simultaneously?
Just as birth and death
Do not occur simultaneously.
6. How, when things cannot
Be established as existing,
With, or apart from one another,
Can th��ey be established at all?
7. There is no becoming of the disappeared.
There is no becoming of the nondisappeared.
There is no destruction of the disappeared.
There is no destruction of the nondisappeared.
8. When no entities exist,
There is no becoming or destruction.
Without becoming and destruction,
There are no existent entities.
9. It is not tenable for the empty
To become or to be destroyed,
It is not tenable for the nonempty
To become or to be destroyed.
10. It is not tenable
That destruction and becoming are identical.
It is not tenable
That destruction and becoming are different.
11. If you think you see both
Destruction and becoming,
Then you see destruction and becoming
Through impaired vision.
12. An entity does not arise from an entity.
An entity does not arise from a nonentity.
A nonentity does not arise from a nonentity.
A nonentity does not arise from an entity.
13. An entity does not arise from itself.
It is not arisen from another.
It is not arisen from itself and another.
How can it be arisen?
14. If one accepts the existence of entities,
Permanence and the view of complete nonexistence follow.
For these entities
Must be both permanent and impermanent.
15. If one accepts the existence of entities
Nonexistence and permanence will not follow.
Cyclic existence is the continuous
Becoming and destruction of causes and effects.
16. If cyclic existence is the continuous
Becoming and destruction of causes and effects,
Then from the nonarising of the destroyed
Follows the nonexistence of cause.
17. If entities exist with entitihood,
Then their nonexistence would make no sense.
But at the time of nirva,
Cyclic existence ceases completely, having been pacified.
18. If the final one has ceased,
The existence of a first one makes no sense.
If the final one has not ceased,
The existence of a first one makes no sense.
19. If when the final one was ceasing,
Then the first was arising,
The one ceasing would be one.
The one arising would be another.
20. If, absurdly, the one arising
And the one ceasing were the same,
Then whoever is dying with the aggregates
Is also arising.
21. Since the series of cyclic existence is not evident
In the three times,
If it is not in the three times,
How could there be a series of cyclic existence?


Chapter XXII

Examination of the Tathgata


1. Neithe��r the aggregates, nor different from the aggregates,
The aggregates are not in him, nor is he in the aggregates.
The Tathgata does not possess the aggregates.
What is the Tathgata?
2. If the Buddha depended on the aggregates,
He would not exist through an essence.
Not existing through an essence,
How could he exist through otherness-essence?
3. Whatever is dependent on another entity,
Its selfhood is not appropriate.
It is not tenable that what lacks a self
Could be a Tathgata.
4. If there is no essence,
How could there be otherness-essence?
Without possessing essence or otherness-essence,
What is the Tathgata?
5. If without depending on the aggregates
There were a Tathgata,
Then now he would be depending on them.
Therefore he would exist through dependence.
6. Inasmuch as there is no Tathgata
Dependent upon the aggregates,
How could something that is not dependent
Come to be so?
7. There is no appropriation.
There is no appropriator.
Without appropriation
How can there be a Tathgata?
8. Having been sought in the fivefold way,
What, being neither identical nor different,
Can be thought to be the Tathgata
Through grasping?
9. Whatever grasping there is
Does not exist through essence.
And when something does not exist through itself,
It can never exist through otherness-essence.
10. Thus grasping and grasper
Together are empty in every respect.
How can an empty Tathgata
Be known through the empty?
11. “Empty” should not be asserted.
Nonempty” should not be asserted.
Neither both nor neither should be asserted.
They are only used nominally.
12. How can the tetralemma of permanent and impermanent, etc.,
Be true of the peaceful?
How can the tetralemma of finite, infinite, etc.,
Be true of the peaceful?
13. One who grasps the view that the Tathgata exists,
Having seized the Buddha,
Constructs conceptual fabrications
About one who has achieved Nirva.
14. Since he is by nature empty,
The thought that the Buddha
Exists or does not exist
After nirva is not appropriate.
15. Those who develop mental fabrications with regard to the Buddha,
Who has gone beyond all fabrications,
As a consequence of those cognitive fabrications,
Fail to see the Tathgata.
16. Whatever is the essence of the Tathgata,
That is the essence of the world.
The Tathgata has no essence.
The world is without essence.


Chapter XXIII

Examination of Errors


1. Desire, hatred and confusion all
Arise from thought, it is said.
They all depend on
The pleasant, the unpleasant, and errors.
2. Since whatever depends on the pleasant and the unpleasant
Does not exist through an essence,
The defilements
Do not really exist.
3. The self’s existence or nonexistence
Has in no way been established.
Without that, how could the defilements
Existence or nonexistence be established?
4. The defilements are somebody’s.
But that one has not been established.
Without that possessor,
The defilements are nobody’s.
5. View the defilements as you view your self:
They are not in the defiled in the fivefold way.
View the defiled as you view your self:
It is not in the defilements in the fivefold way.
6. The pleasant, the unpleasant, and the errors
Do not exist through essence.
Which pleasant, unpleasant, and errors
could the defilements depend upon?
7. Form, sound, taste, touch,
Smell, and concepts of things: These six
Are thought of as the foundation of
Desire, hatred, and confusion.
8. Form, sound, taste, touch,
Smell, and concepts of things: These six
Should be seen as only like a city of the Gandharvas and
Like a mirage or a dream.
9. How could the
Pleasant and unpleasant arise
In those that are like an illusory person
And like a reflection?
10. We say that the unpleasant
Is dependent upon the pleasant,
Since without depending on the pleasant there is none.
It follows that the pleasant is not tenable.
11. We say that the pleasant
Is dependent upon the unpleasant.
Without the unpleasant there wouldn’t be any.
It follows that the unpleasant is not tenable.
12. Where there is no pleasant,
How can there be desire?
Where there is no unpleasant,
How can there be anger?
13. If to grasp onto the view
“The impermanent is permanent” were an error,
Since in emptiness there is nothing impermanent,
How could that grasping be an error?
14. If to grasp onto the view
“The impermanent is permanent” were an error,
Why isn’t grasping onto the view
�� vie�jus�In emptiness there is nothing impermanent” an error?
15. That by means of which there is grasping, and the grasping,
And the grasper, and all that is grasped:
All are being relieved,
It follows that there is no grasping.
16. If there is no grasping,
Whether erroneous or otherwise,
Who will come to be in error?
Who will have no error?
17. Error does not develop
In one who is in error.
Error does not develop
In one who is not in error.
18. Error does not develop
In one in whom error is arising.
In whom does error develop?
Examine this on your own!
19. If error is not arisen,
How could it come to exist?
If error has not arisen,
How could one be in error?
20. Since an entity does not arise from itself,
Nor from another,
Nor from another and from itself,
How could one be in error?
21. If the self and the pure,
The permanent and the blissful existed,
The self, the pure, the permanent,
And the blissful would not be deceptive.
22. If the self and the pure,
The permanent and the blissful did not exist,
The nonself, the impure, the permanent,
And suffering would not exist.
23. Thus, through the cessation of e��cesrror
Ignorance ceases.
When ignorance ceases
The compounded phenomena, etc., cease.
24. If someone’s defilements
Existed through his essence,
How could they be relinquished?
Who could relinquish the existent?
25. If someone’s defilements
Did not exist through his essence,
How could they be relinquished?
Who could relinquish the nonexistent?


Chapter XXIV

Examination of the Four Noble Truths


1. If all of this is empty,
Neither arising nor ceasing,
Then for you, it follows that
The Four Noble Truths do not exist.
2. If the Four Noble Truths do not exist,
Then knowledge, abandonment,
Meditation and manifestation
Will be completely impossible.
3. If these things do not exist,
The four fruits will not arise,
Without the four fruits, there will be no attainers of the fruits.
Nor will there be the faithful.
4. If so, the spiritual community will not exist.
Nor will the eight kinds of person.
If the Four Noble Truths do not exist,
There will be no true Dharma.
5. If there is no doctrine and spiritual community,
How can there be a Buddha?
If emptiness is conceived in this way,
The three jewels are contradicted.
6. Hence you assert that there are no real fruits.
And no Dharma. The Dharma itself
And the conventional truth
Will be contradicted.
7. We say that this understanding of yours
Of emptiness and the purpose of emptiness
And of the significance of emptiness is incorrect.
As a consequence you are harmed by it.
8. The Buddha’s teaching of the Dharma
Is based on two truths:
A truth of worldly convention
And an ultimate truth.
9. Those who do not understand
The distinction drawn between these two truths
Do not understand
The Buddha’s profound truth.
10. Without a foundation in the conventional truth,
The significance of the ultimate cannot be taught.
Without understanding the significance of the ultimate,
Liberation is not achieved.
11. By a misperception of emptiness
A person of little intelligence is destroyed.
Like a snake incorrectly seized
Or like a spell incorrectly cast.
12. For that reason—that the Dharma is
Deep and difficult to understand and to learn—
The Buddha’s mind despaired of
Being able to teach it.
13. You have presented fallacious refutations
That are not relevant to emptiness.
Your confusion about emptiness
Does not belong to me.
14. For him to whom emptiness is clear,
Everything becomes clear.
For him to whom emptiness is not clear,
Nothing becomes clear.
15. When you foist on us
All of your errors
You are like a man who has mounted his horse
And has forgotten that very horse.
16. If you perceive the existence of all things
In terms of their essence,
Then this perception of all things
Will be without the perception of causes and conditions.
17. Effects and causes
And agent and action
And conditions and arising and ceasing
And effects will be rendered impossible.
18. Whatever is dependently co-arisen
That is explained to be emptiness.
That, being a dependent designation,
Is itself the middle way.
19. Something that is not dependently arisen,
Such a thing does not exist.
Therefore a nonempty thing
Does not exist.
20. If all this were nonempty, as in your view,
There would be no arising and ceasing.
Then the Four Noble Truths
Would become nonexistent.
21. If it is not dependently arisen,
How could suffering come to be?
Suffering has been taught to be impermanent,
And so cannot come from its own essence.
22. If something comes from its own essence,
How could it ever be arisen?
It follows that if one denies emptiness
There can be no arising (of suffering).
23. If suffering had an essence,
Its cessation would not exist.
So if an essence is posited,
One denies cessation.
24. If the path had an essence,
Cultivation would not be appropriate.
If this path is indeed cultivated,
It cannot have an essence.
25. If suffering, arising, and
Ceasing are nonexistent,
By what path could one seek
To obtain the cessation of suffering?
26. If nonunderstanding comes to be
Through its essence,
How will understanding arise?
Isn’t essence stable?
27. In the same way, the activities of
Relinquishing, realizing, and meditating
And the four fruits
Would not be possible.
28. For an essentialist,
Since the fruits through their essence
Are already unrealized,
In what way could one attain them?
29. Without the fruits, there are no attainers of the fruits,
Or enterers. From this it follows that
The eight kinds of persons do not exist.
If these don’t exist, there is no spiritual community.
30. From the nonexistence of the Noble Truths
Would follow the nonexistence of the true doctrine.
�� he
If there is no doctrine and no spiritual community,
How could a Buddha arise?
31. For you, it would follow that a Buddha
Arises independent of enlightenment.
And for you, enlightenment would arise
Independent of a Buddha.
32. For you, one who through his essence
Was unenlightened,
Even by practicing the path to enlightenment
Could not achieve enlightenment.
33. Moreover, one could never perform
Right or wrong actions.
If this were all nonempty what could one do?
That with an essence cannot be produced.
34. For you, from neither right nor wrong actions
Would the fruit arise.
If the fruit arose from right or wrong actions,
According to you, it wouldn’t exist.
35. If, for you, a fruit arose
From right or wrong actions,
Then, having arisen from right or wrong actions,
How could that fruit be nonempty?
36. If dependent arising is denied,
Emptiness itself is rejected.
This would contradict
All of the worldly conventions.
37. If emptiness itself is rejected,
No action will be appropriate.
There would be action which did not begin,
And there would be agent without action.
38. If there is essence, the whole world
Will be unarising, unceasing,
And static. The��nd entire phenomenal world
Would be immutable.
39. If it (the world) were not empty,
Then action would be without profit.
The act of ending suffering and
Abandoning misery and defilement would not exist.
40. Whoever sees dependent arising
Also sees suffering
And its arising
And its cessation as well as the path.


Chapter XXV

Examination of Nirva


1. If all this is empty,
Then there is no arising or passing away.
By the relinquishing or ceasing of what
Does one wish nirva to arise?
2. If all this is nonempty,
Then there is no arising or passing away.
By the relinquishing or ceasing of what
Does one wish nirva to arise?
3. Unrelinquished, unattained,
Unannihilated, not permanent,
Unarisen, unceased:
This is how nirva is described.
4. Nirva is not existent.
It would then have the characteristics of age and death.
There is no existent entity
Without age and death.
5. If nirva were existent.
Nirva would be compounded.
A noncompounded existent
Does not exist anywhere.
6. If nirva were existent,
How could nirva be nondependent?
A nondependent existent
Does not exist anywhere.
7. If nirva were not existent,
How could it be appropriate for it to be nonexistent?
Where nirva is not existent.
It cannot be a nonexistent.
8. If nirva were not existent,
How could nirva be nondependent?
Whatever is nondependent
Is not nonexistent.
9. That which comes and goes
Is dependent and changing.
That, when it is not dependent and changing,
Is taught to be nirva.
10. The teacher has spoken of relinquishing
Becoming and dissolution.
Therefore, it makes sense that
Nirva is neither existent nor nonexistent.
11. If nirva were both
Existent and nonexistent,
Passing beyond would, impossibly,
Be both existent and nonexistent.
12. If nirva were both
Existent and nonexistent,
Nirva would not be nondependent.
Since it would depend on both of these.
13. How could nirva
Be both existent and nonexistent?
Nirva is uncompounded.
Both existents and nonexistents are compounded.
14. How could nirva
Be both existent and nonexistent?
These two cannot be in the same place.
Like light and darkness.
15. Nirva is said to be
Neither existent nor nonexistent.
If the existent and the nonexistent were established,
This would be established.
16. If nirva is
Neither existent nor nonexistent,
Then by whom is it expounded
“Neither existent nor nonexistent”?
17. Having passed into nirva, the Victorious Conqueror
Is neither said to be existent
Nor said to be nonexistent.
Neither both nor neither are said.
18. So, when the victorious one abides, he
Is neither said to be existent
Nor said to be nonexistent.
Neither both nor neither are said.
19. There is not the slightest difference
Between cyclic existence and nirva.
There is not the slightest difference
Between nirva and cyclic existence.
20. Whatever is the limit of nirva,
That is the limit of cyclic existence.
There is not even the slightest difference between them,
Or even the subtlest thing.
21. Views that after cessation there is a limit, etc.,
And that it is permanent, etc.,
Depend upon nirva, the final limit,
And the prior limit.
22. Since all existents are empty,
What is finite or infinite?
What is finite and infinite?
What is neither finite nor infinite?
23. What is identical and what is different?
What is permanent and what is impermanent?
What is both permanent and impermanent?
What is neither?
24. The pacification of all objectification
And the pacification of illusion:
No Dharma was taught by the Buddha
At any time, in any place, to any person.


Chapter XXVI

Examination of the Twelve Links


1. Wrapped in the darkness of ignorance,
One performs the three kinds of actions
Which as dispositions impel one
To continue to future existences.
2. Having dispositions as its conditions,
Consciousness enters transmigration.
Once consciousness has entered transmigration,
Name and form come to be.
3. Once name and form come to be,
The six sense spheres come into being.
Depending on the six sense spheres,
Contact comes into being.
4. That is only dependent
On eye and form and apprehension.
Thus, depending on name and form,
And which produces consciousness
5. That which is assembled from the three—
Eye and form and consciousness,
Is contact. From contact
Feeling comes to be.
6. Conditioned by feeling is craving.
Craving arises because of feeling.
When it appears, there is grasping,
The four spheres of grasping.
7. When there is grasping, the grasper
Comes into existence.
If he did not grasp,
Then being freed, he would not come into existence.
8. This existence is also the five aggregates.
From existence comes birth,
Old age and death and misery and
Suffering and grief and …
9. Confusion and agitation.
All these arise as a consequence of birth.
Thus this entire mass of suffering
Comes into being.
10. The root of cyclic existence is action.
Therefore, the wise one does not act.
Therefore, the unwise is the agent.
The wise one is not because of his insight.
11. With the cessation of ignorance
Action will not arise.
The cessation of ignorance occurs through
Meditation and wisdom.
12. Through the cessation of this and that
This and that will not be manifest.
The entire mass of suffering
Indeed thereby completely ceases.


Chapter XXVII

Examination of Views


1. The views “in the past I was” or “I was not”
And the view that the world is permanent, etc.,
All of these views
Depend on a prior limit.
2. The view “in the future I will become other” or “I will not do so”
And that the world is limited, etc.,
All of these views
Depend on a final limit.
3. To say “I was in the past
Is not tenable.
��What existed in the past
Is not identical to this one.
4. According to you, this self is that,
But the appropriator is different.
If it is not the appropriator,
What is your self?
5. Having shown that there is no self
Other than the appropriator,
The appropriator should be the self.
But it is not your self.
6. Appropriating is not the self.
It arises and ceases.
How can one accept that
Future appropriating is the appropriator?
7. A self that is different
From the appropriating is not tenable.
If it were different, then in a nonappropriator
There should be appropriation. But there isn’t.
8. So it is neither different from the appropriating
Nor identical to the appropriating.
There is no self without appropriation.
But it is not true that it does not exist.
9. To say “in the past I wasn’t”
Would not be tenable.
This person is not different
From whoever existed in previous times.
10. If this one were different,
Then if that one did not exist, I would still exist.
If this were so,
Without death, one would be born.
11. Annihilation and the exhaustion of action would follow;
Different agents’ actions
Would be experienced by each other.
That and other such things would follow.
12. Nothing comes to exist from something that did not exist.
From this errors would arise.
The self would be produced
Or, existing, would be without a cause.
13. So, the views “I existed,” “I didn’t exist,”
Both or neither,
In the past
Are untenable.
14. To say “in the future I will exist or
Will not exist,”
Such a view is like
Those involving the past.
15. If a human were a god,
On such a view there would be permanence.
The god would be unborn.
For any permanent thing is unborn.
16. If a human were different from a god,
On such a view there would be impermanence.
If the human were different from the god,
A continuum would not be tenable.
17. If one part were divine and
One part were human,
It would be both permanent and impermanent.
That would be irrational.
18. If it could be established that
It is both permanent and impermanent,
Then it could be established that
It is neither permanent nor impermanent.
19. If anyone had come from anyplace
And were then to go someplace,
It would follow that cyclic existence was beginningless.
This is not the case.
20. If nothing is permanent,
What will be impermanent,
Permanent and impermanent,
Or neither?
21. If the world were limited,
How could there be another world?
If the world were unlimited,
How could there be another world?
22. Since the continuum of the aggregates
Is like the flame of a butterlamp,
It follows that neither its finitude
Nor its infinitude makes sense.
23. If the previous were disintegrating
And these aggregates, which depend
Upon those aggregates, did not arise,
Then the world would be finite.
24. If the previous were not disintegrating
And these aggregates, which depend
Upon those aggregates, did not arise,
Then the world would be infinite.
25. If one part were finite and
One part were infinite,
Then the world would be finite and infinite.
This would make no sense.
26. How could one think that
One part of the appropriator is destroyed
And one part is not destroyed?
This position makes no sense.
27. How could one think that
One part of the appropriation is destroyed
And one part is not destroyed?
This position makes no sense.
28. If it could be established that
It is both finite and infinite,
Then it could be established that
It is neither finite nor infinite.
29. So, because all entities are empty,
Which views of permanence, etc., would occur,
And to whom, when, why, and about what
Would they occur at all?
30. I prostrate to Gautama
Who through compassion
Taught the true doctrine,
Which leads to the relinquishing of all views.


PART TWO

The Text and Commentary



Introduction to the Commentary


Ngrjuna, who lived in South India in approximately the second century C.E., is undoubtedly the most important, influential, and widely studied Mahyna Buddhist philosopher. He is the founder of the Mdhyamika, or Middle Path schools of Mahyna Buddhism. His considerable corpus includes texts addressed to lay audiences, letters of advice to kings, and the set of penetrating metaphysical and epistemological treatises that represent the foundation of the highly sceptical and dialectical analytic philosophical school known as Mdhyamika. Most important of these is his largest and best known text, Mlamadhyamakakrik (literally Fundamental Verses on the Middle Way). This text in turn inspires a huge commentarial literature in Sanskrit, Tibetan, Chinese, Korean and Japanese. Divergences on interpretation of Mlamadhyamakakrik often determine the splits between major philosophical schools. So, for instance, the distinction between two of the three major Mahyna philosophical schools, Svtantrika-Mdhyamika and Prsagika-Mdhyamika reflect, inter alia, distinct readings of this text, itself taken as fundamental by scholars within each of these schools.1

The treatise itself is composed in very terse, often cryptic verses, with much of the explicit argument suppressed, generating significant interpretive challenges. But the uniformity of the philosophical methodology and the clarity of the central philosophical vision expressed in the text together provide a considerable fulcrum for exegesis. Moreover, the rich commentarial literature generates a number of distinct and illuminating readings. The central topic of the text is emptiness—the Buddhist technical term for the lack of independent existence, inherent existence, or essence in things. Ngrjuna relentlessly analyzes phenomena or processes that appear to exist independently and argues that they cannot so exist, and yet, though lacking the inherent existence imputed to them either by naive common sense or by sophisticated realistic philosophical theory,2 these phenomena are not nonexistent—they are, he argues, conventionally real.

This dual thesis of the conventional reality of phenomena together with their lack of inherent existence depends upon the complex doctrine of the two truths or two realities—a conventional or nominal truth and an ultimate truth—and upon a subtle and surprising doctrine regarding their relation. It is, in fact, this sophisticated development of the doctrine of the two truths as a vehicle for understanding Buddhist metaphysics and epistemology that is Ngrjuna’s greatest philosophical contribution. If the analysis in terms of emptiness is the substantive heart of Mlamadhyamakakrik, the method of reductio ad absurdum is the methodological core. Ngrjuna, like Western sceptics, systematically eschews the defense of positive metaphysical doctrines regarding the nature of things, arguing rather that any such positive thesis is incoherent and that, in the end, our conventions and our conceptual framework can never be justified by demonstrating their correspondence to an independent reality. Rather, he suggests, what counts as real depends precisely on our conventions.3

For Ngrjuna and his followers this point is connected deeply and directly with the emptiness of phenomena. That is, for instance, when a Mdhyamika philosopher says of a table that it is empty, that assertion by itself is incomplete. It invites the question, Empty of what? And the answer is, Empty of inherent existence, or self-nature, or, in more Western terms, essence.4 Now, to say that the table is empty is hence simply to say that it lacks essence and importantly not to say that it is completely nonexistent.5 To say that it lacks essence, the Mdhyamika philosopher will explain, is to say, as the Tibetans like to put it, that it does not exist “from its own side”—that its existence as the object that it is—as a table— depends not on it, nor on any purely nonrelational characteristics, but depends on us as well. That is, if our culture had not evolved this manner of furniture, what appears to us to be an obviously unitary object might instead be correctly described as five objects: four quite useful sticks absurdly surmounted by a pointless slab of stick-wood waiting to be carved. Or we would have no reason to indicate this particular temporary arrangement of this matter as an object at all, as opposed to a brief intersection of the histories of some trees. It is also to say that the table depends for its existence on its parts, on its causes, on its material, and so forth. Apart from these, there is no table. The table, we might say, is a purely arbitrary slice of space-time chosen by us as the referent of a single name and not an entity demanding, on its own, recognition and a philosophical analysis to reveal its essence. That independent character is precisely what it lacks on this view.6

So from the standpoint of Mdhyamika philosophy, when we ask of a phenomenon, Does it exist?, we must always pay careful attention to the sense of the wordexist” that is at work. We might mean exist inherently, that is, in virtue of being a substance independent of its attributes, in virtue of having an essence, and so forth, or we might mean exist conventionally, that is to exist dependently, to be the conventional referent of a term, but not to have any independent existence. No phenomenon, Ngrjuna will argue, exists in the first sense. But that does not entail that all phenomena are nonexistent tout court. Rather, to the degree that anything exists, it exists in the latter sense, that is, nominally, or conventionally. It will be important to keep this ambiguity in “exists” in mind throughout the text, particularly in order to see the subtle interplay between the two truths and the way in which the doctrine of the emptiness of emptiness resolves apparent paradoxes in the account.

And this analysis in terms of emptiness—an analysis refusing to characterize the nature of anything precisely because it denies that we can make sense of the idea of a thing’s nature—proceeding by the relentless refutation of any attempt to provide such a positive analysis, is applied by Ngrjuna to all phenomena, including, most radically, emptiness itself. For if Ngrjuna merely argued that all phenomena are empty, one might justly indict him for merely replacing one analysis of things with another, that is, with arguing that emptiness is the essence of all things. But Ngrjuna, as we shall see, argues that emptiness itself is empty. It is not a selfexistent void standing behind a veil of illusion comprising conventional reality, but merely a characteristic of conventional reality. And this, as we shall see, is what provides the key to understanding the deep unity between the two truths.7

While Ngrjuna is a powerfully original thinker, he is clearly and self-consciously operating squarely within the framework of Buddhist philosophy. As such, Ngrjuna accepts and takes it as incumbent upon him to provide an account of the Four Noble Truths, nirva, buddhahood, and other fundamental Buddhist soteriological conceptions. Moreover, he takes it as a fundamental philosophical task to provide an understanding of what Buddhist philosophy refers to as pratīyasamutpda—dependent coorigination. This term denotes the nexus between phenomena in virtue of which events depend on other events, composites depend on their parts, and so forth. Exactly how this dependency is spelled out, and exactly what its status is, is a matter of considerable debate within Buddhist philosophy, just as the nature of causation and explanation is a matter of great dispute within Western philosophy. Ngrjuna is very much concerned to stake out a radical and revealing position in this debate. We will, in fact, see that this position and its connection to his understanding of emptiness and the nirvna-sasra relation provides the key to understanding his entire text.

Mlamadhyamakakrik is divided into twenty-seven chapters, which fall roughly, though by no means officially, into four sections. In the first section of the text, comprising Chapters I through VII, Ngrjuna discusses the fundamental theoretical constructs in Buddhist ontology, such as dependent origination, change and impermanence, perception, the aggregates that compose the self, the elements that constitute the universe, and the relation between substance and attribute. In the second major section, Chapters VIII through XIII, Ngrjuna focuses on the nature of the self and of subjective experience. Chapters XIV through XXI are primarily concerned with the external world and the relation of the self to objects. The final section, Chapters XXII through XXVII, addresses phenomena associated with the ultimate truth, such as buddhahood, emptiness, and nirva, and the relation of the conventional to the ultimate and of sasra to nirvna. The chapters that form the climax of the text are found in this section. But it is important to note that in fact the dialectical structure of the text requires a reading of these chapters in order to fully grasp the impo��grart of the earlier ones. This is because the doctrine of the emptiness of emptiness does not fully emerge until this point, and it is crucial to Ngrjuna’s argument that all phenomena are empty and that their emptiness is also empty.8

The order of the chapters is often, though not always, important. Often a chapter will consider a phenomenon held by a proponent of another philosophical school to be inherently existent. Or an opponent may charge Ngrjuna with denying the actuality of a phenomenon in virtue of asserting its emptiness. In his analysis, Ngrjuna will typically argue that the phenomenon proposed as inherently existent cannot be so and indeed is empty, or that the phenomenon whose existence he is charged with denying is, in fact, on his analysis, while nonexistent from the ultimate point of view, conventionally existent. In each case, he will argue that the functions the opponent thought could only be served by an inherently existent phenomenon can, in fact, be served only by empty phenomena. But quite often these analyses will inspire natural rejoinders of the form, “Yes, x might well be empty and only conventionally existent, but we can’t make sense of its conventional existence without presupposing the inherent existence of y. “In such cases, the next chapter will typically address that natural rejoinder. So, for instance, the first chapter argues that conditions and the relation between phenomena and that on which they depend are empty. But a natural rejoinder is that even conventional but actual conditions can only be understood in the context of change or impermanence. So Chapter II addresses change. The text hence forms a single sustained argument with only a few digressions or changes of subject, generally marked by the section divisions I have suggested above.

The first chapter addresses dependent origination. While many Western commentators assert that this chapter opens the text simply because it addresses a “fundamental doctrine of Buddhism,”9my analysis of the text suggests that Ngrjuna begins with causation for deeper, more systematic reasons. In Chapters II through XXI, Ngrjuna addresses a wide range of phenomena, including external perceptibles, psychological processes, relations, putative substances, and attributes, arguing that all are empty. In the final six chapters, Ngrjuna generalizes the particular analyses into a broad theory concerning the nature of emptiness itself and the nature of the ultimate, of liberation, and of the relation between emptiness and dependent arising. At the close, he replies to objections. It is generally, and in my view correctly, acknowledged that Chapter XXIV, the examination of the Four Noble Truths, is the central chapter of the text and the climax of the argument, with Chapter XXV on nirva and sasdex="00006" align="baseline" height="13" width="7">ra sharing that spotlight. One verse of Chapter XXIV, verse 18, has received so much attention that interpretations of it alone represent the foundations of major Buddhist schools in East Asia:

18. Whatever is dependently co-arisen
That is explained to be emptiness.
That, being a dependent designation
Is itself the middle way.


Here Ngrjuna asserts the fundamental identity of (1) emptiness, or the ultimate truth; (2) the dependently originated, that is, all phenomena; and (3) verbal convention. Moreover, he asserts that understanding this relation is itself the middle-way philosophical view he articulates in Mlamadhyamakakrik. This verse and the discussion in the chapters that follow provide the fulcrum for Candrakīrti’s more explicit characterization of the emptiness of emptiness as an interpretation of Ngrjuna’s philosophical system—the interpretation that is definitive of the Prsangika-Mdhyamika school.10 In what follows I will provide an interpretation of the text inspired by the centrality of this verse and of the chapters forming its context that harmonizes with Candrakīrti’s. In fact, on my reading of the text this doctrine is already found in the opening chapter—the examination of conditions. Reading the text in this way locates the doctrine of the emptiness of emptiness not only as a dramatic philosophical conclusion to be drawn at the end of twenty-four chapters of argument, but as the perspective implicit in the argument from the very beginning and only rendered explicit in XXIV. Reading the text in this way will show us exactly how XXIV: 18 is to be understood and just why a proper understanding of causality is so central to Buddhist philosophy.

When a Westerner first encounters Mlamadhyamakakrik or other Mdhyamika texts, the philosophical approach can appear highly metaphysical and downright weird. The unfamiliar philosophical vocabulary, the highly negative dialectic, and the cryptic verse form are indeed forbidding. Most bizarre of all, however, at first glance is the doctrine that all phenomena, including self and its objects, are empty. For indeed Ngrjuna and his followers do argue that the entire everyday world is, from the ultimate standpoint, nonexistent. And that does appear to stand just a bit deeper into philosophical left field than even Berkeley dares to play. But if the interpretation I will urge is adopted, the real central thrust of Mdhyamika is the demystification of this apparently mystical conclusion. While it might appear that the Mdhyamikas argue that nothing really exists except a formless void, in fact the actuality of the entire phenomenal world, persons and all, is recovered within that emptiness.11

Now a word about the methodology and intent of this commentary: Since the intended audience is Western philosophers and students of philosophy whose primary study has been in the Western tradition, I have tried throughout, insofar as that is possible without distortion of the meaning of the text, to explain Ngrjuna’s arguments and positions in language familiar to Western philosophers. I have occasionally used analogies to positions and arguments found in Western texts, but have avoided doing so where I thought that the comparisons might force a Procrustean analysis of Ngrjuna’s own views. And it is, of course, impossible and pointless to completely recast Ngrjuna’s positions as those with which we in the West are familiar and to replace his technical terminology with ours. For Ngrjuna is not a Western philosopher. He is an Indian Buddhist philosopher whose work we approach through a vast Asian Buddhist commentarial literature. And while many of his concerns, problems, theses, and arguments are recognizable cousins of ours, many are not, and there are genuine differences in outlook.

This is what makes Ngrjuna’s work so exciting to read and to think about—it provides a genuinely distinctive perspective on a set of problems and projects that we share. In commenting on Ngrjuna’s text, I am constantly aware of walking a philosophical and hermeneutical tightrope. On the one hand, one could provide a perfectly traditional commentary on the text—or better, a translation of one of the major Sanskrit or Tibetan commentaries—or a transcript of oral commentary by a recognized scholar of the tradition. Such a commentary would explain in great detail the way the text is seen from the perspective of its home tradition and the background of Buddhist controversies to which the text responds. A commentary like this would undoubtedly be of great use to Buddhologists and philosophers already steeped in Buddhist philosophy and its history. And indeed Sprung’s translation of most of Candrakīrti’s Prasannapad (Lucid Exposition), including the root verses from Mlamadhyamakakrik, partially fulfills this need. But many of these scholars and students already have access to the relevant texts in their original languages or to teachers situated within the Buddhist tradition.

On the other hand, one could try to comment on the text by prese�� tenting a theory of what Ngrjuna would have said had he been a twentieth-century Western philosopher. One could then feel free to step back from the internecine debates in the classical Buddhist academy, which were so absorbing to the historical Ngrjuna and so distant from our own context, and simply ask how his arguments would be formulated in the context of the contemporary philosophical scene. Leaving aside the question of how one would identify the possible philosopher denoted by this bizarre counterfactual, this would again be a profoundly unsatisfying enterprise. For what makes this a great text is not simply that we can extrapolate its significance to our own context, but that in reading it, to borrow Gadamer’s metaphor, we are able to fuse its textual horizon with our own. It is the bringing to the present of Ngrjuna’s own concerns, insights, and arguments that is revelatory, not speculation about a related counterfactual nonentity. And for this fusion of interpretive horizons to be possible, we must, as much as possible, respect the original horizon of the text.

Having said this, one must confess the double difficulty of giving sense to the phrase “Ngrjuna’s own concerns, insights, and arguments.” The recovery of authorial intent as a hermeneutic task is problematic (especially when the author is so culturotemporally remote and when his corpus is as controversial in composition and interpretation as is Ngrjuna’s). But it is equally problematic as a hermeneutic desideratum. For who is to say that Ngrjuna was/is the best possible interpreter of Mlamadhyamakakrik? After all, he did not have the benefit of the long commentarial tradition he spawned.12 A great text—or, as Gadamer has referred to such texts, an “eminent text”—grows over time and merits reinterpretation and rereading as the tradition in which it participates develops and provides an ever-expanding context for its reading. Moreover, I am reading Ngrjuna largely through the lens of the Tibetan commentarial tradition and through the Tibetan translation of his text—the text read and discussed by the scholars of this long, deep, and intellectually diverse and rich tradition, few of whom had access to Sanskrit. So the Ngrjuna whose views I am exploring is an evolving figure, rooted in the life and writing of a first or second century Indian monk, of whom we know but little, but whose literary life and identity extends through a complex, sophisticated, and contested textual and philosophical tradition in India and Tibet and in the West.

As a consequence, in interpreting this text on the Midd��xt le Path for a Western audience, I have sought insofar as possible to find a middle path between these extremes. I have tried to explain Ngrjuna’s own arguments and their context as straightforwardly as possible without burdening the Western philosophical reader with extended discussion of the specifically ancient Indian Buddhist philosophical debates. I have indicated ways in which very specific arguments can be generalized and have commented on general structural features of arguments, chapters, and the text. I have throughout explained arguments in Western philosophical terms, while situating those arguments in their Buddhist context. There may be times when my desire to make arguments accessible has led to some distortion in Ngrjuna’s sense. There may also be times at which, by leaving arguments set firmly within the soteriological context of Buddhism, I have left those arguments looking like curios to my Western audience. Some of this may be unavoidable, but in any case I have sought specifically to minimize these difficulties.

The interpretation I offer is situated squarely within a Prsangika-Mdhyamika interpretation of Ngrjuna (the philosophical school that reads Mlamadhyamakakrik through the commentaries of Buddhaplita and Candrakīrti). But more specifically, my reading is heavily influenced by the Tibetan Geluk-pa tradition that takes as central the commentaries of dGe-’dun-grub, mKhas-grub-rje, and especially, Je Tsong Khapa. My interpretation of the text reflects not only Candrakīrti’s and Je Tsong Khapa’s commentaries, but also the extended oral commentary I have received on this text from the eminent Tibetan Mdhyamika scholars, especially the Ven. Professor Geshe Yeshes Thap-Khas of the Central Institute of Higher Tibetan Studies and the Ven. Professor Gen Lobzang Gyatso of the Institute of Buddhist Dialectics (I should point out that both of these scholars—as well as others to whom I am indebted for valuable conversations, including the Most Ven. Prof. Samdhong Rinpoche and the Ven. Geshe Namgyal Wangchen—received their education at Drepung Loseling Monastic College, and so my interpretation also reflects more particularly the academic tradition of that institution).

Having characterized this as a tradition of interpretation, I must emphasize that it is not, as it is often represented, and as it often represents itself, a homogeneous tradition. Though there is a hermeneutic convention in Indo-Tibetan Buddhist literature of presenting oneself as merely expounding faithfully the views of all of the earlier commentators, this is almost never the truth. There are considerable divergences in interpretation and in philosophical position within Buddhist schools and within lineages. Indeed the Tibetan scholars I have regularly consulted, despite the fact that they shared many of the same teachers and an identical curriculum, differ widely among themselves on many issues. It would hence be impossible in any case to represent accurately the Prsangika-Mdhyamika interpretation, or even the Geluk-pa interpretation or the Drepung Loseling interpretation of Mlamadhyamakakrik.

I emphasize that even if one could identify such a homogeneous interpretation, I am not here presenting the interpretation or interpretations of any of these commentators or scholars, individually or collectively. There are substantial debates within these traditions regarding interpretative issues, and I do not consistently side with any particular faction (though I do think that it is true that my reading never conflicts directly with that of Candrakīrti); sometimes (as in my reading of the final chapter) I depart from the most common Geluk-pa interpretation entirely in favor of a line more closely associated with the Nyingma-pa reading of the text. Nor is the purpose of this text to compare, criticize, and resolve differences between interpretations. Instead, I here present the text as I read it, having been influenced by all of these commentators and teachers, and as I present it to my Western colleagues. And my intention in doing so is to let the text stand alone as a work of philosophy valuable in its own right to anyone interested in fundamental metaphysical, epistemological, and soteriological questions, not as a text to be studied only as part of “the history of philosophy” or “comparative philosophy.”

Moreover, my exposition will be deliberately sympathetic. My goal is not to assess Ngrjuna’s philosophy, but to present and elucidate it and to do so in a way that, while making the text accessible to Western philosophers, does not disguise the fact that the text made accessible is an early Indian Mdhyamika philosophical treatise, read by a Western philosopher through an extended Indo-Tibetan commentarial and academic tradition. It is neither a contemporary treatise nor a second century text transported miraculously to us without the distortion of time and cultural distance. Buddhologists may lament the lack of critical discussion of Buddhist antecedents and commentarial sequellae, and my Tibetan colleagues may be uncomfortable with some of the tendentious extensions of arguments beyond the dialectical contexts in which they originally arose. Despite this, I hope that for Western philosophers interested in approaching Mdhyamika in particular or Buddhist philosophy in general, and for students of Ngrjuna’s philosophy in the West, this exposition will make his text more accessible.

 
Dedicatory Verses


I prostrate to the Perfect Buddha,
The best of teachers, who taught that
Whatever is dependently arisen is
Unceasing, unborn,
Unannihilated, not permanent,
Not coming, not going,
Without distinction, without identity,
And free from conceptual construction.


Dedicatory verses are often treated as mere performatives. But these are special and announce in a subtle but powerful way the program of the Mlamadhyamakakrik. There is a common point being made in the four pairwise denials, but also a specific insight being expressed in each. The relation between the conventional and the ultimate that will be developed in the text is also expressed poetically in the dedication. In fact, Candrakïrti, in Prasannapad, argues that the dedication determines the Prsangika reading of Ngrjuna’s text.

Candrakïrti’s point is this: In the four pairwise denials, Ngrjuna is announcing that the Mdhyamika philosopher will make no positive assertions about the fundamental nature of things. But this claim must be qualified in several ways. For one thing, we must take the phrase “the nature of things” very seriously. That is, Ngrjuna will be refusing to say anything about the essence of anything exactly because he will deny the coherence and utility of the concept of an essence. For another, it is important to see that the predications that are rejected are intended to be understood as made from the ultimate standpoint. That is, the assertions that are being denied are assertions about the final nature of phenomena that emerge from philosophical analysis. They are not meant to be ordinary assertions dependent upon conventions. Ngrjuna will deny that it is possible to assert anything from the ultimate standpoint. He will urge that all truth is relative and conventional. In fact, as we shall see, these qualifications turn out to be mutually entailing.

But each pair is significant in its own right. To say that “whatever is dependently arisen is unceasing and unborn” is to emphasize that dependent arising amounts to emptiness, and emptiness amounts to nonexistence in the ultimate sense. While, as we shall see, Ngrjuna defends the conventional existence of phenomena, he will urge that none of them ultimately exist—that none of them exist independently of convention with identities and natures that they possess in themselves. Therefore, he will argue, nothing ultimately is born, and from the ultimate standpoint there is nothing to cease. This is a deep point, which only emerges completely through a reading of the whole text. But we can say at this point that this insight contains within it the seeds of the eventual equation of the phenomenal world with emptiness, of sasrawith nirva, and of the conventional and the ultimate that are the hallmarks of the Prsaìgika-Mdhyamika view.

When Ngrjuna claims that “whatever is dependently arisen is … unanihilated and not permanent” he indicates that the dependently arisen world and all of its contents are, in virtue of being dependently arisen and dependent upon conditions, impermanent. Phenomena come into existence when the conditions upon which they depend obtain, and they cease to exist when the conditions for their continued existence no longer obtain. This impermanence, he will argue, entails their nonexistence from the ultimate standpoint. For there will be no principled way to assert criteria for identity for phenomena that distinguish them in any principled way from their conditions. Nor can we find any essence they themselves have that determines their identity. The criteria for identity we posit will end up being purely conventional. Hence the same is true for any claims of substantial difference between things. But this impermanence and lack of intrinsic identity, while it amounts to the impossibility of ultimate existence, is not equivalent to annihilation. The empirical reality of things, on Ngrjuna’s analysis, is not denied by asserting their emptiness.

Finally, to assert that things are “not coming, not going” is to assert that the phenomenal world does not contain intrinsically identifiable entities that persist independently with those identities over time. As a consequence, there can be no sense in saying that any entity, independent of conventional imputation, comes into existence, remains in existence, or goes out of existence.13

The final remark—that the phenomenal world is free from conceptual imputation—raises a tension that is central to Mdhyamika philosophy and that animates the whole of the text: The tension between the desire to characterize the ultimate nature of things and the recognition that all characterization is conventional. For Ngrjuna will urge that all conventional phenomena are conceptually designated, depending for whatever identity and existence they have on such designation, and that this merely imputed status is their ultimate nature. Despite this, however, he will urge that seeing this fact is at the same time to see that the nature naively imputed to things and the nature they appear to us to have—inherent existence—is wholly false. In themselves, from their side, things are free of that imputation, even though there is really nothing at all that can be said from their side. This dynamic philosophical tension—a tension between the Mdhyamika account of the limits of what can be coherently said and its analytical ostension of what can’t be said without paradox but must be understood—must constantly be borne in mind in reading the text. It is not an incoherent mysticism, but it is a logical tightrope act at the very limits of language and metaphysics.


Chapter I

Examination of Conditions


Central to this first chapter is the distinction between causes and conditions (Skt: hetu and pratyaya, Tib: rgyu and rkyen). This distinction is variously drawn and is controversial,14 and it is arguably differently understood in Sanskrit and Tibetan. The way I will understand it here, I argue, makes good, coherent sense not only of this chapter, ��but of Mlamadhyamakakrik as a whole. Briefly, we will understand this distinction as follows: When Ngrjuna uses the wordcause” (hetu, rgyu), he has in mind an event or state that has in it a power (kriy, bya-ba)15 to bring about its effect and has that power as part of its essence or nature (svabhva, rang bzhin). When he uses the term “condition” on the other hand (pratyaya, rkyen), he has in mind an event, state, or process that can be appealed to in explaining another event, state, or process without any metaphysical commitment to any occult connection between explanandum and explanans. In Chapter I, Ngrjuna, we shall see, argues against the existence of causes and for the existence of a variety of kinds of conditions.16

Things are not, however, quite this simple. For in the philosophical context in which Ngrjuna is writing, there are those—indeed including most Buddhist philosophical schools—who would accept his classification of conditions, but who would then assert that in order for conditions to function as explanatory, they must themselves have an independent inherent existence. Some—such as the Sarvastivadas or Sautntrika-Svtantrikas (despite other differences between these schools regarding causation)—would argue that the conditions must exist as substantially distinct from the conditioned; others, such as the Cittamtra, would argue that they can be of the same nature.17 Ngrjuna will evade these particular debates, however, by emphasizing that the conditions he has in mind must be thought of as empty of inherent existence and connected to the phenomena they condition neither through absolute difference nor through identity.

The argument against causation is tightly intertwined with the positive account of dependent arising and of the nature of the relation between conditions and the conditioned. Ngrjuna begins by stating the conclusion (I: 1): Entities are neither self-caused nor do they come to be through the power of other entities. That is, there is no causation when causation is thought of as involving causal activity:18

1. Neither from itself nor from another,
Nor from both,
N���or without a cause,
Does anything whatever, anywhere arise.


The fourfold classification of positions with regard to the relation between an active cause and its effect is meant to be exhaustive. But it is important to keep in mind that Ngrjuna was aware of philosophical schools espousing each of these four positions. And each of them has something to say for itself if we begin by supposing a model of causation involving powers as essential properties of substantially real causes. The first view—held prominently by Samkhya philosophers19 —is that all causation is really self-causation. A proponent of this view would argue that for a cause to be genuinely the cause of an effect, that effect must exist potentially in that cause. If it does not, then the cause might exist without the effect, in which case the cause would fail to necessitate the effect, in which case it would not be a genuine cause. This is not to say that effects exist in full actuality in their causes, but that they have a genuine potential existence when their causes exist. In this case, since the effect is present in the cause, it already has a kind of existence prior to its appearance. And it is the fact of this prior potential existence that accounts for the causal character of the cause. So we can say, on this view, that a thing’s prior potential existence is what gives rise to its later actual existence. So effects are in this sense self-caused. The typical kind of example appealed to in order to defend this model of causation is the seed and sprout relation. The sprout, although only actual after germination, is potential in the seed. Its potentiality is what makes the seed a seed of that sprout. Moreover, on this view, the seed and sprout cannot be distinguished as substantially different. Intuitively it makes sense to say that they are two stages of the same entity. But the seed is the cause of the sprout. Hence, the proponent of this view concludes, the sprout is self-caused.

Causation from another is a more familiar way of thinking of causation and was the dominant doctrine of causation in the Buddhist philosophical milieu in which Ngrjuna was working. On this view, causes and their effects are genuinely distinct phenomena.20 They can be characterized and can in principle exist independently of one another. But they are related by the fact that one has the power to bring the other about. The relations between parents and children is an example often appealed to in illustrating this doctrine. Parents bring their children into existence. But they are not identical entities.

The doctrine of causation by both self and other emerges through a juxtaposition of the doctrine of causation-from-another and the doctrine of self-causation. Let us return to the example of the seed. A proponent of other-causation might point out that seeds that are not planted, watered, and so forth, do not sprout. If the sprout were present in the seed, these other conditions, which are manifestly other than the sprout, would be otiose. On the other hand, the proponent of self-causation might reply: No matter how much you water, nourish, and exhort an infertile seed—one without the potentially existent sprout—nothing happens. So all of the distinct conditions in the world will not suffice absent the potential existence of the effect. The happy compromise doctrine that emerges is the��� doctrine of causation-by-both: Effects are the result of the joint operation of the effect itself in potentio and the external conditions necessary to raise the effect’s mode of existence from potentiality to actuality.

The fourth alternative view of causation is that things simply spontaneously arise from no particular causes—that there are no links at all between events. What might motivate such a view? Well, as we shall see (and as any reader of Sextus Empiricus, Hume, or Wittgenstein will recall), there are powerful reasons for believing that none of the three alternatives just rehearsed can be made coherent. And if one believed that only if there were either some identity or difference between causes and effects could there be a relation of dependency between phenomena, one would be forced to the nihilistic conclusion that things simply arise causelessly.

Nonetheless, Ngrjuna notes, there are conditions—in fact four distinct kinds—that can be appealed to in the explanation and prediction of phenomena:

2. There are four conditions: efficient condition;
Percept-object condition; immediate condition;
Dominant condition, just so.
There is no fifth condition.


The general classification of conditions Ngrjuna employs is pretty standard in Indian and especially in Buddhist accounts of explanation. But there are two specific features of Ngrjuna’s presentation that should be noted: First, since he is writing with specifically soteriological goals in mind, which require the practicioner to develop a deep insight into the nature of his/her own mind, there is a specifically psychological emphasis in the presentation. We must be aware both of this emphasis and of the natural generalization away from that particular domain that the account supports. Second, it will be of paramount importance to Ngrjuna that the analysis of the relation of conditions to the conditioned involves ascribing neither inherent existence nor causal power to the conditions.

Efficient conditions are those salient events that explain the occurrence of subsequent events: Striking a match is the efficient condition for its lighting. My fingers depressing the keys of this computer is the efficient condition for the creation of this text.

The percept-object condition is in its primary sense the object in the environment that is the condition for a mind’s perception of it. So when you see a tree, the physical tree in the environment is the percept-object condition of your perceptual state. Now things get vexed here in a number of ways. First, there is no unanimity in the world, or even in Buddhist philosophy, regarding the analysis of perception and, hence, no consensus on the view just adumbrated—that external objects are the percept-object conditions of perceptual awareness. Idealists, for instance, argue that the percept-object conditions are to be located in the subject. Second, many fans of percept-object conditions, on both sides of��� the idealist/realist divide, argue that the substantial existence of such a condition, and the appropriate exercise of its power to produce perception, is a necessary condition of perception. Ngrjuna will be concerned to reject any such analysis—whether idealist or realist—in virtue of his attack on the notions of substantial existence, substantial difference, and causal power. Third, within the psychological domain, the account generalizes beyond perception. Conceptual states, imaginings, reasoning—all can have percept-object conditions. To Western philosophical ears this seems odd. But from the standpoint of Buddhist epistemology and psychology, intentional21 activity generally is the natural kind comprised by “perception.” So the point is that the intentional existence of the golden mountain is a percept-object condition of my being able to doubt that there is such a thing. Finally, the analysis bears generalization well beyond the psychological. For at the most abstract level, what is distinct about a percept-object condition is its existence simultaneously with and as a support for what it conditions. So Ngrjuna’s attack on a substantialist understanding of this kind of explanans will apply, mutatis mutandis, to the case of a table supporting a book.

The dominant condition is the purpose or end for which an action is undertaken. My hoped for understanding of Mdhyamika might be the dominant condition for my reading Ngrjuna’s text, its presence before my eyes the percept-object condition, and the reflected light striking my eyes the efficient condition. The immediate conditions are the countless intermediary phenomena that emerge upon the analysis of a causal chain, in this case, the photons striking my retina, the excitation of photoreceptor cells, and so forth.22

A nonpsychological example might be useful to illustrate the difference between the four kinds of condition and the picture Ngrjuna suggests of explanation in the most general sense: Suppose that you ask, “Why are the lights on?” I might reply as follows: (1) “Because I flicked the switch.” I have appealed to an efficient condition. Or, (2) “Because the wires are in good working order, the bulbs haven’t burned out, and the electricity is flowing.” These are supporting conditions. Or, (3) “The light is the emission of photons each of which is emitted in response to the bombardment of an atom by an electron, and so forth.” I have appealed to a chain of immediate conditions. Or, (4) “So that we can see.” This is the dominant condition. Any of these would be a perfectly good answer to the “Why?” question. But note that none of them makes reference to any causal powers or necessitation.23

The next three verses are crucial to Ngrjuna’s understanding of the nature of conditions and their role in explanation. Ngrjuna first notes (I: 3) that in examining a phenomenon and its relations to its conditions, we don’t find that phenomenon somehow contained potentially in those conditions:

3. The essence of entities
Is not present in the conditions, etc….
If there is no essence,
There can be no otherness-essence.


The point being made in the first two lines of the verse is fairly straightforward. When we examine the set of conditions that give rise to an entity—for example, the set of conditions we detailed above for the shining of a lamp, or the conditions for seeing a tree we discussed previously—no analysis of those conditions yields the consequent effect. Dissecting light switches, wires, brains, and so forth, does not reveal any hidden light. Nor is there a tree perception to be found already in the existence of the tree, the eye, and so forth. Rather these phenomena arise as consequences of the collocation of those conditions. To borrow a Kantian turn of phrase, phenomena are not analytically contained in their conditions; rather, a synthesis is required out of which a phenomenon not antecedently existent comes to be.

But Ngrjuna, through his use of the phrase “the essence of entities” (dngos-po rnams kyi rang bzhin), emphasizes a very important metaphysical consequence of this observation: Given that phenomena depend upon their conditions for their existence and given that nothing answering to an essence of phenomena can be located in those conditions and given that there is nowhere else that an essence could come from, it follows that phenomena that arise from conditions are essenceless. One might argue at this point that just as phenomena come into existence dependent upon conditions, their essences come into existence in this way. But what goes for phenomena24 does not go for essences. For essences are by definition eternal and fixed. They are independent. And for a phenomenon to have an essence is for it to have some permanent independent core. So neither essences nor phenomena with essences can emerge from conditions.

The next two lines require a careful gloss, both because of the complexity of the philosophical point at stake and because of the Buddhist philosophical term of art I translate as “otherness-essence” (Skt: parabhva, Tib: gzhan dngos). Let us begin by glossing that term. In its primary sense it means to have, as a thing’s nature, dependence upon another for existence. So for a table, for instance, to have otherness-essence, according to a proponent of this analysis of the nature of things, might be for it to have as an essential characteristic the property of depending for its existence on some pieces of wood, a carpenter, and so forth. This way of thinking of the nature of things has great appeal—was used by those who defended the analysis of causation as production from other and the analysis of causes and their effects according to which they are linked by causal powers inhering in the causes—particularly for other Buddhist schools who would want to join with Ngrjuna in denying essence to phenomena. For such a philosopher, it would be c���ongenial to argue that the table has no essence of its own, but has the essential property of depending on its parts, causes, and so forth—an essential property that depends critically on another. And it would then be important to note that this nature relies on the other having an intrinsic connection to the phenomenon in question, a connection realized in the causal powers (or other inherently existent relation to the effect) of that other and, hence, in the other’s own nature. Moreover, it is crucial to such an analysis, if it is not to lapse into the absurdities that plague self-causation, that there be a real, substantial difference in entity—a difference in intrinsic nature between the dependent phenomenon and the conditions on which it depends. Absent such a difference, the otherness required in the analysis cannot be established.25

Given this understanding of otherness-essence, we can see the arguments Ngrjuna is ostending in the last two lines of this verse. First, since all entities are without their own essences (that is, without essences that can be specified intrinsically without reference to anything else), the other with respect to which any phenomenon is purportedly essentially characterized will be without an essence, and so there will be no basis on which to build this otherness-essence. Second, without individual essences, there will be no basis on which to draw the absolute, essential distinctions necessary to establish phenomena as intrinsically other than their conditions. Without individual essences there are not substantial differences. Without substantial differences, there are no absolute others by means of which to characterize phenomena. Third, in order to characterize phenomena as essentially different from their conditions, it is important to be able to characterize them independently. Otherwise, each depends for its identity on the other, and they are not truly distinct in nature. But the whole point of otherness-essence is that things in virtue of having it are essentially dependent. So the view is in fact internally contradictory. Given that things have no intrinsic nature, they are not essentially different. Given that lack of difference, they are interdependent. But given that interdependence, there cannot be the otherness needed to build otherness-essence out of dependence.

Now, on the reading of this chapter that I am suggesting, we can see conditions simply as useful explanans. Using this language, Ngrjuna is urging that even distinguishing between explanans and explanandum as distinct entities, with the former containing potentially what the latter has actually, is problematic. What we are typically confronted with in nature is a vast network of interdependent and continuous processes, and carving out particular phenomena for explanation or for use in explanations depends more on our explanatory interests and language than on joints nature presents to us. Through addressing the question of the potential existence of an event in its conditions, Ngrjuna hints at this concealed relation between praxis and reality.

Next, Ngrjuna notes (I: 4) that in invoking an event or entity as a condition in explanation we do not thereby ascribe it any causal power:

4. Power to a���ct does not have conditions.
There is no power to act without conditions.
There are no conditions without power to act.
Nor do any have the power to act.


This is the beginning of Ngrjuna’s attack on the causal power/cement-of-the-universe view of causation and his contrastive development of his regularity view of conditioned dependent arising. Causal powers, according to those who posit them, are meant to explain the causal nexus—they are meant to explain how it is that causes bring about their effects, which is itself supposed to be otherwise inexplicable. But, Ngrjuna argues, if there were a causal power, it itself, as a phenomenon, would either have to have conditions or not. If the former, there is a vicious explanatory regress, for then one has to explain how the powers to act are themselves brought about by the conditions, and this is the very link presupposed by the friend of powers to be inexplicable. One could posit powers the conditions have to bring about powers and powers the powers have to bring about effects. But this just moves one step further down the regress.

If, on the other hand, one suggests that the powers have no condition, one is stuck positing uncaused and inexplicable occult entities as the explanans of causation. If what is to be explained is how it is that all phenomena are brought about by causal processes, it is a bit embarrassing to do so by reference to unobserved entities that are explicitly exempted from this otherwise universal condition. Moreover, there is then no explanation of how these powers arise and why they come to be where they are. This is all startlingly anticipatory of Wittgenstein’s famous echo of Hume in the Tractatus:

6.371 The whole modern conception of the world is founded on the illusion that the so-called laws of nature are the explanations of natural phenomena.

6.372 Thus people today stop at the laws of nature, treating them as something inviolable, just as God and Fate were treated in past ages. And in fact both are right and both are wrong: though the view of the ancients is clearer in so far as they have a clear and acknowledged terminus, while the modern system tries to make it look as if everything were explained.26


In the next two lines, as we will often see in the text, Ngrjuna is speaking in two senses—first, from the conventional standpoint, and second, from the ultimate. In the third line of the verse, he notes that conditions can certainly, in a perfectly legitimate sense, be appealed to as the things that bring about their effects; in that sense, we can say that they are efficacious—that they have the power to act. But in the fourth line he emphasizes that we cannot, so to speak, quantify over this power, identifying it as a phenomenon or property possessed by the conditions. There are no powers in that sense. Just as we can act for someone else’s sake, despite there being no sakes, we can appeal to the potency of conditions despite their being no such potency. The trick is to make correct use of conventional locutions��� without reifying denotata for all of the terms. For example, we might ask a farmer, “Do these seeds have the power to sprout?” as a way of asking whether they are fertile. It would be then perfectly appropriate for him to answer in the affirmative. But if we then asked him to show us where in the seed the power is located, he would be quite justified in regarding us as mad.27

Our desire for light does not exert some occult force on the lights. Nor is there anything to be found in the flicking of the switch other than the plastic, metal, movement, and connections visible to the naked eye. Occult causal powers are singularly absent. On the other hand, Ngrjuna points out in this discussion that this does not mean that conditions are explanatorily impotent. In a perfectly ordinary sense—not the sense that the metaphysicians of causation have in mind—our desire is active in the production of light. But not in the sense that it contains light potentially, or some special causal power that connects our minds to the bulbs.

What is it, then, about some sets of event pairs (but not others) that make them dependently related if not some causal link present in those cases but not in others?

5. These give rise to those,
So these are called28 conditions.
As long as those do not come from these.
Why are these not nonconditions?


One might answer this question, Ngrjuna notes in the opponent’s suggestion in the first two lines, by noting the presence of some relation of “giving rise to,” realized in a power. But, he rejoins in the final two lines, this move is blocked: For having shown the absence and the theoretical impotence of such a link, it would follow that there would be no conditions. Ngrjuna hence suggests here that it is the regularities that count. Flickings give rise to illuminations. So they are conditions of them. If they didn’t, they wouldn’t be. Period. Explanation relies on regularities. Regularities are explained by reference to further regularities. Adding active forces or potentials adds nothing of explanatory utility to the picture.29

In reading the next few verses, we must be hermeneutically cautious and pay careful attention to Ngrjuna’s use of the term “existent” (Tib: yod-pa, Skt: sat) and its negative contrastive “nonexistent” (Tib: med-pa, asat). For Ngrjuna is worried here about inherent existence and inherent nonexistence, as opposed to conventional existence or nonexistence. For a thing to exist inherently is for it to exist in virtue of possessing an essence—for it to exist independently of��� other entities and independently of convention. For a thing to be inherently nonexistent is for it to not exist in any sense at all—not even conventionally or dependently. With this in mind, we can see how Ngrjuna defends dependent arising while rejecting causation:

6. For neither an existent nor a nonexistent thing
Is a condition appropriate.
If a thing is nonexistent, how could it have a condition?
If a thing is already existent, what would a condition do?


He notes here that if entities are conceived as inherently existent, they exist independently and, hence, need no conditions for their production. Indeed, they could not be produced if they exist in this way. On the other hand, if things exist in no way whatsoever, it follows trivially that they have no conditions.30 The following three verses make this point with regard to each of the four kinds of conditions:

7. When neither existents nor
Nonexistents nor existent nonexistents are established,
How could one propose a “productive cause?”
If there were one, it would be pointless.


8. An existent entity (mental episode)31
Has no object.
Since a mental episode is without an object,
How could there be any percept-condition?


9. Since things are not arisen.
Cessation is not acceptable.
Therefore, an immediate condition is not reasonable.
If something has ceased, how could it be a condition?


In I: 7, Ngrjuna is reasoning that since an inherently existent phenomenon is by definition independent, it could not have been produced by anything else. An inherently nonexistent phenomenon certainly cannot be produced; if it were, it would be existent. An existent nonexistent (for instance, something posited by a Meinongian ontology—existing in a logical space, though not in the actual world) cannot be produced since its actual production would contradict its nonexistence and its production in some other way would contradict the inherent existence of the other sort posited for it.

The argument in I: 8 is a bit different and is directed more specifically at the special s���tatus of simultaneous supporting conditions, such as those posited in perception, as discussed above. Ngrjuna is making the following point: If we consider a particular moment of perception, the object of that perceptual episode no longer exists. This is so simply because of the mundane fact that the chain of events responsible for the arising of perceptual consciousness takes time. So the tree of which I am perceptually aware now is a tree that existed about one hundred milliseconds ago; not one that exists now. The light took some time to reach my eye; the nerve impulses from the eye to the brain took some time; visual processing took still more time. So if the story about how the tree is the percept-object condition of my perception according to which the tree exists simultaneously with the perception and exerts a causal power on my eye or visual consciousness were accepted, perception would be impossible. Moreover, the objects of many mental episodes are themselves nonexistent (like the golden mountain). But non-existents can’t be causally responsible for anything.

Verse 9 contains two arguments. In the first half of the verse, Ngrjuna is offering a quick reductio on the idea that immediately preceding conditions can exist inherently. By definition, an immediately preceding condition is a momentary element of a causal chain. And, by definition, something that is inherently existent is independent; hence, it cannot arise depending on something else and, therefore, cannot cease to exist. But immediately preceding conditions must arise and cease. In the final line of the verse, Ngrjuna develops a related problem. Immediately preceding conditions must cease before their effect arises. If their existence and exertion of causal power is what explains the arising of the cause, the arising of the cause is then inexplicable. (This argument is also used by Sextus Empiricus in Against the Logicians.)

What is important about this strand of the argument? Ngrjuna is drawing attention to the connection between a causal power view of causation and an essentialist view of phenomena on the one hand, and between a condition view of dependent arising and a conventional view of phenomena on the other. If one views phenomena as having and as emerging from casual powers, one views them as having essences and as being connected to the essences of other phenomena. This, Ngrjuna suggests, is ultimately incoherent since it forces one at the same time to assert the inherent existence of these things, in virtue of their essential identity, and to assert their dependence and productive character, in virtue of their causal history and power. But such dependence and relational character, he suggests, is incompatible with their inherent existence. If, on the other hand, one regards things as dependent merely on conditions, one regards them as without essence and without power. And to regard something as without essence and without power is to regard it as merely conventionally existent. And this is to regard it as existing dependently. This provides a coherent mundane understanding of phenomena as an alternative to the metaphysics of reification that Ngrjuna criticizes.

Verse 10 is central in this discussion:

10. If things did not exist
Without essence,
The phrase, “When this exists so this will be,”
Would not be acceptable.


Ngrjuna is replying here to the causal realist’s inference from the reality of causal powers to their embodiment in real entities whose essences include those powers. He turns the tables on the realist, arguing that it is precisely because there is no such reality to things—and hence no entities to serve as the bearers of the causal powers the realist wants to posit—that the Buddhist formula expressing the truth of dependent arising32 can be asserted. It could not be asserted if in fact there were real entities. For if they were real in the sense important for the realist, they would be independent. So if the formula were interpreted in this context as pointing to any causal power, it would be false. It can only be interpreted, it would follow, as a formula expressing the regularity of nature.33

In the next three verses (I: 11–13), Ngrjuna anticipates and answers the causal realist’s reply:

11. In the several or united conditions
The effect cannot be found.
How could something not in the conditions
Come from the conditions?


Here the realist argues that the conclusion Ngrjuna draws from the unreality of causal power—the nonexistence of things (where “existence” is read “inherent existence”)—entails the falsity of the claim that things dependently arise. For if there are no things, surely nothing arises. This charge has a double edge: If the argument is successful it not only shows that Ngrjuna’s own position is vacuous, but that it contradicts one of the most fundamental tenets of Buddhist philosophy—that all phenomena are dependently arisen. Moreover, the opponent charges, on Ngrjuna’s view that the explanandum is not to be found potentially in the explanans, there is no explanation of how the former is to be understood as depending upon the latter. As Ngrjuna will emphasize in I: 14, however, the very structure of this charge contains the seeds of its reply. The very emptiness of the effect, an effect presupposed by the opponent to be non���empty, in fact follows from the emptiness of the conditions and of the relationship between conditions and effect. Ngrjuna will, hence, reply to the opponent’s attempted refutation by embracing the conclusion of his reductio together with the premises it supposedly refutes.

12. However, if a nonexistent effect
Arises from these conditions,
Why does it not arise
From nonconditions?


How, the opponent asks, are we to distinguish coincidental sequence from causal consequence, or even from conventional dependence? And why don’t things simply arise randomly from events that are nonconditions since no special connection is posited to link consequents to their proper causal antecedents?

13. If the effect’s essence is the conditions,
But the conditions don’t have their own essence,
How could an effect whose essence is the conditions
Come from something that is essenceless?


Finally, the opponent asks, since the phenomena we observe clearly have natures, and since those natures clearly derive from their causes, how could it be, as Ngrjuna argues, that they proceed by means of a process with no essence, from conditions with no essence? Whence do the natures of actual existents arise? Ngrjuna again will reply to this last charge by pointing out that since on his view the effects indeed have no essence, the opponent’s presupposition is ill-founded. This move also indicates a reply to the problem posed in I: 12. That problem is grounded in the mistaken view that a phenomenon’s lack of inherent existence entails that it, being nonexistent, could come into existence from nowhere. But “from nowhere,” for the opponent, means from something lacking inherent existence. And indeed, for Ngrjuna, this is exactly the case: Effects lacking inherent existence depend precisely upon conditions that themselves lack inherent existence.

Ngrjuna’s summary of the import of this set of replies is terse and cryptic. But unpacking it with the aid of what has gone before provides an important key to understanding the doctrine of the emptiness of causation that is the burden of this chapter:

14. Therefore, neither with conditions as their essence,
Nor with nonconditions as their essence are there any effects.
If there are no such effects,
How could conditions or nonconditions be evident?


First, Ngrjuna points out, the opponent begs the question in asserting the genuine existence of the effects in question. They, like their conditions, and like the process of dependent origination itself, are nonexistent from the ultimate point of view. That is, they have no essence whatever. Hence, the third charge fails. As a consequence, in the sense in which the opponent supposes that these effects proceed from their conditions—namely that their essence is contained potentially in their causes, which themselves exist inherently—these effects need not be so produced. And so, finally, the effect-containing conditions for which the opponent charges Ngrjuna with being unable to account are themselves unnecessary. In short, while the reificationist critic charges the Mdhyamika with failing to come up with a causal link sufficiently robust to link ultimately real phenomena, for the Mdhyamika philosopher the core reason for the absence of such a causal link is the very absence of such phenomena in the first place.

We are now in a position to characterize explicitly the emptiness of causation and the way this doctrine is identical with the doctrine of dependent origination from conditions adumbrated in this chapter. It is best to offer this characterization using the via media formulation most consonant with Ngrjuna’s philosophical school. We will locate the doctrine as a midpoint between two extreme philosophical views. That midpoint is achieved by taking conventions as the foundation of ontology, hence rejecting the very enterprise of a philosophical search for the ontological foundations of convention (Garfield 1990). To say that causation is nonempty, or inherently existent, is to succumb to the temptation to ground our explanatory practice and discourse in genuine causal powers linking causes to effects. That is the reificationist extreme that Ngrjuna clearly rejects. To respond to the arguments against the inherent existence of causation by suggesting that there is then no possibility of appealing to conditions to explain phenomena—that there is no dependent origination at all—is the extreme of nihilism, also clearly rejected by Ngrjuna. To assert the emptiness of causation is to accept the utility of our causal discourse and explanatory practice, but to resist the temptation to see these as grounded in reference to causal powers or as demanding such grounding. Dependent origination simply is the explicability and coherence of the universe. Its emptiness is the fact that there is no more to it than that.

Keep this analysis in mind, for when we reach Chapter XXIV, in which the most explicit analysis of emptiness itself and of the relation of emptiness to the conventional world is articulated, we will see that the principal philosophical move in Ngrjuna’s demystification of emptiness was this attack on ���a reified view of causality. Ngrjuna replaces the view shared by the metaphysician and the person in the street, a view that presents itself as common sense, but is in fact deeply metaphysical, with an apparently paradoxical, thoroughly empty, but in the end commonsense view not only of causation, but of the entire phenomenal world. This theme—the replacement of apparent common sense that is deeply metaphysically committed with an apparently deeply metaphysical but actually commonsense understanding of the phenomenal world—will recur in each chapter of the text.


Chapter II

Examination of Motion


The target of Ngrjuna’s arguments in this chapter is any view of motion according to which motion is an entity, or a property with an existence independent of that of moving things, or according to which motion is part of the nature of moving things. These are versions of what it would be to think of motion as nonempty. It might be quite natural for a reificationist to reply to the arguments in Chapter I by proposing that such a view must be the case. For in Chapter I Ngrjuna does presuppose, in developing the view that conventionally things do arise dependent upon conditions, that there is motion, or change. For if there were not, there would be no arising. And as we have seen, this would indeed be an absurd consequence for Ngrjuna. So, one might think, even if the links between conditions and their consequences are empty, the change represented by the arising of these consequences must be real.

Ngrjuna argues that from such a view a number of absurd consequences would follow: Things not now in motion, but which were in motion in the past or which will be in the future, would have to undergo substantial change, effectively becoming different things when they change state from motion to rest or vice versa; a regress would ensue from the need for the entity motion itself to be in motion; motion would occur in the absence of moving things; the moment at which a thing begins or ceases motion would be indescribable. Ngrjuna concludes that a reification of motion is incoherent. Motion is therefore empty.

1. What has been moved is not moving.
What has not been moved is not moving.
Apart from what has been moved and what has not been moved,
Movement cannot be conceived.


That is, if motion exists, there must be sometime at which it exists. Ngrjuna in this opening verse considers the past and the future. This makes goo���d sense. For motion requires a change of position, and a change of position must occur over time. But the present has no duration. So if motion were to exist, it would have to exist either in the past or in the future. But a thing that has moved only in the past is not now moving. Nor is a thing yet to be moved. One might, of course, suggest that there is a simple tense fallacy here—that things that were moving in the past were then in motion, that things that will move in the future will then be in motion. But this would be problematic. For that would mean that all motion would be in the past or in the future, and this could be said at any time. So there would be no time at which it would be true of any thing that it is in motion.34 But this intuition is behind the opponent’s reply in the next verse:

2. Where there is change, there is motion.
Since there is change in the moving,
And not in the moved or not-moved,
Motion is in that which is moving.


This verse is important not only because it announces the obvious reply that motion exists in presently moving things, but because it introduces the connection between change in general and motion. Though this interpretative point is controversial, and several scholars have given widely different interpretations,35 it is highly plausible that Ngrjuna is calling attention to the fact that the attack on motion as an inherently existent phenomenon is a general attack on seeing change or impermanence as inherently existent. This suggests that even the properties that according to Buddhist philosophy characterize all things—being dependently arisen and being impermanent—are not themselves inherently existent. Ngrjuna replies:

3. How would it be acceptable
For motion to be in the mover?
When it is not moving, it is not acceptable
To call it a mover.


The point here is that if motion is thought of both as inherently existent and as a property of the mover, then it should, as inherently existent, continue to exist. For something that is inherently existent depends for nothing on its existence, and so it cannot be deprived of the conditions of its manifestation. That is because inherent existence is existence with an essence, as an independent entity whose identity can be intrinsically specified. (See Chapter XV for more detail.) But movers come to rest. It would seem then that it would have to be appropriate to call something a mover, even when it is at rest, since inherently existing motion could not cease.

4. For whomever there is motion in the mover,
There could be nonmotion
Evident in the mover.
But having motion follows from being a mover.


In this verse Ngrjuna begins his attack on the idea that motion is a property with an existence independent of movers. If, he asserts, one were to posit motion as such a property that simply happened to inhere in movers, it would follow from its independence that movers might not have it, but instead its contrary, namely, nonmotion. But that is not tenable. So it follows that motion can’t be thought of as an independent property. This line of argument is continued in the next two verses:

5. If motion is in the mover.
There would have to be a twofold motion:
One in virtue of which it is a mover,
And one in virtue of which it moves.


6. If there were a twofold motion.
The subject of that motion would be twofold.
For without a subject of motion,
There cannot be motion.


Here Ngrjuna develops a reductio on a position according to which motion is a property of the mover only at the time that the mover is in motion. This might seem to be a much more plausible view than the earlier discussed view of motion as an essential property. But Ngrjuna argues that this can’t work either. For it involves a multiplication of movements and agents of motion that is unacceptable to the proponent of such a theory. For if the motion is a property of the mover at all, both the mover and the motion must be moving. And this amounts to two separate motions. One motion—that in virtue of which the mover is a mover in the first place—is the motion posited by the theory. But if that motion were stationary, the mover would either also not be moving or it would “outrun” its motion and leave it behind. So there must also be a motion of the motion. Each of these two motions requires a subject. They can’t be the same subject because then the mover and the motion would be identical, which would be absurd. So in explaining the motion of a single individual, the opponent is stuck with two movers.

This argument clearly can be understood as the start of an infinite regress. It is not at all clear whether Ngrjuna so intended it, as the context in which the argument is formulated is one in which the consequence that two movers emerge in the analysis of the motion of a single mover is enough to refute the opponent.36 But it is important to see that once this multiplication of explanatory motions and agents begins, it cannot be stopped, and so this argument constitutes a perfectly ���general attack on a view according to which motion is an entity associated with movers. It is also worth noting that the argument generalizes in other ways: It can be formulated as an argument against a parallel analysis of change as an independent property and, in general, as an argument against properties as entities that inhere in subjects—a twofold redness is required for a red shirt to be red because of the possession of redness. So this is, in fact, a “third man” argument.

7. If without a mover
It would not be correct to say that there is motion,
Then if there were no motion,
How could there be a mover?


Ngrjuna is here emphasizing the codependence of motion and the mover. If there are no movers, there is no motion. If there is no motion, there are no movers. This has import at both the conventional level and with respect to any discussion of the inherent existence of either the mover or motion. At the conventional level, it means that any analysis of either motion or the mover that leaves the other out, or that does not involve codependence, will fail. Neither can be established as an independent basis for the analysis of the other. But it also means that neither, therefore, can be thought to inherently exist since to exist inherently would be to exist independently.

8. Inasmuch as a real mover does not move,
And a nonmover does not move,
Apart from a mover and a nonmover,
What third thing could move?


Here the terms “mover” and “nonmover” must be understood in the context of the previous arguments. Ngrjuna is clearly talking about entities that are essentially in motion or in nonmotion. He has argued that we cannot think of a thing in motion as a thing whose nature is to move. And clearly a thing whose nature is not to move cannot be in motion. So if motion is thought of as a property that is either part of the nature of a thing or incompatible with a thing’s nature, we are left with the conclusion that there is no motion. And so we have a philosophical problem: How is ordinary motion (and change) possible? Ngrjuna emphasizes this in the following verses:

9. When without motion,
It is unacceptable to call something a mover,
How will it be acceptable
To say that a mover moves?


10. For him from whose perspective a mover moves,
There would be the consequence that
Without motion there could be a mover.
Because a mover moves.


These verses recapitulate the argument in II: 4 and II: 7. If we simply regard motion and mover as independent phenomena, we are forced to the absurd consequence that either could be present without the other.

11. If a mover were to move,
There would be a twofold motion:
One in virtue of which he is a mover,
And one in virtue of which the mover moves.


This last verse recapitulates the important argument in II: 6 in preparation for the attack on the possibility of the beginning and end of motion. The next few verses are reminiscent both of Zeno of Elea and Sextus Empiricus:

12. Motion does not begin in what has moved,
Nor does it begin in what has not moved.
Nor does it begin in what is moving.
In what, then, does motion begin?


13. Prior to the beginning of motion,
There is no beginning of motion in
The going or in the gone.
How could there be motion in the not-gone?


These two verses are alternative formulations of the same argument: If there is motion, it must begin sometime. But that moment is inconceivable. For motion doesn’t begin in a stationary thing. And once a thing is in motion, it is too late. It can’t always have begun in the past or be yet to begin, and there simply isn’t time to go anywhere in the present.

14. Since the beginning of motion
Cannot be conceived in any way.
What gone thing, what going thing,
And what nongoing thing can be posited?


After having emphasized this point, Ngrjuna points out that all that has been said about motion (and hence implicitly about change) applies, mutatis mutandis, to rest (and hence implicitly to stasis). Things that are in motion cannot be simultaneously at rest. But to say that a stationary thing is at rest, where rest is conceived as a property or entity having independent existence, would involve us in the same paradoxes encountered above: The stasis itself would have to be either in motion or at rest. If in motion, then the static thing would have to be in motion, which is contradictory. But if at rest, then it must be at rest in virtue of possessing sta���sis, and we are off on the same regress:

15. Just as a moving thing is not stationary,
A nonmoving thing is not stationary.
Apart from the moving and the nonmoving,
What third thing is stationary?
16. If without motion
It is not appropriate to posit a mover,
How could it be appropriate to say
That a moving thing is stationary?


And, in the same fashion, all that applies to the initiation of motion applies mutatis mutandis, to its cessation:

17. One does not halt from moving,
Nor from having moved or not having moved.
Motion and coming to rest
And starting to move are similar.


Ngrjuna now develops further problems with any view regarding motion as an entity; it must be either identical to or different from the mover. Both options, he will argue, turn out to be incoherent:

18. That motion just is the mover itself
Is not correct.
Nor is it correct that
They are completely different.
19. It would follow from
The identity of mover and motion
That agent and action
Are identical.


The identity of agent and action is absurd on its face. For then whenever an agent were to perform another act, s/he would become a distinct agent. There would be no basis for identifying individuals over time.

20. It would follow from
A real distinction between motion and mover
That there could be a mover without motion
And motion without a mover.


This is more complicated. It is important to recall that the target positions here are positions that reify motion as a distinct entity, however abstract. If motion were an entity, and were distinct from all movers, then it should be possible��� to separate motion from movers.37 Then we should see motion when nothing is moving and movers that are not in motion. Noticing that this is a problem for Ngrjuna’s opponent provides us with a hint as to the positive account of conventional motion that we should take from this chapter to be discussed below: Motion can only be understood in relation to movers—as a relation between their positions at different times. Movers can only be understood as movers in relation to motion so understood. But to understand motion and movers this way is not to reify them as entities—and so to escape the dilemma of their identity or difference. Ngrjuna emphasizes this moral in the next verse, where we must read “established” as meaning established as existent entities.

21. When neither in identity
Nor in difference
Can they be established,
How can these two be established at all?


22. The motion by means of which a mover is manifest
Cannot be the motion by means of which he moves.
He does not exist before that motion,
So what and where is the thing that moves?


In this verse and in the next, Ngrjuna is simply emphasizing the interdependence of motion and the mover. In II: 22 he notes the absurdity of the supposition that the mover and the motion are known independently. If they could be, then the mover would have to have one motion in virtue of which he was a mover and a second independent motion in virtue of which he now moves. But since prior to being in motion, no mover exists, it cannot be that the mover exists as a mover independently of the motion. This then demands an answer to the question, What moves?

In II: 23 Ngrjuna answers this in a very straightforward way: The mover who is a mover in virtue of his motion (and that motion is a motion in virtue of being carried out by a mover) is what moves. Hence, the mover is dependent for his identity as a mover on the motion; the motion is dependent for its identity on the mover. Neither has an intrinsic identity, and both are empty of inherent existence:

23. A mover does not carry out a different motion
From that by means of which he is manifest as a mover.
Moreover, in one mover
A twofold motion is unacceptable.


2���4. A really existent mover
Doesn’t move in any of the three ways.
A nonexistent mover
Doesn’t move in any of the three ways.


The three ways in question are past, present, and future. Something that is inherently a mover has been shown to be incapable of motion in any of these periods. This is simply a way of emphasizing the moral of the entire chapter: Movers cannot be thought of as being movers intrinsically. Moreover, nonexistent movers—movers that are not even conventionally movers—certainly don’t move. It must therefore be that neither do movers move intrinsically nor that there is no motion. There must be a sense in which motion and movers exist, but do not do so intrinsically. The final verse must hence be read with “entity,” “nonentity,” and “existent” as asserted in the ultimate sense:

25. Neither an entity nor a nonentity
Moves in any of the three ways.
So motion, mover,
And route are nonexistent.


So far so good. But then is motion completely nonexistent? Is the entire universe static according to Mdhyamika philosophy? If we simply read this chapter in isolation, that conclusion might indeed seem warranted. It would be hard to distinguish emptiness from complete nonexistence. We would be left with an illusory world of change and movement, behind which would lie a static ultimate reality. But such a reading would be problematic. For one thing, it would be absurd on its face. Things move and change. For another, it would contradict the doctrine of dependent origination and change that is the very basis of any Buddhist philosophical system, which Ngrjuna has already endorsed in the opening chapter. How, then, are we to read this discussion more positively? Answering this question is hermeneutically critical not only for an understanding of this chapter, but for a reading of the entire text, which if not read with care, can appear unrelentingly nihilistic.

The positive account we are after emerges when we read this second chapter in the context of the first chapter: All phenomena, including motion, are dependently arisen and, hence, empty of inherent existence. The conclusion that motion is empty is simply the conclusion that it is conventional and dependent, like the putatively moving entities themselves. Since there is no implicit contrastive, inherently existent, ultimate reality—say of the static, or of stasis—this conclusion does not lead us to ascribe a “second class” or merely apparent existence to motion or to movers. Their nonexistence is simply their lack of existence as substantial entities. Existence—of a sort—is hence recovered exactly in the context of an absence of inherent existence.

But existence of what kind? Herein lies the clue to the positive construction of motion that emerges. The existence that emerges is a conventional and dependent existence. Motion does not exist as an entity on this account, but rather as a relation—as the relation between the positions of a body at distinct times and, hence, as dependent upon that body��� and those positions.38 Moreover, it emerges as a conventional entity in the following critical sense: Only to the extent that we make the decision to identify, as a single entity, things that differ from each other in position over time, but are in other respects quite similar and form causal chains of a particular sort, can we say that whatever is so identified moves. And this is a matter of choice. For we could decide to say that entities that differ in any respect are thereby distinct. If we did adopt that convention for individuation, an entity here now and one there then would ipso facto be distinct entities. And so no single entity could adopt different positions (or different properties) at different times, and so motion and change would be nonexistent. It is this dependence of motion on the moved, of the status of things as moved on their motion, and of both on conventions of individuation that, on this account, constitutes their emptiness. But this simply constitutes their conventional existence and provides an analysis of the means by which they so exist. The emptiness of motion is hence seen to be its existence as conventional and as dependent, not other than its conventional existence. In understanding its emptiness in this way, we bring motion, change, and movable and changeable entities back from the brink of extinction.39


Chapter III

Examination of the Senses


In this chapter, which is most immediately about vision, Ngrjuna really addresses the status of sense perception generally, as he makes clear in the opening and closing verses. Just as in Chapter II, where the target positions Ngrjuna argues against are positions according to which motion and the mover inherently exist as distinct, independent, but somehow related entities, here he argues against positions according to which the sense faculties, the sense organs, the subject of sensory experience, and the sense object inherently exist and are distinct, independent, but somehow related entities. For we do perceive motion and change, and the argument for the conventional existence of motion did suggest that it could be seen as a relation between the positions at which we perceive objects at different times. So one can imagine an opponent saying, “Even if the motion we perceive is not real, the perception must be.” Again, it will be important for Ngrjuna that his analysis of perception as empty of inherent existence, and as merely dependently arisen, does not entail its complete nonexistence. He must, that is, steer a middle path between reification and nihilism using emptiness as his compass.

1. Seeing, hearing, smelling,
Tasting, touching, and mind
Are the six sense faculties.
Their spheres are the visible objects, etc….


This is a standard Buddhist catalog of the sense faculties. It differs from the standard Western catalog only in that t���he Buddhists regard introspection literally as an inner sense with the same epistemic structure as outer senses and presumably subserved by analogous physical structures. Ngrjuna will not dispute the reality of these faculties or of their respective spheres. But he will insist that that reality must be characterized interdependently and conventionally.

2. That very seeing does not see
Itself at all.
How can something that cannot see itself
See another?


This cryptic argument is aimed at any theory according to which vision is inherently existent. The idea is this: If the visual faculty were to be inherently existent, then seeing would be its essence. Its action would hence require no distinct conditions and no external object to be seen. That is, if vision were inherently existent, vision would occur simply in virtue of the existence of the visual faculty. Suppose then that there is an inherently existent visual faculty and no external sense object for it. It would then have only itself as a possible object of sight, yet it would be seeing and so would have to be seeing itself. Therefore, Ngrjuna argues, a view of vision as inherently existent would entail the possibility of visual apperception. But there is no such possibility. So the fact that vision can see other things cannot be in virtue of its containing percipience as an inherent property.

There is also a plausible Pyrrhonian interpretation of this verse: The point of a sensory faculty is to make knowledge possible. But that is only possible if the data the faculty provides are themselves perceived. But the data that the visual faculty delivers are visual. If they themselves are to be perceived, one would require either another visual faculty, hence generating a vicious regress, or apperception by vision, which is absurd. The point is not then that vision is impossible, but rather that visual perception—or any kind of perception—can only be completely explained and characterized by reference to things outside of the visual faculty itself. Vision is relational, and not an intrinsically identifiable phenomenon.40

3. The example of fire
Cannot elucidate seeing.
Along with the moved and not-moved and motion
That has been answered.


This is a reply to a standard substantialist counterexample to a Mdhyamika analysis, specifically: Fire burns other things, but does not burn itself. And it can be intrinsically identified. Perhaps then vision is like fire, in that it can see others but not itself, while it does not need to be relationally identified. This example is a standard in early Buddhist debates about intrinsic versus relational identity, and Ngrjuna devotes an ent���ire chapter to its refutation as a dialectical device (Chapter X), arguing there that fire cannot be intrinsically identified. But at this point, he is willing to grant the opponent that premise for the sake of argument. For, he claims, its utility as an analogy has already been undermined by the argument in the second chapter.

How? Whatever fire is burning must be burned in the past, the future, or the present. But, as with motion, burning cannot be, by its very nature, in the past, on pain of regress. Nor can it be in the future for the same reason. But burning cannot take place in the present either, for there is not enough time in an instant for anything to burn. Mutatis mutandis for vision. In the case of vision, for Ngrjuna, there is a further problem with vision of another in the present. The visual process—and any sensory process—takes time. So if vision is seeing another thing, the other thing is already past. The only thing that vision could see in the present is a visual sense-impression. But then we are back to the problem of visual apperception. So even if fire were intrinsically identifiable, there is no point at which it could burn another. And if vision were intrinsically identifiable, there would be no moment at which it could see another.

4. When there is not even the slightest
Nonseeing seer,
How could it makes sense to say
That seeing sees?


When all there is to vision is visual perception, what is the motivation for positing an entity to undertake the process of perception? All there is to vision is the perceptual process: We don’t need to posit an entity—the visual faculty over and above the set of interdependent phenomena that subserve vision. The desire to do so is of a piece with the more general substantialist imperative to posit an independent substratum to support every capacity or property.

5. Seeing itself does not see.
Nonseeing itself does not see.
Through seeing itself
The clear analysis of the seer is understood.


Perception is not accomplished by any independent entity known as vision. But that doesn’t mean that things that are incapable of sight thereby perceive. In order to know what the proper subject of vision is, it is important to undertake a careful analysis of the perceptual process and not simply to posit a faculty with the nature of vision.

6. Without detachment from vision there is no seer.
Nor is there a seer detached from it.
If there is no seer
How can there be seeing or the seen?


On Ngrjuna’s analysis, we can���’t make sense of an autonomous subject of visual perception. For such a subject would by definition have its identity as a visual subject independent of perception. But there is no sense in calling something that does not see a seer. On the other hand, if we pack vision into its definition, we thereby fail to identify the subject nonrelationally. Vision and its subject are thus relational, dependent phenomena and not substantial or independent entities. So neither seeing nor seer nor the seen (conceived of as the object of sense perception) can be posited as entities with inherent existence. The point is just that sense perception cannot be understood as an autonomous phenomenon, but only as a dependent process.

7. Just as the birth of a son is said to occur
In dependence on the mother and father,
So consciousness is said to arise
In dependence on the eye and material form.41


Here the opponent offers yet another argument in favor of the inherent existence of the visual faculty (and, by extension, the other sense faculties): Consciousness is a consequence of vision, and it surely exists—in fact, its existence, one might say, is self-validating. Given the reality of the effect, the cause must also be real.42 The final two verses reply to this objection and state the obvious generalization to all other senses, sense objects, sense faculties, and faculties of knowledge. The reply consists in pointing out that the other faculties and aggregates, including introspection and consciousness, exist and fail to exist in exactly the senses that vision and its objects exist and fail to exist: All are empty of inherent independent existence. But all exist conventionally. So the effect that, according to this interlocutor, exists inherently and demands an inherently existent cause does not so exist. And in the sense that it exists, its causes also exist:

8. From the nonexistence of seeing and the seen it follows that
The other four faculties of knowledge do not exist.
And all the aggregates,43 etc.,
Are the same way.


9. Like the seen, the heard, the smelled,
The tasted, and the touched,
The hearer, sound, etc.,
And consciousness should be understood.


Again, the point of this chapter is emphatically not that there is no perception, or that there are no sense faculties, sense organs, or sense objects. Rather the point is that none of these can be analyzed successfully as autonomous entities. They are interdependent phenomena that depend for their existence and their character on each other. None of them exists independently. They are all, hence, empty of inherent existence, and carving the process of perception into ���these components represents a conventional taxonomy of a process that does not present itself with natural joints demanding cleavage on their own.


Chapter IV

Examination of the Aggregates


The five aggregates are the basic Buddhist categories of personal constituents. The first—that discussed as an example in this chapter—is in Sanskrit rpa, in Tibetan gzugs. Unfortunately, given the lexicography of Western philosophy, this word has historically been translated as “form.” This practice is so ubiquitous that I am loathe to depart from it, despite the confusion it engenders. For what the word means is matter. The other aggregates are sensation, perception, intellect, and the dispositions. It is important to realize that this taxonomy is to be understood pragmatically: There is no deep doctrinal or philosophical point that hangs on dividing the properties or capacities of humans up in just this way. In fact, most often the only important point about analysis in terms of the aggregates is that humans are composite. The precise nature of the best decomposition is of interest to psychology and to soteriological practitioners, but is at bottom, from the standpoint of the tradition, an empirical matter.44

This chapter is motivated by the natural suggestion that even if vision itself is empty, as was argued in the previous chapter, there must be a truly existent basis for vision in the person and his/her faculties. For the emptiness of vision was established in part by showing that perception depends upon the perceiver and the perceived. And that might seem to suggest that these bases—or at least the most essential one, the perceiver—truly exist. For then one could say that whereas vision itself is not inherently existent, it does exist as a relation between an inherently existent perceiver and an inherently existent object, or at least as a property of such a perceiver, even if there is truly no object.45 Ngrjuna aims to demonstrate the emptiness of all of the constituents of the person by taking form as an example and applying arguments that are general in scope. Form is taken as an example precisely because it is the most solid, apparently nonempty of the aggregates—the one that we are most likely to reify. So the program is to use arguments with application to any of the aggregates and to apply them to the hardest case. The conclusion Ngrjuna is after is that no decomposition of the person will yield constituents that are themselves independent and nonempty.

1. Apart from the cause of form,
Form cannot be conceived.
Apart from form,
The cause of form is not seen.


Ngrjuna begins by making use of the results of the first chapter. Nothing arises causelessly, and no cause is ineffectual. So if any form exists, it exists with a��� cause. And if the cause of any form exists, so does that form. But there is an interesting problem to be posed: How about form itself—matter considered in general, not in its specific instances? Does it have a cause or not? This question is important because it gets at the question of whether we can imagine ultimate ontological categories that exist independently. If form has a cause at all, it must be either the same or different from form. If the former, we have an infinite regress. If the latter, then we have the absurd conclusion that immaterial things can cause material things to come into existence. If it has no cause, then it cannot be said to exist at all.46

2. If apart from the cause of form, there were form,
Form would be without cause.
But nowhere is there an effect
Without a cause.


If form as such exists without any cause, we would have an example of an inherently existent category. But that would also violate the principle of dependent origination. That is, both Ngrjuna and his opponent agree that all phenomena are dependently originated, and the discussion in the present chapter is in fact directed at figuring out just what material form depends on. So an attempt to posit material form as inherently existent on the grounds that it comes into existence causelessly is an ad hoc move that is unavailable to any participant in this debate.

Moreover, Ngrjuna points out in the next verse, if we held form to be dependent upon a cause that was itself inherently existent, we would have an inherently existing cause without an inherently existing effect. That putative cause would, hence, fail to be a cause in the full sense. Between genuine causes and their effects there is a relation of dependence. For something to count as a cause independent of its producing an effect would be incoherent. But since in the context of inherent existence merely conventional existence counts as no existence at all, an inherently existent cause with a merely conventionally existent effect would count just as much as an ineffective cause. So neither can we make sense of an inherently existent cause of the existence of material form if material form is held not to be inherently existent.

3. If apart from form
There were a cause of form,
It would be a cause without an effect.
But there are no causes without effects.


4. When form exists,
A cause of the arising of form is not tenable.
When form is nonexistent,
A cause of the arising of form is not tenable.


Any relation���ship between form and a putative cause is unintelligible, Ngrjuna argues, following closely the reasoning in Chapter I. If form exists, the cause has ceased to exist. If form does not exist, the cause cannot have existed. This might seem at first glance to be a wholesale rejection of the possibility of dependency of effects on causal conditions. But if we recall the moral of Chapter I and keep the dialectical context of the current chapter firmly in mind we will see that this is not so: The paradox of causal contact arises—as Sextus also notes—only if we suppose that the causes we appeal to in explanation must have some special force by means of which they bring about their effects. That, as we have seen, is the view of the causal link as inherently existent and, hence, of causes as inherently existent. The opponent Ngrjuna is attacking in this chapter is one who thinks that form/matter is inherently existent, but who has granted that all individual phenomena—all particular forms, such as human bodies, tables, and chairs—are dependently arisen. So the opponent agrees that all phenomena must be explicable. But the opponent wants to reify form, and that is to treat it as a phenomenon—albeit an inherently existent one. Therefore, it must, for the opponent, have an explanation of its existence, and since its existence is inherent existence, it must be an explanation in terms of inherently existent causation. So all that Ngrjuna has to do is to remind the opponent of the incoherence of that notion in order to undermine the view that form as such is inherently existent. The coherence of conventional dependent origination is not at issue.

5. Form itself without a cause
Is not possible or tenable.
Therefore, think about form, but
Do not construct theories about form.


The moral of these arguments, Ngrjuna concludes, is that we cannot think of form as such as an entity at all. Individual forms are entities—dependently arisen ones, hence, empty of inherent existence. But form itself is an abstraction, neither caused nor uncaused, but dependent upon the existence of material things with form. (Moreover, were one to argue that form itself exists as an entity, one would be faced with an uncomfortable dilemma: Its existence would be caused or uncaused. The latter alternative patently begs the question regarding the explanation of the existence of the material world. But the former issues in a further dilemma: The cause would either itself be material or immaterial. On the first horn, we have an infinite regress; on the second, the inexplicable causation of the material by the immaterial.)47 So, he advises, think carefully about what form is and about the nature of particular material objects. But do not construct theories that purport to describe the essence of material form. For there is no such thing. It is simply a characteristic of individual material objects and, hence, something that depends upon their existence, with no essence of its own.

6. The assertion that the effect and cause are similar
Is not acceptable.
The assertion that they are not similar
Is also not acceptable.


We cannot say that nonmaterial things give rise to the existence of matter, for that would be an inexplicable miracle. Nor can we say that matter gives rise to matter, since that would beg the question. But there is no other possibility. So despite the reificationist’s intuition that though individual material objects may be empty, the matter they are made of is nonempty, we see that we cannot even clearly conceive of the nature of matter as such independently of material objects. Matter, too, is hence dependent and empty of inherent existence. Ngrjuna immediately generalizes this to the other aggregates:

7. Feelings, discriminations and dispositions
And consciousness and all such things
Should be thought of
In the same way as material form.


8. When an analysis is made through emptiness,
If someone were to offer a reply,
That reply will fail, since it will presuppose
Exactly what is to be proven.


9. When an explanation is made through emptiness,
Whoever would find fault with it
Will find no fault, since the criticism will presuppose
Exactly what is to be proven.


In these last two oft-quoted verses, Ngrjuna claims that once a demonstration of the emptiness of a phenomenon or class of phenomena has been produced, any reply will inevitably beg the question. And this is meant to have been demonstrated by the argument in this chapter in the following way: Once we have shown something to be empty of inherent existence, we have, ipso facto, shown it to be dependently arisen and merely conventionally real. Anything an opponent would want to demonstrate to be inherently existent would fall prey to the causal paradoxes developed in this chapter. That is, he must either assume that the thing is completely independent and causeless, which is, upon analysis, exactly equivalent to the conclusion he is out to prove, or that it arises from another inherently existent phenomenon. But then in order to demonstrate that fact, he must demonstrate the inherent existence of that second phenomenon (as well as the inherent dependence relation between them—a kind of relation we have seen to be internally c���ontradictory). And this is true no matter to which ontological category the putatively inherently existent phenomenon belongs.

That this is so should not be surprising, for the central thrust of Ngrjuna’s arguments thus far, and throughout Mlamadhyama-kakrik, is not that inherent existence is a property some things might have had but by global accident is uninstantiated or that emptiness just happens to characterize all phenomena. Rather he is arguing that inherent existence is simply an incoherent notion and that emptiness is the only possible analysis of existence. It would follow straightforwardly that arguments for inherent existence will be question begging.48


Chapter V

Examination of Elements


This chapter examines the ontological status of characteristics and the characterized, or in more familiar terms, properties and individuals. The question, as always, is this: Does it make sense to think of either as existing independently, substantially, or fundamentally? Or, on the other hand, are they mutually interdependent and therefore empty? The example Ngrjuna chooses to focus on is space since it is one of the six primal elements according to classical Buddhist cosmology.49 If he can show that these elements must be understood as neither inherently existing entities nor as inherently existing characteristics of entities, he will have shown that no ontological decomposition of phenomena into their primary constituents yields inherently existing constituents. Moreover, according to some early Buddhist schools, each of the primal elements has a distinguishing characteristic and, hence, an essence. So, Ngrjuna is addressing his opponent on the opponent’s home turf. If any entities or characteristics have essences, these do.

1. Prior to a characteristic50 of space
There is not the slightest space.
If it arose prior to the characteristic
Then it would, absurdly, arise without a characteristic.


Space cannot exist as a completely uncharacterized entity that then somehow acquires characteristics. Anything that exists has some properties and cannot be identified or characterized independently of them.

2. A thing without a characteristic
Has never existed.
If nothing lacks a characteristic,
Where do��� characteristics come to be?


So we can conclude that everything has characteristics. But maybe these characteristics exist inherently, independently of the things, and then come to be associated with them. On such a view, while individuals would not have inherent existence, properties would.

3. Neither in the uncharacterized nor in the characterized
Does a characteristic arise.
Nor does it arise
In something different from these two.


But there is a problem. If a characteristic were inherently existent, it would have to become instantiated in either a characterized or an uncharacterized object. But there are no uncharacterized objects, and if the object already is characterized, there is no need for the characteristic to become instantiated. So to think of individuals and properties as existing independently and then somehow coming together to constitute particulars makes no sense.

4. If characteristics do not appear,
Then it is not tenable to posit the characterized object.
If the characterized object is not posited,
There will be no characteristic either.


But if we were to go completely eliminativist with respect to characteristics, we would lose the ability to posit both actual objects with characteristics and characteristics that actual objects share.

5. From this it follows that there is no characterized
And no existing characteristic.
Nor is there any entity
Other than the characterized and the characteristic.


In the first two lines of this verse, Ngrjuna draws the conclusion that there are no inherently existent characteristics and no inherently existent characterized entities. Entities and their properties are mutually dependent and, hence, empty of inherent existence. But this does not mean, he emphasizes in the final two lines, that there is some other ontology of inherently existent basic types that could replace them. Indeed particulars can be thought of as characterized entities, with characteristics; but this does not entail the independent existence of entities of either of those types.

6. If there is no existent thing,
Of what will there be nonexistence?
Apart from existent and nonexistent things
Who knows existence and nonexistence?


Here Ngrjuna generalizes the conclusion and indicates its larger ontological implications. Having shown that there are no inherently existent things, it might seem that it follows that all things are inherently nonexistent. But existence and nonexistence, after all, are characteristics. So it follows that neither existence nor nonexistence can be said to exist independently and hence to characterize, inherently, anything. Moreover, since no particulars can be said inherently to exist, and thereby characterized as inherently existing things, none can be said to be inherently nonexistent. Existence and nonexistence are hence themselves dependent, relative characteristics. It is, of course, important to recall that this entire dialectic is aimed at nonrelative understandings of existence and nonexistence. Ngrjuna is not arguing that nothing exists in any sense and that nothing fails to exist in any sense. Rather, he is arguing that nothing exists in virtue of instantiating an independently existent property of existence. Similarly, things do not fail to exist in virtue of instantiating the property nonexistence.

7. Therefore, space is not an entity.
It is not a nonentity.
Not characterized, not without character.
The same is true of the other five elements.


Ngrjuna now returns to the example at hand to sum up the conclusions of the chapter. Things cannot be analyzed ontologically as particulars existing independently of their properties. But this does not mean that individual things do not exist. They do not possess independently existing properties. But this does not mean that things are all propertyless.

8. Fools and reificationists who perceive
The existence and nonexistence
Of objects
Do not see the pacification of objectification.


This is the soteriological import of this discussion of fundamental ontology: If one reifies phenomena—including such things as one’s own self, characteristics (prominently including one’s own), or external objects—and if one thinks that things either fail to exist or exist absolutely, one will be unable to attain any peace. For one will thereby be subject to egoism, the overvaluing of oneself and one’s achievements and of material things. One will not appreciate the possibility of change, of the impermanence and nonsubstantiality of oneself and one’s possessions. These are the seeds of grasping and craving and, hence, of suffering. The alternative, Ngrjuna suggests, and the path to pacification, is to see oneself and other entities as non-substantial, impermanent, and subject to change and not as appropriate objects of such passionate craving.


Chapter VI

Examination of Desire and the Desirous


This chapter represents a continuation of the discussion begun in the previous one. That is, while the chapter is nominally about desire, an example chosen for its obvious soteriological significance, it is in a larger sense a further discussion of the relation between entities and their properties, with specific attention to the relation between human beings and their psychological characteristics. Locating the discussion at this point is consonant with a tradition of Mahyna discussions of emptiness in which one first addresses external phenomena, which are both easier to analyze and less succeptible of reification than the self, and then generalizes the discussion to human psychological phenomena.51 The chapter opens with an echo of the discussion of space:

1. If prior to desire
And without desire there were a desirous one,
Desire would depend on him.
Desire would exist when there is a desirous one.


One possibility for the relationship between the subject of desire and the desire is that the desirous one exists qua desirous one independently of the desire, which is then adventitious and dependent. That is, on this view the desirous one is inherently desirous, but the desire is merely dependent. This, however, is problematic, for then there is a real contrast in the mode of existence of the desirous one and the desire: The desirous one truly exists, but the desire does not truly exist. But if there is no real desire, in virtue of what is there a desirous one?

2. Were there no desirous one, moreover,
Where would desire occur?
Whether or not desire or the desirous one exist.
The analysis would be the same.


But if there is no desirous one, there is no ontological basis for the desire. So whether we posit an inherently existent desirous one or no desirous one at all, we cannot identify desire as existing. And, of course, this goes for any characteristic or psychological attribute and for any subject of any such attribute identified under any description. Moreover, the converse is also true: Whether or not we posit inherently existent desire, we cannot thereby establish the existence of a substantially existent desirous one. If the desire does not exist inherently but only dependently, that dependence in no way presupposes an independent basis. If on the other hand desire is posited as inherently existent, there would be no need for a basis in a desirous one at all. In neither case would the substantial existence of the entity in question (subject or attitude) have any import for the reality of the correlative entity (attitude or subject). And the reason for this is simply that inherent existence is not relational existence. Since desire and the desirous one must be understood as interrelated, they must be understood as mutually dependent.

3. Desire and the desirous one
Cannot arise together.
In that case, desire and the desirous one
Would not be mutually contingent.


Another possibility the opponent might suggest is this: Desire and the desirous one come into inherent existence at the same time. It is very important in following this argument to remember Ngrjuna’s dialectical task. The opponent against whom his reductios are aimed is one who attributes inherent existence either to the desirous one, to desire, or both. Ngrjuna is only attempting to show that attributing to them that kind of existence is incoherent—not that there is no desire and that there are no desirous people at all. That would be crazy. Fundamental to the Buddhist conception of the predicament of human existence is the centrality of craving to the arising of suffering. But also fundamental is the conviction that there can be a release from craving. That is only possible, however, if craving is dependently originated since only then could the conditions that determine its arising be eliminated. So it is critically important from a Buddhist perspective to come to a complete understanding of the nature of desire, and the mode of its existence, and it would be inconceivable to deny its existence completely. But Ngrjuna is emphasizing here that that understanding must reveal them as mutually dependent in order to avoid the absurd conclusion that either could exist without the other. That precludes the assertion that while they in fact always co-occur, that co-occurrence is not through interdependence, but through contingent simultaneity of independent phenomena.

Ngrjuna’s claim in VI: 3 is also the conclusion of the argument that is about to follow. It proceeds by means of a destructive dilemma. Given that the opponent must have desire and the desirous one arising simultaneously, they must be either identical or different. Ngrjuna will show that neither alternative is coherent; VI: 4 spells out this strategy:

4. In identity there is no simultaneity.
A thing is not simultaneous with itself.
But if there is difference,
Then how would there be simultaneity?


In the first line of this verse, Ngrjuna points out the relational character of simultaneity. If simultaneity is predicated, it must be predicated of two distinct things that arise at the same time. We don’t say that a thing arises simultaneously with itself. But if things are completely distinct in nature, they cannot co-occur in the same place, that is, if desire and the desirous one had distinct essences, they could not be in the same place ���at the same time.

5. If in identity there were simultaneity,
Then it could occur without association.
If in difference there were simultaneity,
It could occur without association.


The first claim is meant to be a reductio on the view that simultaneous things can be identical. For suppose that there was an apparent pair of events whose simultaneity was in question, say William Clinton’s uttering of the oath of office of the presidency and the inauguration of the first president from Arkansas. If there is every reason to believe that these events are distinct but occur at the same time, it is then appropriate to say that they are simultaneous. But if we know that there is in fact only one event, it is at best a joke to assert its simultaneity with itself. The proper thing to say then would be not that the oath taking was simultaneous with the inauguration but that it was identical to the inauguration. The term translated as “association” here (grogspa) can also mean friendship, or companionship—the idea is of something distinct but accompanying. For the inauguration and the oath taking to be associated would be for them to be, say, accompanying rituals that could in principle occur independently. But if they could occur independently, they cannot be identical. Simultaneity requires association of some kind. But identity is incompatible with association.

The second claim is meant to be a reductio on the view that simultaneous and associated things could be different in nature. Difference, like identity, is incompatible with association, though for a different reason. The kind of difference at issue here is essential difference. Ngrjuna’s claim is that things that are completely different from one another, that are completely independent, ipso facto, stand in no relation to one another and so are not associated. This is another application of the Humean (and Tractarian) argument Ngrjuna has mobilized above: If phenomena are distinct—indeed, being simultaneous, they are not even argued to be causally related—they can be imagined to be separate. So they are then logically independent. But that would then entail that if desire and the desirous one were different in this strong sense, we could imagine a desirous one without desire, and vice versa. But that is of course absurd. So if desire and the desirous one are supposed to arise simultaneously, they can neither be identical nor different. Of course, since any inherently different entities, in virtue of having determinate natures, are either identical or different, it follows that desire or the desirous one are either nonsimultaneous or empty of inherent existence.

6. If in difference there were simultaneity,
How could desire and the desirous one,
Being different, be established?
If they were, they would be simultaneous.


This last verse emphasizes and spells out the point scouted above: We are left with a hard choice once we conceive of des���ire and the desirous one as entities. If desire and the desirous one are conceived as substantially different but simultaneous, we would have to be able to establish the nature and existence of each independent of the other. That is no easy task. If we could accomplish it, simultaneity would be a satisfactory solution to the dilemma. But of course we cannot. Moreover, Ngrjuna argues in the next verse, if they are completely different, we are left with the peculiar task of explaining why they always go together. And asserting their simultaneity forces this problem:

7. If desire and the desirous one
Are established as different,
Then why would you think
That they are simultaneous?


8. Since difference is not established,
If you assert that they are simultaneous,
Since they are established as simultaneous,
Do you also assert that they are different?


We have not established—nor could we—that desire and the desirous one are substantially different. But the opponent wishes to assert their simultaneity. Given the entailment of difference by simultaneity as per the argument above, this would force the opponent to assume the impossible burden of demonstrating this substantial difference.

The whole quandary is summed up in VI: 9. Since we can’t establish their difference in entity, we can’t establish the claim that desire and the desirous one arise as distinct, simultaneous phenomena. We don’t even have two phenomena to serve as the relata of difference:

9. Since nothing different has been established,
If one is asserting simultaneity,
Which different thing
Do you want to say is simultaneous?


The conclusion, as stated in the special case in the preceding verse, is generalized in the final verse of the chapter. Once we think of entities and their properties—in particular, ourselves and our characteristics—as independently characterized things, we can make no sense of how they fit together temporally, logically, or ontologically. It is important that objects and their characteristics, persons and their states, be unified. But if we introduce essence and entity into our ontology, this will be impossible:

10. Thus desire and the desirous one
Cannot be established as simultaneous or not simultaneous.
So, like desire, nothing whatever
Can be established either as simultaneous or as nonsimultaneous.


As always, however, we must remind ourselves of the sense of the conclusion and of its dialectical context. There is no denial here of the possibility of simultaneity, of the existence of desire, or of the possibility of desirous persons. Rather, there is a denial that any of these things make sense in the context of inherent existence.


Chapter VII

Examination of the Conditioned


Having begun the text with an examination of the relation of dependency between phenomena, and having then conducted an analysis of the fundamental ontological constituents of reality, Ngrjuna now brings these two analyses together in a long chapter investigating the nature of the world of conditioned things as a whole. The target position is the view that dependent arising itself, as well as dependently arisen things, are either inherently existent or completely nonexistent. There are really two positions here with which Ngrjuna must contend: First, the reificationist opponent charges that even if we grant Ngrjuna’s earlier arguments for the conclusion that phenomena themselves are empty because they are dependently arisen, dependent arising itself must inherently exist. For only if phenomena are truly dependently arisen, one might argue, are they truly empty. Second, Ngrjuna must answer the following objection: If dependent arising is empty, then arising, stasis, and cessation are nonexistent. Hence there are, in fact, no phenomena since phenomena are defined—particularly in a Buddhist context—as those things that arise, remain, and cease. But clearly there are actual empirical phenomena; indeed, such phenomena must exist for Ngrjuna’s claim that they are empty to make any sense at all. How can this be reconciled with the emptiness of dependent arising?

1. If arising were produced52
Then it would also have the three characteristics.
If arising is not produced,
How could the characteristics of the produced exist?


The three characteristics in question are arising, stasis, and cessation. On a standard Buddhist view, all phenomena come into being in dependence upon conditions, remain in existence dependent upon conditions, and cease to exist dependent upon conditions. This is the core of the two central doctrines of dependent arising and impermanence. Ngrjuna here poses a problem: If dependent arising itself were produced by conditions, then it itself would have these three characteristics and, apparently paradoxically, be impermanent. This is prima facie paradoxical just because if dependent arising is impermanent, it would appear that sometimes thing���s don’t arise dependently, which contradicts the thesis that all phenomena are dependently arisen. Moreover, as Ngrjuna will argue below, this assertion threatens a vicious regress—if arising arises, there must already be arising in virtue of which it does so.

But, Ngrjuna asks in the third and fourth lines, if dependent arising is not produced, where did it come from? If one were to say that dependent arising were not produced and, hence, that it does not depend for its existence on anything else, this would appear to contradict the thesis that everything arises dependently. Dependent arising itself would then be the counterexample to the thesis.

2. If the three, arising, etc., are separate,
They cannot function as the characteristics of the produced.
But how could they be joined
In one thing simultaneously?


These three characteristics, if they characterize the phenomenon of dependent arising itself, must either be present separately or together. This furnishes the basis of a destructive dilemma. If they are separate, then some parts of dependent arising have one of the three; some another. Some are arising; some abiding; some ceasing. But this is problematic since all phenomena are said to arise, to abide, and to cease.53 So it would seem to be the case that if dependent arising itself has all three of these characteristics, it cannot have them separately, but must have them jointly and simultaneously.

But the three characteristics could not be present simultaneously since they are mutually contradictory. At any one point, dependent arising could have only one of them. The same thing cannot be—in the same sense, at the same time—arising and ceasing when these are understood in the sense at issue here, that introduced by the substantialist opponent. It is important in order to understand this argument to keep the dialectical context firmly in mind. The opponent throughout the text, whether on the nihilist side or on the reificationist side, considers existence to be inherent existence and predication to be the ascription of really existent properties to substantial bases. For the opponent Ngrjuna has in mind here, dependent arising—if it is the nature of things at all—must inherently exist. It must therefore have the three characteristics inherently. To have a characteristic inherently is to have it essentially. But then dependent arising, for the opponent, would have a contradictory set of essential properties.

3. If arising, abiding, and ceasing
Have characteristics other than those of the produced,
There would be an infinite regress.
If they don’t, they would not be produced.


The other possibility is that dependent arising has��� some other characteristics—that is, characteristics other than those that all phenomena have in virtue of being dependently arisen. But we could then ask about the characteristics of those characteristics. Do those characteristics arise, abide, or perish? If so, the original regress has not been stopped. Another possibility is that arising, abiding, and perishing do not have characteristics at all. But if not, then they are not phenomena in any ordinary sense at all. While that would cut off the regress, it would do so without achieving any explanation, or any analysis of the kind originally sought, and would leave an uncomfortable paradox: We started seeking an understanding of dependent arising as inherently existent. But its inherent existence requires the inherent existence of arising, cessation, and stasis, all of which now come out to be ontologically sui generis. The further paradox is this: For dependent arising to exist inherently, these three should turn out to be essential properties of all phenomena. But on the alternative under consideration, they are not properties at all.

We might, of course, try to extend this horn of the dilemma by suggesting that although arising, abiding, and ceasing are not phenomena in the ordinary sense, they are characteristics of some special kind. We then seem to have a more curious regress; new ad hoc characteristics arise at each level of analysis. The regress here is an interesting one because its viciousness consists not in the same basis being required for each putatively basic posit, but in there being no principle available to determine a basis for any putative basic posit despite a principle that urges that there must be one. The point that Ngrjuna is after, of course, is that this principle itself—that there must be an explanatory basis, an independent entity that has characteristics, as an explanation of the occurrence of any characteristic—is what generates the regress and must be rejected.

There is, of course, a third alternative. These three might neither have characteristics different from those possessed by ordinary phenomena nor have no characteristics at all: They might indeed have the very trio of characteristics that all ordinary phenomena have, namely, arising, abiding, and ceasing. It is this alternative that occupies Ngrjuna for the remainder of the chapter. This alternative is interesting dialectically in that, on the one hand, it represents the most natural way to approach an analysis of dependent arising, namely, by consistently predicating it of everything, hence suggesting that it is indeed a candidate for an essence of things. On the other hand, as we shall see, that very move precludes treating it as a genuine essence since essences turn out to lack precisely the properties that we must universalize here.

4. The arising of arising only gives rise
To the basic arising.
The arising of the basic arising
Gives rise to arising.



 
This is the opponent speaking. He suggests that dependent arising arises from a more basic arising. This basic arising comes to be, but not on the basis of anything else. The idea, defended by some earlier Buddhist schools, is this: There are two levels of dependent arising. The more superficial is the relationship of mutual dependence of all phenomena, issuing in their impermanence. But this interdependence, on this view, is itself dependently arisen. It depends on a basic arising—a mere fact of interdependent origination, which gives rise to the more specific empirical relations we see. So in the first two lines of this verse, the opponent says that when arising itself is considered in isolation, all that we have is the basic arising. In the third and fourth lines, the opponent says that when that arising has arisen, it gives rise to the more superficial ordinary dependent arising. It is, then, that basic arising that is posited as ontologically foundational.

5. If, as you say, the arising of arising
Gives rise to the basic arising,
How, according to you, does this,
Not arisen from the basic arising, give rise to that?


But Ngrjuna makes the obvious move in reply: Does the basic arising arise from a more basic arising, or is it somehow unarisen (eternal or inexplicable)? If the former, then we seem to have an infinite regress; if the latter, a petitio principii. Ngrjuna makes some of the numerous difficulties that afflict this view explicit in the next two verses:

6. If, as you say, that which is arisen from basic arising
Gives rise to the basis,
How does that nonarisen basis
Give rise to it?


The account is either circular or regressive. If the basic arising is held to arise in dependence on other dependently originated phenomena, and dependent arising is explained as dependent upon the basic arising, then the basis is posited as dependent upon that which it explains, and we have a vicious circle. If on the other hand the phenomena on which the basis depends are other than those it explains, and the phenomena themselves depend upon yet another basis, we have a vicious regress.

In the next verse, Ngrjuna points out the question-begging alternative reading of the enterprise. He notes that one may explain that dependent arising arises through basic arising without circles or regresses, but only by positing the basis as itself nonarisen. This, of course, flies in the face of the demand that motivates positing it in the first place—namely, the demand that every phenomenon, including dependent arising, be explained by some ontologically more fundamental phenomenon:

7. If this nonarisen
Could give rise to that,
Then, as you wish,
It will give rise to that which is arising.


The opponent now suggests another reply. Using the ���analogy of a lamp that illuminates both itself and others, he argues that arising can give rise to itself and to others. This would, from the standpoint of the reificationist, have the happy consequence that while other phenomena would be dependent on dependent arising, dependent arising would be independent and nonempty:

8. Just as a butterlamp
Illuminates itself as well as others,
So arising gives rise to itself
And to other arisen things.


Ngrjuna now launches a lengthy critique of the example, arguing that the relation between the butterlamp and what it illuminates is not one that supports a notion of an inherently existent basis on which things that are not inherently existent can depend:

9. In the butterlamp and its place,
There is no darkness.
What then does the butterlamp illuminate?
For illumination is the clearing of darkness.


Here Ngrjuna is emphasizing a disanalogy between the relation between the butterlamp and what it illuminates, and the putative relation between dependent arising and what it depends upon. The opponent who wields the example does so in order to demonstrate a difference in status between dependent arising and the dependently arisen. Dependent arising is meant not to be dependently arisen, despite the fact that all dependently arisen phenomena are. So the appropriate analogy in the case of the lamp would map this difference in status between being dependently arisen and being independent onto the difference between being illuminated and not being illuminated. The problem, though, is that in the example there is nothing that is not illuminated: Everything in the neighborhood of the lamp is illuminated just as is the lamp.

It was standard philosophical fare in the Buddhist tradition within which Ngrjuna was working to see darkness as a positive phenomenon. So to the extent that one adopted a reified ontology, darkness would be reified as easily as light. The attack on the butterlamp analogy can thus effectively exploit the difficulties Ngrjuna has already developed for theories that require inherently existent things to be related to one another. But it is important to see that even if one is not disposed to reify darkness, and regards it as the mere absence of light, to the extent that one reifies light, Ngrjuna can argue that one will be compelled to reify darkness as well. For if light exists inherently, then wherever light is not present it is essentially not present. And the essential nonpresence of light is essential darkness.

10. If the arising butterlamp
Does not reach darkness,
How could that arising butterlamp
Have cleared the darkness?


Moreover, argues Ngrjuna, the example itself does not bear close scrutiny as a case of an entity with some inherent power giving rise to a set of effects that depend upon it. For the task of the butterlamp is the clearing of darkness—or the production of illumination. Now the production of light and the clearing of darkness are, Ngrjuna claims, equivalent. So, if the butterlamp illuminates objects by its light reaching them, it should clear darkness by means of its light reaching darkness. But that would be for light and darkness to be present in the same place, which is contradictory.

11. If the illumination of darkness occurs
Without the butterlamp reaching darkness,
All of the darkness in the world
Should be illuminated.


If it is not necessary, on the other hand, for the light of the butterlamp to reach darkness in order to dispel it, since there is a lot of darkness in the world not reached by any single butterlamp, that butterlamp should be capable of dispelling all of that darkness.

12. If, when it is illuminated,
The butterlamp illuminates itself and others,
Darkness should, without a doubt,
Conceal itself and others.


Finally, Ngrjuna argues, if we are seriously to maintain that the butterlamp illuminates itself and others through a luminous essence, then since the essence of darkness is to conceal things, and things with such essences affect themselves and others, we should expect darkness to be self-concealing. But then we would not see darkness.

The point of all of this is not that we can’t see lamps when they are lit or that we can when they aren’t. Rather it is that the mechanism by which we see what we see when a lamp is lit is the same whether we are seeing the lamp or other things. To put it in contemporary terms, photons reach our eyes from the lamp or from its flame in the same way they do from the other physical objects in the neighborhood. And just as the visibility of the things in the neighborhood is dependent on a host of conditions, so is the visibility of the lamp. So we do not have even an analogy to a case where the status of dependent arising would be distinct from that of the dependently arisen.

13. How could this arising, being nonarisen,
Give rise to itself?
And if it is arisen from another,
Having arisen, what is the need for another arising?


Here Ngrjuna is bringing us back to the original argument and reminding us of the reificationist’s uncomfortable choice between a vicious regress and a begged question. If every arisen thing depends on an ontologically prior arising, we have an infinite regress. For each arising will require such a foundation. But if we cut off the regress by presupposing at some level a nonarisen dependent arising, we have to ask why that level is exempt from the need for explanation. Ngrjuna now announces the conclusion he will defend in the next section of the chapter:

14. The arisen, the non-arisen and that which is arising
Do not arise in any way at all.
Thus they should be understood
Just like the gone, the not-gone, and the going.


Recall the analysis of motion: Ngrjuna argued that no entity answering to “motion” could be found in an entity that was in motion in the past, nor in an entity yet to move, nor in a currently moving entity. Motion had to be understood relationally and not as an entity. Using similar reasoning, Ngrjuna will now argue that arising cannot be found as an entity in something not yet arisen, nor in something that has already arisen, nor in something yet to arise. Arising will also fail to be an entity and will have to be understood relationally. This will provide the key both to the refutation of the position that underlies both extreme positions—that for arising to exist, it must exist inherently—and to the construction of a coherent positive account of dependent arising. The next three verses begin a sketch of dependent arising as empty, connecting this fact with the emptiness of dependently arisen phenomena:

15. When there is arising but not yet
That which is arising,
How can we say that that which is arising
Depends on this arising?


Ngrjuna here suggests that the way the reificationist has gone about posing the philosophical problem about the status of dependent arising itself is all wrong. The initial presumption at the basis of this debate is that arisen entities arise from an independently existing process of dependent arising. But this is wrongheaded in at least two ways: First, phenomena arise from other phenomena, not from a���rising. So, for instance, if I strike a match, the fire emerges from the friction, the sulphur, the oxygen, my desire for light, and so forth, but not from dependent arising itself. That is a fact at a different level of analysis, which itself comprises the network of relationships just indicated. Second, if the existence of the process of arising antedates the existence of the arisen, it cannot be a sufficient condition or a complete explanation of the arisen. For if it were, the arisen would then exist. That being so, Ngrjuna asks, “Why posit dependent arising itself as a phenomenon within the framework of dependent arising?”

16. Whatever is dependently arisen,
Such a thing is essentially peaceful.
Therefore that which is arising and arising itself
Are themselves peaceful.


The sense of “peaceful” (zhi-ba) here is important. Ngrjuna is asserting that things are not, from the ultimate point of view, in the constant flux of arising, remaining, and decaying that characterizes them from the conventional point of view. This will be the conclusion of the extended argument that follows and is here merely announced in advance. But it is important at this stage to be clear about just what Ngrjuna is asserting for it is indeed a delicate point: It is true that ordinarily and prereflectively, and sometimes as the result of bad philosophy, we tend to think of things as permanent and as having fixed essential natures. But a careful reflection on the nature of conventional phenomena shows them on analysis to be impermanent and, hence, to be characterized by the three properties of arising, stasis, and cessation.54

But while this takes us to a deeper understanding of the nature of phenomena, it does not take us all the way. For phenomena, having no essence, cannot have even these properties essentially. One way of seeing that is this: If we take the import of the threefold nature of phenomena seriously, we see that the phenomena are themselves literally momentary. And if they are momentary, then there is literally no time for them to arise, to endure, or to decay. So from an ultimate point of view, the point of view from which they have no existence as extended phenomena at all, they do not possess these three properties. Hence no single real entity is in flux. In this sense they are peaceful. Ngrjuna points out the other way of seeing phenomena in the next verse: It does not follow from the fact that there are no inherently existent arisen entities that there are non-arisen ones. All phenomena are arisen, but they arise as empty, and as dependent. Coming to be just is arising, and all arising is dependent arising.

Ngrjuna now turns his attention to an analysis of the three characteristics of arising, stasis, and cessation, showing of each in turn that it cannot be understood as��� ontologically independent. He begins with arising:

17. If a nonarisen entity
Anywhere exists,
That entity would have to arise.
But if it were nonexistent, what could arise?


We can exclude nonarisen entities from the analysis since the only sense that we can make of the existence of any phenomenon is in terms of its having arisen. Arising is hence a ubiquitous characteristic of phenomena. This, of course, is part of what motivates treating it, as well as stasis and cessation, as inherently existent.

18. If this arising
Gave rise to that which is arising,
By means of what arising
Does that arising arise?


If we take arisen things to require ontological grounds, then ground them not in other arisen things (since that would generate an obvious regress within the phenomenal world), but in dependent arising itself, there remains the infinite regress to which Ngrjuna alluded earlier. Assuming dependent arising is to be the ground, then if grounds are needed, it too needs a ground. Ngrjuna makes this explicit in the following verse:

19. If another arising gives rise to this one,
There would be an infinite regress.
If something nonarisen is arisen,
Then all things could arise in this way.


The last two lines of this verse emphasize that the regress cannot ever be cut off by positing some nonarisen arising. That would, as Ngrjuna argued above, patently beg the question.

20. Neither an existent nor a nonexistent
Can be properly said to arise.
As it is taught before with
“For neither an existent nor a nonexistent.”


The reference of the last line is to I: 6:


For neither an existent nor a nonexistent thing
Is a condition appropriate.
If a thing is nonexistent, how could it have a condition?
If a thing is already existent, what would a condition do?


The implicit argument is, then, that inherently existent phenomena cannot be said to arise since they would exist eternally and independently; nonexistent phenomena cannot be said to arise since if they did, they would exist. Arising can hence only be a property of noninherently, but conventionally, existent phenomena. But it then follows that arising as a property can only be a noninherently existent, conventional property.

Ngrjuna now turns his attention to the properties of cessation and endurance. He begins, though, with a final remark on arising as a transition, concerning the relation between arising and cessation. This next verse must be read along with VII: 23 and 26. Together they constitute an exhaustive discussion of the possible inherence of the three properties under discussion in ceasing entities:

21. The arising of a ceasing thing
Is not tenable.
But to say that it is not ceasing
Is not tenable for anything.


The first alternative Ngrjuna considers is that a ceasing thing is arising. But if a thing is already ceasing, it is therefore no longer arising. And since all phenomena are, when their impermanence is taken into consideration, ceasing, it would follow that nothing can be said to be arising.

22. A static existent does not endure.55
A nonstatic existent does not endure.
Stasis does not endure.
What nonarisen can endure?


Ngrjuna now turns to stasis—the moment between arising and ceasing. This verse must be read along with VII: 23, 25, and 27, which together provide a complete examination of the status of stasis. Here he emphasizes that the moment between the arising and ceasing of a momentary phenomenon—an event—has no temporal extent. So a thing that we might conventionally refer to as static literally does not endure with identity through time. But of course neither does something that is not even conventionally static. And finally, since as a consequence of these two premises stasis is not instantiated in any phenomenon, it itself does not endure. So, Ngrjuna concludes, stasis fails to exist over time in any sense and so is no candidate for an inherently existent phenomenon.

23. The endurance of a ceasing entity
Is not tenable.
But to say that it is not ceasing
Is not tenable for anything.


This verse plays a central role in each of two interwoven arguments. In the context of VII: 21 and 26, it provides part of the exhaustive analysis of the impossibility of arising, abiding and ceasing as instantiated in ceasing (hence in impermanent) phenomena. In the context of VII: 22, 25, and 27, it provides part of the analysis of the impossibility of locating endurance in any phenomenon, hence emphasizing the impermanence of all phenomena. Since to exist is to exist in time and things that are ceasing are by definition not in a state of continued existence, ceasing phenomena do not provide the kind of continuity with numerical identity that endurance demands. And all phenomena are, upon analysis, seen to be constantly ceasing. So endurance has no possibility of instantiation, and ceasing phenomena cannot have this property as an essential attribute.

24. Inasmuch as the nature of all things
Is aging and death,
Without aging and death,
What existents can endure?


Moreover, since all things decay, this analysis is perfectly general. Nothing exists in the way that it would have to in order to have endurance as part of its essence.

25. Stasis cannot endure through itself
Or through another stasis.
Just as arising cannot arise from itself
Or from another arising.


This verse recalls the discussion of VII: 13–19 and has an important echo in VII: 32. Ngrjuna argued earlier that we cannot analyze arising either as sui generis or as dependent upon some other arising. In the first case, we beg the question; in the second we invite an infinite regress. He now points out that the same is true of stasis. We can’t, in order to demonstrate the inherent existence of stasis, argue that it endures because of itself. If this kind of reflexive explanation were possible, we would not need to posit stasis in the first place as an explanation of the continued existence of empirical phenomena. Each could count as self-explanatory. But if we say that stasis, like other static things, is static because of its possessing a distinct stasis, we are off on a vicious regress.

26. The ceasing of what has ceased does not happen.
What has not yet ceased does not cease.
Nor does that which is ceasing.
What nonarisen can cease?


Nagarjuna thus completes the tripartite argument for the impossibility of the instantiation of arising, abiding, and ceasing begun in VII: 21 and 23. Cessation, conceived of as an inherently existent, independent property, needs a substratum. We have seen in the previous two verses in this argument that neither arising nor static things can provide this substratum. The only alternative remaining is the ceasing. But these phenomena, passing out of existence, are by definition not inherently existent and so fail as candidates. And again, since all phenomena are ceasing, this means that ceasing as an independent property has no basis. The argument here is an obvious echo of the argument against the inherent existence of motion. So the conclusion to draw is not that there is no cessation or that there are no ceasing phenomena. That would be crazy. Rather, neither cessation nor any impermanent phenomenon can be identified independently as an entity itself. Their existence is purely relational. Nagarjuna now turns to the cessation of the static:

27. The cessation of what is static
Is not tenable.
Nor is the cessation of
Something not static tenable.


Two points are being made here: First, if there were intrinsically real entities that could serve as ontological bases for cessation, they would have to have either remained stable or not. If the former, then in virtue of having the nature of stasis, they would be incapable of cessation. If the latter, since they never really existed, there is nothing to cease. But there is also a second point being made that depends upon the conventional reality of cessation. Since cessation is conventionally real and is incompatible both with inherently existent stasis and with there being no stasis at all, both of these alternatives with respect to stasis are eliminated. Cessation and stasis must be understood relatively and not absolutely. This point is reiterated in the following verse:

28. Being static does not cease
Through being static itself.
Nor does being static cease
Through another instance of being static.


This verse also echoes VII: 25 and that discussion of the impossibility of arising being either self-explanatory or always explained by reference to yet another arising. All things, having remained momentarily in existence, change constantly. This, however, cannot be explained by reference to the nature of stasis, either reflexively or regressively.

29. When the arising of any entity
Is not tenable.
Then the cessation of any entity
Is not tenable.


Since nothing arises inherently, nothing ceases inherently. Since upon careful examination nothing withstands analysis as an inherently existing phenomenon, nothing remains independent of conventional designation to be characterized as arising or ceasing. This is how it goes from the ultimate standpoint. From that standpoint—though achieved by noting the universality of arising and cessation of conventional phenomena—since there are no phenomena, there is no arising and cessation. But by contraposition we get the corelativity and mutual entailment of arising and ceasing at the conventional level.

30. For an existent thing
Cessation is not tenable.
A single thing being an entity and
A nonentity is not tenable.


This verse and the next reinforce the point about the ultimate nonexistence of cessation and, by implication, of arising and stasis. In the preceding, Nagarjuna emphasizes that for an inherently existent entity to cease to exist would be for it to inherently exist and not exist. In the subsequent verse, he points out that it makes no sense for a nonexistent thing to cease to be, just as it makes no sense to behead someone a second time:

31. Moreover, for a nonentity,
Cessation would be untenable.
Just as a second beheading
Cannot be performed.


32. Cessation does not cease by means of itself.
Nor does it cease by means of another.
Just as arising cannot arise from itself
Or from another arising.


This verse has an exact parallel in VII: 25. Again, Nagarjuna recalls the uncomfortable choice between a trivially begged question and a vicious regress presented originally in the context of the discussion of arising and recalled in the discussion of stasis. The argument applies, mutatis mutandis, to cessation. The conclusion of this trio of arguments is that we cannot conceive of any of the three characteristics of dependent arising as self-grounded. All must be understood dependently and hence as empty.

33. Since arising, ceasing, and abiding
Are not established, there are no compounded things.
If all compounded things are unestablished,
How could the uncompounded be established?


That is, arising, abiding, and ceasing are not entities at all—they are mere relations. Since these fundamental attributes of dependently arisen phenomena are empty of inherent existence, what could have inherent existence?

34. Like a dream, like an illusion,
Like a city of Gandharvas,
So have arising, abiding,
And ceasing been explained.


This chapter thus brings the first principal section of Mlamadhyamakakrik to a close, drawing together the threads spun in the earlier chapters to produce a thorough demonstration of the emptiness of the conventional phenomenal world. Having demonstrated the emptiness of conditions and their relations to their effects, change and impermanence, the elements, the aggregates,56 and characteristics and their bases—in short, of all the fundamental Buddhist categories of analysis and explanation—Nagarjuna has now considered the totality they determine—dependent arising itself and the entire dependently arisen phenomenal world—arguing that dependent arising and what is dependently arisen are themselves empty of inherent existence.

This is a deep result. It again presages the doctrine of the emptiness of emptiness that is made explicit in Chapter XXIV, and it develops further the theme explored in Chapter I, namely, that when from the Madhyamika perspective one asserts that a thing is empty or that it is dependently arisen, one is not contrasting their status with the status of some other things that are inherently existent. Nor is one asserting that they are merely dependent on some more fundamental independent thing. Nor is one asserting that instead of having an independent essence things have as their essence dependence or emptiness, either or both of which exist in some other way. Rather, as far as one analyzes, one finds only dependence, relativity, and emptiness, and their dependence, relativity, and emptiness.

But this is not to say either that emptiness, dependent arising or conventional phenomena are nonexistent—that they are hallucinations. Indeed it is to say the opposite. For the upshot of this critical analysis is that existence itself must be reconceived. What is said to be “like a dream, like an illusion” is their existence in the mode in which they are ordinarily perceived/conceived—as inherently existent. Inherent existence simply is an incoherent notion.57 The only sense that “existence” can be given is a conventional, relative sense. And in demonstrating that phenomena have exactly that kind of existence and that dependent arising has exactly that kind of existence, we recover the existence of phenomenal reality in the context of emptiness. In the next major section, comprising Chapters VIII through XIII, Ngrjuna addresses the emptiness of the subject of experience.


Chapter VIII

Examination of the Agent and Action


The discussion of external phenomena comprised by the first seven chapters of the text leads naturally to a discussion of the subject side of experience, a discussion that occupies the next six chapters. For it might be granted that the phenomenal external world is empty, but argued that it depends for its nominal existence on an inherently existing subject. This idealist tactic, familiar in the West through Berkeley and Hume (and criticized by Kant in the refutation of idealism), was adopted by some (the Cittamtra school) in the history of Buddhist philosophy. We can well imagine an opponent at this stage in the dialectic conceding to Ngrjuna that external phenomena lack inherent existence and that the dependent arising that characterizes them lacks inherent existence, but that their very emptiness entails their nominal character and, hence, some subject capable of engaging in nominal imputation. So the subject as agent must exist.

1. This existent agent
Does not perform an existent action.
Nor does some nonexistent agent
Perform some nonexistent action.


Nagarjuna here announces that, with respect to agency and action as well, he will steer a middle course between inherent existence and complete nonexistence. Neither action nor agent will come out to be an inherently existing entity. Nor will either end up being completely nonexistent.

2. An existent entity has no activity.
There would also be action without an agent.
An existent entity has no activity.
There would also be agent without action.


If the agent were inherently existent, then it would be unchanging. Activity is always a kind of change. So if there were action in the context of an inherently existing agent, the action would be agentless, which would be absurd. Moreover, the agent would be inactive, which would also be absurd. This, of course, is just one more case of Nagarjuna demonstrating the incoherence of a position that tries both to posit inherently existent, independent entities and then to get them to interact.

3. If a nonexistent agent
Were to perform a nonexistent action,
Then the action would be without a cause
And the agent would be without a cause.


However, if agent and action are totally nonexistent, there will be no cause for the action and no justification for calling the agent an agent.

4. Without a cause, the effect and
Its cause will not occur.
blockquote width="0pt" align="justify">Without this, activity and
Agent and action are not possible.


Agent, the agent’s activity, and the action all depend upon conditions. They are all, therefore, dependently arisen and empty. If, as the opponent would have it, these are inherently existent, there would be no action. But if we think of them as dependent, we can make perfectly good sense of agent, activity and action in interrelation.

5. If activity, etc., are not possible,
Entities and nonentities are not possible.
If there are neither entities nor nonentities,
Effects cannot arise from them.


If there were no action, then since entities arise from the action of previous events, there would be no entities and no effects. In short, without making sense of the possibility of actions and agency as empty, we can’t account for the existence of any phenomena.

6. If there are no effects, liberation and
Paths to higher realms will not exist.
So all of activity
Would be without purpose.


And all of this has a moral and a soteriological dimension as well. For if there are no acts and no effects, then the practice of morality and of the Buddhist path will make no sense. There would be no point to life if human action is impossible. And again, its impossibility follows straightforwardly from the reification of either agent or action. It is ironic that it is the urge to guarantee more reality and significance for ourselves than emptiness appears to allow that leads to a view of life as perfectly impossible and pointless. That is, though we are led to ascribe inherent, independent existence to ourselves and to the world of phenomena we cherish—in part, in order to assign them the greatest possible importance—this very importance would be completely undermined by such inherent existence and independence. For in that case, all activity and all consequences of activity would be impossible. The resultant life would be static, detached, and utterly meaningless. Only in the context of emptiness—what might appear to be the greatest threat to meaningfulness—can a meaningful life be understood.

7. An existent and nonexistent agent
Does not perform an existent and nonexistent action.
Existence and nonexistence cannot pertain to the same thing.
For how could they exist together?


There is no way to escape from this dilemma by trying to have it both ways: The agent cannot be existent as an actor, but nonexistent as one who undergoes the action. Nor can the action be existent as an entity, but nonexistent as dependent upon the agent.

8. An actual agent
Does not perform a nonactual action.
Nor by a nonactual one is an actual one performed.
From this, all of those errors would follow.


Nor is it coherent to suppose that the agent is existent, but the action nonexistent. For then there would be no reason to call the agent an agent. An agent, after all, is someone who performs an action. The next two verses put this point and those made in the opening verses together:

9. An existent agent
Does not perform an action that
Is unreal or both real and unreal
As we have already agreed.


10. A nonexistent agent
Does not perform an action that
Is unreal or both real and unreal
As we have already agreed.


11. An existent and nonexistent agent
does not perform an action that
Is unreal or both real and unreal
As we have agreed.


Ngrjuna now moves to assert his positive position on this matter: Agent and action are interdependent. Neither is logically or ontologically prior to or independent of the other. What it is to be an agent is to be performing an action. What it is to be an action is to be the action of an agent:

12. Action depends upon the agent.
The agent itself depends on action.
One cannot see any way
To establish them differently.


13. From this elimination of agent and action,
One should elucidate appropriation in the same way.
Through action and agent
All remaining things should be understood.


By “appropriation,” Nagarjuna indicates any cognitive act by means of which one takes an attribute or entity as one’s own, or as part of one’s self. That includes the grasping of the aggregates as the self or of one’s mental states as part of one’s identity or of one’s possessions as central to one’s being. Appropriation in this broad sense is, hence, a central object of concern for Buddhist philosophy and psychology, and the relation between the appropriator and the act of appropriation is an important object of analysis. For in many ways the self that is constructed through appropriation presents itself as the subject of appropriation. But it is merely constructed, and its substantial reality is illusory.

Then what indeed does the appropriation? And where there is no appropriator, how does appropriation occur? Nagarjuna here suggests that this account of the relation between agent and action provides a model for understanding that relation. That is, this analysis provides a perfect paradigm for understanding the nature of subjectivity. In all cases of the relation between an agent of any kind and an act of any kind, the identity of the two will be seen to be mutually dependent, and each will come out as conventionally real, though not as inherently existent. We will see this paradigm articulated over the next five chapters as Nagarjuna argues that we cannot make any sense of the self as an entity independent of its actions, perceptions, and interactions. Nor can we make any sense of the ontology of these phenomena as independent of the subject. This is a natural extension of the analysis of emptiness of the external world and demonstrates Nagarjuna determination to treat all phenomena on the same basis.


Chapter IX

Examination of the Prior Entity


Now one can surely imagine an opponent responding to the argument of the previous chapter by granting that agency and its corelative phenomena might be empty, yet still denying that awareness itself—the subjectivity that grounds perception—could be empty. For, one might argue, the emptiness of all phenomena still requires that there be a subject for whom they are phenomena. Nagarjuna articulates this response in the opening verses of this chapter:

1. Since sight and hearing, etc., and
Feeling, etc., exist,
He who has and uses them
Must exist prior to those, some say.


2. If there were no existent thing,
How could seeing, etc., arise?
It follows from this that prior to this,
there is an existent thing.


That is, without a subject of experience, there can be no experience and no experienced objects. This argument has familiar instances in Descartes and Kant. But Nagarjuna, siding with Hume on this issue, begins by asking how this entity could be an object of knowledge:

3. How is an entity existing prior to
Seeing, hearing, etc., and
The felt, etc.,
Itself known?


So first, Nagarjuna points out, we have no direct evidence for the existence of such an entity because evidence of it would require that it could be an object, but is supposed by its proponent to be purely subjective. Moreover, Nagarjuna points out, it is supposed to be independent of and ontologically prior to perception and the perceived. So:

4. If it can abide
Without the seen, etc.,
Then, without a doubt,
They can abide without it.


That is, independence is a two-way street. If the self is independent of its perceiving and perception, then its perceiving and perception are independent of it. Now there is one reading of this claim on which it is straightforwardly and foolishly fallacious. Nagarjuna is not arguing that all relations are symmetric. It does not follow from the fact that this book is on your table that your table is on the book, and Nagarjuna is not foolish enough to think that it does. The point is, rather, once again the Humean one that whatever is indeed logically independent is separable. The opponent wants to argue that the self is logically independent of its perceptions and their contents. But if so, then they are separable, and we can imagine not only a nonperceiving subject, but also unperceived perceptions. Just as we can imagine a clear table and a book not on a table. But, Nagarjuna suggests, the idea of unperceived perceptions is both absurd on its face and contradictory to the opponent’s theoretical framework.

5. Someone is disclosed by something.
Something is disclosed by someone.58
Without something how can someone exist?
Without someone how can something exist?


Nagarjuna here emphasizes the corelativity and interdependence of subject and object.59 Subjectivity only emerges when there is an object of awareness. Pure subjectivity is a contradiction in adjecto. Moreover, the idea of an object with no subject is contradictory. The very concept of being an object is that of being the object of a subject. The affinities to Kant and Schopenhauer here are quite strong, but should not be pushed too far. Nagarjuna would clearly have no truck with the substantialist flavor of their analysis of the subject and object.

6. While prior to all of seeing, etc.,
That prior entity doesn’t exist,
Through seeing, etc., by another one,
That other one becomes disclosed.


An opponent might at this point argue that although there is no continuous prior entity that endures through time and stands behind all perception, we must posit an entity as the basis of each individual perceptual episode. The self on this model would be a succession of momentary but inherently existent subjects of moments of experience. But, Nagarjuna argues in the next verse, the same argument against positing a single prior entity can be mobilized against each punctal prior entity:

7. If prior to all of seeing, etc.,
No prior entity exists,
How could an entity prior
To each seeing exist?


That is, given that there is no need to identify an independent self as the basis of all seeing, there is no need to establish one as a basis for each one independently. The same arguments for the relativity and relational character of perception apply, mutatis mutandis, for each perceptual episode. Moreover, even if we did posit such entities, they would get us nowhere toward positing the self that the reifier of self really cares about—a continuous self with which we can really identity and whose fate we can care about.

8. If the seer itself is the hearer itself,
And the feeler itself, at different times,
Prior to each of these he would have to arise.
But this makes no sense.


Moreover, since this proposal is for a distinct prior entity for each perceptual episode, we would need distinct subjects for, for example, hearing and seeing. But as we can do these things at the same time, it would follow that there are multiple simultaneous selves. The unity of experience that is the putative explanandum and motivation for positing this entity in the first place (emphasized in the first two lines) would dissolve. Nagarjuna emphasizes this conclusion at IX: 9:

9. If the seer itself is distinct,
The hearer is distinct and the feeler is distinct,
Then when there is a seer there would also be a hearer,
And there would have to be many selves.


10. Seeing and hearing, etc.,
And feeling, etc.,
And that from which these are arisen:
There is no existent there.


However, one should not be tempted to try to ground perception, the perceived object, and the perceiver in some more fundamental ontological ground—some intrinsically identical basis for their existence. For the need to develop a substantial foundation for these phenomena should vanish once one sees that not only do they have no ultimate ontic status, but that they need none. They, like all phenomena, emerge relationally and dependently.

11. Seeing and hearing, etc.,
And feeling, etc.,
If that to which they belong does not exist,
They themselves do not exist.


Not only has this analysis refuted the inherent existence of the self as a basis for experience, but in virtue of so doing, it has refuted the inherent existence of perception and the perceptual faculties.

12. For whomever prior to,
Simultaneous with, or after seeing, etc., there is nothing,
For such a one, assertions like “it exists” or “it does not exist”—
Such conceptions will cease.


Nagarjuna here generalizes the point and offers a diagnosis of the confusion he has worked to resolve: Just as we want to say that the self as pure subject does not exist—nor do perception or perceptual objects exist as entities—yet want to affirm the conventional reality of perception, perceivers, and perceiveds, in general, we want to deny the inherent existence of phenomena and affirm their conventional reality. Just as we want to say that the self neither exists inherently nor that it is nonexistent inherently, we want to refrain from attributing inherent existence or inherent nonexistence to all entities. The apparent paradox involved in saying that things both exist and do not exist in one breath and saying that they neither exist nor do not exist in another—indeed of refusing in another sense to permit even these predications in another mood—arises, Nagarjuna points out, from the conceptual imputation of inherently existent bases for these predications, which then h�Àave to be thought of as having contradictory properties. Absent the bases, we can see these assertions merely as useful analytical tools in various dialectical contexts to help us to see the ultimately empty and conventionally real nature of phenomena. And Ngrjuna concludes this chapter by asserting that once one ceases hypostasizing the subjective self—that entity that might seem to be, as Descartes notes, the most obviously existent and most easily known entity of all—the temptation to hypostasize other entities dissolves.


Chapter X

Examination of Fire and Fuel


This chapter, the only one in this set of chapters ostensibly addressing an external phenomenon, is in fact concerned entirely with a standard counterexample to the kind of arguments Nagarjuna offered in the two previous chapters on subjectivity in action and in perception. Recall that in those discussions Nagarjuna argues that subject and object cannot be intrinsically and distinctly identified as entities because of their mutual dependence. Buddhist schools asserting substantial identity in the context of dependent co origination, such as Vaibhśika and Sautrantika schools, used the example of fire and fuel to demonstrate the compossibility of substantial independent identity and dependent origination, as well as the possibility of the one-way dependence relation that these schools assert that actions and perception bear to the self. Just as fire depends on fuel but not vice versa, they would argue, and just as fire and fuel have distinct identities despite the fact that the former depends for its existence on the latter, action and perception can depend on the subject but not vice versa. Despite this dependence, proponents of this view would argue each relatum can be individually established as an entity.60 In this chapter, Nagarjuna undertakes the task of demonstrating that the example does not demonstrate these possibilities.

1. If fuel were fire
Then agent and action would be one.
If fire were different from fuel,
Then it could arise without fuel.61


The opponent does not want to assert the identity of fire and fuel, first, since it would contradict common sense, but second, since that, by the intended analogy, would identify agent and action, self and perception. On the other hand, if they are identified as intrinsically different—as having distinct and independent essential identities—they should be able to arise independently. Fuel should count as fuel even if there were no fire; fire should be possible without fuel. This follows from drawing the distinction at the level of intrinsic identity. Of course, distinguishing them conventional ƀy permits their mutual dependence, but fails to establish the intrinsic identity intended by the verificationist.

2. It would be forever aflame;
Flames could be ignited without a cause.
Its beginning would be meaningless.
In that case, it would be without any action.


The second and third verses spell out the consequences of attributing inherent existence to fire: It would be independent of all conditions, including its fuel; it would burn causelessly, since there would be no condition under which it would not burn. So all fire would, in that case, be eternal. Moreover, it would not consume anything, having no connection to the presence or absence of fuel. Moreover, Nagarjuna asserts in the final two lines of X: 3, the activity of starting a fire would be nonsensical:

3. Since it would not depend on another
Ignition would be without a cause.
If it were eternally in flames,
Starting it would be meaningless.


4. So, if one thinks that
That which is burning is the fuel,
If it is just this,
How is this fuel being burned?


Nagarjuna now sets up a destructive dilemma: Either the process of burning is identical to the fuel or different. In X: 4, he considers the possibility that they are identical. If so, he suggests, we have a problem in explaining how the fuel is consumed. The ordinary explanation of that is the presence of fire. But by identifying the burning process with the fuel, we have left the fire out of the picture. This analysis hence provides no explanation of combustion. After all, fuel by itself does not burn. It must be ignited, that is, fire must be introduced. If, as Nagarjuna argues in X: 5, they are completely different, there won’t be any fire at all. For then the burning would be dissociated from and independent of the fuel, and the unburned fuel would not be consumed by the burning. We could make no sense of the transition from unburned to burned fuel. The general moral is that we cannot make sense of interactive processes such as combustion without attending to the mutual dependence of the interacting phenomena that constitute those processes:

5. If they are different, and if one not yet connected isn’t connected,
The not yet burned will not be burned.
They will not cease. If they do not cease
Then it will persist with its own characteristic.
6. Just as a man and a woman
Connect to one another as man and woman,
So if fire were different from fuel,
Fire and fuel would have to be fit for connection.


Here the opponent suggests that just as males and females are suited to connect in special ways in virtue of their particular anatomical structures, despite existing independently of one another, fire and fuel may be similarly suited to some special kind of connection. In that case, we would have the bizarre picture of fire being independent of fuel, yet peculiarly suited to coming together with it, and vice versa.62 Moreover, since on this model fire and fuel are distinct from one another in nature, yet interactive (they “preclude” each other in the sense that causes and effects preclude one another—that is, in virtue of being connected yet incapable of simultaneous copresence), there must still be some account of how they connect, an account by no means easy to envisage:

7. And, if fire and fuel
Preclude each other
Then fire being different from fuel,
It must still be asserted that they connect.


Fire and fuel hence appear to be mutually dependent. Indeed the central point of Nagarjuna argument is that they are. But here the question arises: Don’t they then have either to depend upon some third more fundamental thing or to be asymmetrically dependent, one of them established independently of the other?

8. If fire depends on fuel,
And fuel depends on fire,
On what are fire and fuel established as dependent?
Which one is established first?


If either is established as an entity first, without any reliance on the existence or nature of the other, that member of the pair would have a claim to being the basis in an asymmetrical dependency relation, and the opponent would have the counterexample necessary to refute the analysis in Chapters VIII and IX. The most obvious form that such an asymmetric dependence could take would involve the dependence of fire on fuel. Nagarjuna argues that this is impossible to maintain:

9. If fire depends on fuel,
It would be the establishment of an established fire.
And the fuel could be fuel
Without any fire.


There are two arguments here. In the first two lines, Nagarjuna argues that if fire were to depend upon fuel, fire would be doubly established. The point is that in order for the fuel to count as fuel, the existence of the fire must have already been established; indeed, the fuel depends upon the fire for its character as fuel. So to say then that the fire is dependent upon the fuel would be to argue that something whose existence is already presupposed if the fuel is to exist depends for its existence on that fuel. Note that this is only problematic for the opponent. That is, for one who accepts, as Nagarjuna does, the mutual interdependence of phenomena, it is in fact true that fire depends upon fuel and that fuel depends upon fire. But the opponent at this stage in the argument argues that fire exists only dependently, but dependently on independent fuel. So Nagarjuna only needs to show that position to be untenable. And the problem for the opponent is simply that the fuel he wants to exist independently can only do so in the presence of fire, which itself is merely dependent.

Second, Nagarjuna argues, this would entail the absurd independent establishment of fuel as fuel. For fuel to be established independently as fuel in the absence of fire would be for there to be some characteristic of fuel that could be specified independently of fire that makes it fuel. But there is none. What makes fuel fuel is that it is combustible.

10. If that on which an entity depends
Is established on the basis
Of the entity depending on it,
What is established in dependence on what?


So in order to establish the existence of fuel as fuel, we must establish the existence of fire. In order for something to be fire, it must be consuming fuel. Neither depends asymmetrically on the other.

11. What entity is established through dependence?
If it is not established, then how could it depend?
However, if it is established merely through dependence,
That dependence makes no sense.


Now Nagarjuna draws the general ontological moral from this discussion of the putative counterexample. If an entity is inherently existent, it must be independently established as an entity and with its own nature. So no entity could be established as inherently existent through dependence on any other entity. Onl�ˀy inherently existent entities could be independent. To establish something as inherently existent through its dependence on something else is incoherent. So since entities can be established neither through independence nor through dependence, there is no way to establish anything as an entity in its own right.

12. Fire is not dependent upon fuel.
Fire is not independent of fuel.
Fuel is not dependent upon fire.
Fuel is not independent of fire.


That is, neither fuel nor fire can be established as independent bases of predication separate from one another that then stand in accidental relations to one another. There are not two entities, fire and fuel, which then are related either by dependence or interdependence.

13. Fire does not come from something else,
Nor is fire in fuel itself.
Moreover, fire and the rest are just like
The moved, the not-moved, and the goer.


Though, as verse 12 grants, fire exists only in relation to fuel, it would not be correct to assert that fuel as an independent entity somehow produces fire. The analysis and the conclusion are strictly analogous to that regarding motion and the mover. We neither can say that motion is the same as the mover nor that they are different entities. We cannot say that motion is present in the unmoved, the moving, or the yet-to-move. Similarly we cannot say that fire is the same as the fuel nor that it is different. Nor can we say that it is present in the unburned, the burning, or the yet-to-be-burned fuel. The next verse emphasizes this point:

14. Fuel is not fire.
Fire does not arise from anything different from fuel.
Fire does not possess fuel.
Fuel is not in fire, nor vice versa.


15. Through discussion of fire and fuel,
The self and the aggregates, the pot and cloth
All together,
Without remainder have been explained.


The fire and fuel example is used as an analogy for a number of different cases of relations between bases and their attributes, including the relation between the putative self and its aggregates—that is, the components of the personality. But there are other stock examples—the relation between the pot and its properties and between the cloth and its thread—that are used to try to defend these asymmetrical dependence relations between inherently existent bases and the properties they support. Nagarjuna is simply asserting the �ˀcomplete generality of this argument: It applies, mutatis mutandis, to all of these cases.

16. I do not think that
Those who teach that the self
Is the same as or different from the entities
Understand the meaning of the doctrine.


This colophon verse reminds us that when existence is understood in terms of emptiness and when entities are regarded as purely relational in character, identity and difference can only be understood conventionally. This applies not only with respect to apparently distinct entities, but also to the relation between parts and wholes, things and their attributes, events and their causes, and as Nagarjuna emphasizes here, self and the objects of awareness. Strict identity and difference as determined by reference to phenomena themselves are only conceivable from the incoherent standpoint of inherent existence.


Chapter XI

Examination of the Initial and Final Limits


But suppose that one could see that the self, considered as agent or as subject, lacks inherent existence, and still one argued that nonetheless it must do so in virtue of its impermanence and being subject to change. Then, one might argue, birth, aging, and death must be real as the conditions of the self’s unreality. This is the position with which Nagarjuna concerns himself in this chapter. But he is also concerned with the generalization of this question to the birth, aging, and death of all of cyclic existence.63 And it is this more general problem with which he actually opens the chapter, developing the account of individual impermanence as a special case:64

1. When asked about the beginning,
The Great Sage said that nothing is known of it.65
Cyclic existence is without end and beginning.
So there is no beginning or end.


The question about the existence and nature of the origin of the world is one of the questions that Sakyamuni Buddha declared to be unanswerable. Nagarjuna here interprets that to mean that there is nothing coherent that can be said about the origin of the world. Given the striking similarity between the questions that the Buddha declared unanswerable and those that Kant argues to be unanswerable by reason in the Antinomies of Pure Reason, there is much to be said for this diagnosis.66 So Nagarjuna here claims that we cannot make sense of the beginning or end of all of cyclic existence—beginnings and ends are beginnings and ends of actual, conventionally designated and delimited processes within cyclic existence.

2. Where there is no beginning or end,
How could there be a middle?
It follows that thinking about this in terms of
Prior, posterior, and simultaneous is not appropriate.


The concept of a middle, Nagarjuna argues, is bound up with those of beginnings and ends. We can say that we and all phenomena are within cyclic existence, but to posit determinate absolute spatiotemporal locations is senseless.

3. If birth came first,
And then old age and death,
Then birth would be ageless and deathless,
And a deathless one would be born.


Birth, old age, and death here are to be understood in an absolute sense. Of course, conventionally, the birth of a particular human being comes before her/his aging, which precedes her/his death. But that should not lead us to think of that birth as the origin of an entity, that aging as the midpoint in the life of that entity, or that death as the end of that entity. If one adopts a doctrine of rebirth, as does Ngrjuna and as do all of his interlocutors, the point can be made quite straightforwardly: For any sentient continuum, every birth is preceded by an aging and a death, and so forth.

But even setting aside the particular doctrine of rebirth, we can elucidate this insight with equal force: To see particular entities as having determinate, nonconventional beginnings of existence and determinate, nonconventional termini and, hence, that there are distinct times at which there is a clear fact of the matter about whether or not they exist, independent of conventions for their individuation, is to see those entities as having necessary and sufficient characteristics for their identity, that is, as having essences. But the central thesis Nagarjuna is defending is that this very conception of what it is to exist is incoherent—that things are empty of such essences and that the boundaries of objects are conventional and indeterminate. There is no fixed boundary between the existence of a seed, the tree to which it gives rise, a piece of wood from that tree, and a table fashioned therefrom or between the existence of an intact table, a broken table, wooden table parts, ashes, earth, the nutrients for a seed, that seed, the sapling to which it gives rise, and another tree.

Once we see the world from the standpoint of emptiness of inherent existence, the history of any conventionally designated entity is but an arbitrary stage carved out of a vast continuum of interdependent phenomena.67 The arising of any phenomenon, human, nonhuman sentient being, or inanimate object is the consequence of the disintegration of others. That disintegration succeeds their arising and aging. Once we give up the intrinsic identity of entities, the constant cycle of death, birth, aging, and rebirth of entities is unavoidable.

4. If birth were to come after,
And old age and death first,
How could there be a causeless aging and death
Of one not born?


But birth has to precede death as well, on pain of the absurdity of something that is unborn dying. And, as Nagarjuna points out in the next verse, we must think conventionally of these things in sequence because any conventionally designated object undergoes them in order:

5. Birth and age and death
Cannot occur at one time.
Then what is being born would be dying
And both would occur without cause.


6. When the series of the prior, simultaneous, and posterior
Is not possible,
Why are you led to posit
This birth, aging, and death?


The birth, aging, and death that the opponent has in mind can be represented at two levels: At the most general level, it is the birth, aging, and death of cyclic existence, the examination of which frames this discussion. At that level, Nagarjuna is pointing out that these conceptions, having legitimate employment only within the empirical realm, are nonsense. But the opponent could also be interpreted as positing birth, aging, and death as determinate, intrinsically identifiable moments in the evolution of empirical phenomena or, specifically, of sentient beings. v rejects that as well, arguing that moments intrinsically prior to, simultaneous with, or posterior to the existence of entities cannot be identified, given the lack of intrinsic identity of the entities themselves. So long as one in conceiving of phenomena thinks of them as temporally determinate and bounded, and thinks of the identity of things as intrinsic to them, one will have to identify their beginnings, middles, and ends. But this leads to paradox, given the indeterminateness, interdependence, and interpenetration of things. Nagarjuna hence advises the rejection of this ontology:

7. Not only is cyclic existence itself without beginning,
No existent has a beginning:
Neither cause and effect;
Nor character and characterized …


The alternative, both with respect to cyclic existence as a whole and with respect to individual entities, is to reject the ontology of entities and characteristics altogether, along with the boundaries and determinate relations that ontology requires:

8. Nor feeling and the feeler;
Whatever there is;
All entities
Are without beginning.


Chapter XII

Examination of Suffering


The first of the Four Noble Truths is that “all this is suffering.” So one can imagine an interlocutor granting all that has gone before, but in defense of Buddhist orthodoxy, insisting that suffering is inherently existent. After all, the Four Noble Truths are, from a Buddhist perspective, truths. Nagarjuna, of course, is a Buddhist and accepts the Four Noble Truths. (In fact, the principal chapter of this work, Chapter XXIV, is devoted to an exposition of the Four Noble Truths from the standpoint of emptiness and to the argument that only on Nagarjunas analysis can these truths be maintained at all.) So he must, without denying the reality of suffering, explain its emptiness.

1. Some say suffering is self-produced,
Or produced from another or from both.
Or that it arises without a cause.
It is not the kind of thing to be produced.


These are the four possibilities with regard to inherently existent suffering. The echo of I: 1 is obvious, and the argument here will depend heavily upon the analysis of dependent arising developed in that chapter and in Chapter VII.

2. If suffering came from itself,
Then it would not arise dependently.
For those aggregates
Arise in dependence on these aggregates.


Self-arising suffering would indeed be a candidate for inherent existence. But for the proponent of a Buddhist analysis of suffering, that is little help since suffering on a Buddhist analysis is the consequence of delusion, attachment, craving, action, and so forth. So such an analysis is not open to anyone wanting to defend the inherent existence of the suffering explored in the Four Noble Truths.

3. If those were different from these,
Or if these were different from those,
Suffering could arise from another.
These would arise from those others.


The next alternative—that suffering arises from another—requires that there be essential difference. For since suffering does arise from previous conditions, if there is genuine otherness, that would characterize the relation between suffering and its grounds.

4. If suffering were caused by a person himself,
Then who is that person
By whom suffering is caused
Who exists distinct from suffering?


But who is that other? It must be the sufferer himself at another stage, or another individual altogether. If it is the person himself, then as the cause of suffering, he must be distinct from suffering. This poses two problems: First, as per the analysis of motion, desire, and agency in Chapters II, VI, and VIII above, we cannot conceive of the sufferer as inherently different from the suffering he experiences. For part of his identity is constituted by that very suffering, and that suffering is his suffering. But second, given the framework of the first of the Four Noble Truths, a Buddhist philosopher such as Nagarjuna would share with any Buddhist interlocutor the assumption that in samsra sentient beings not only suffer, but are literally constituted of suffering—that every aggregate of a sentient being’s existence is a cause, an effect, and a basis of misery. So on either score, to distinguish sufferer from suffering for the purpose of such an analysis would be impossible.

5. If suffering comes from another person,
Then who is that person
When suffering is given by another—
Who exists distinct from suffering?


Another alternative is that the suffering is caused not by earlier stages of one’s own life, but by another individual. That other individual of course could be someone else entirely, in the ordinary sense, or it could be an earlier moment of what is ordinarily regarded as oneself, but which is for the purposes of this analysis regarded as substantially other. That is, taken in this way, Nagarjuna can be seen to be arguing on each side of a dilemma with regard to the identity of persons across time. But if this were so, it would have to be the case that the person in whom suffering was caused by that other could be identified and that that person could be distinguished from her suffering. But then the same problems developed above apply. Nagarjuna emphasizes this in XII: 6:68

6. If another person causes suffering,
Who is that other one
Who bestowed that suffering,
Distinct from suffering?
7. When self-caused is not established,
How could suffering be caused by another?
Whoever caused the suffering of another
Must have caused his own suffering.


But the suffering of that other person must either be caused by someone else or be self-caused. The former alternative leads to a regress: The whole point from the standpoint of the opponent who is the target of this argument is to find the independent explanatory ground for suffering. The second alternative leads back to the problem scouted in the opening verses: Self-caused suffering is both inconceivable within a general Buddhist soteriological framework and runs afoul of the arguments against self-causation generally. Finally, it is rather embarrassingly ad hoc. Nagarjuna sums this up in the next verse:

8. No suffering is self-caused.
Nothing causes itself.
If another is not self-made,
How could suffering be caused by another?


But, as Nagarjuna points out in XII: 9, it can’t be caused by both since we have seen that neither can be causally relevant at all to inherently existent suffering of a kind relevant to Buddhist doctrine. And it is absurd to suppose that it is uncaused:

9. If suffering were caused by each,
Suffering could be caused by both.
Not caused by self or by other,
How could suffering be uncaused?


10. Not only does suffering not exist
In any of the fourfold ways:
No external entity exists
In any of the fourfold ways.


The fourfold analysis is, of course, that in terms of the tetralemma of causation. And Nagarjuna is simply emphasizing that this refutation of the existence of inherently existing suffering is perfectly general. No entity can arise from itself, from another, from both, or from a noncause. This was the burden of the first chapter. We must, of course, recall that this is not a refutation of the existence of the suffering we all experience and wish to avoid. Rather it is a demonstration of its emptiness of inherent existence. For just as the analysis in Chapter I has provided the key to dismissing the inherent existence of suffering, the positive side of that same analysis can be used to recover its conventional existence. If by suffering we mean something dependently arisen, impermanent, and conventional, existing only as imputed and only in relation to its empty subjects, there is plenty of suffering to go around.

But moreover, not only is the existence of suffering rendered comprehensible on this analysis, but so is the possibility of the alleviation of suffering. For if the proponent of the inherent existence of suffering were correct, while it might seem that suffering would then have a more solid status than that vouchsafed it by Nagarjuna’s analysis in terms of emptiness, that very substantial existence and hence independence of other conditions would make its alleviation impossible. For if it exists independently, then there are no conditions in the absence of which it fails to exist. So Nagarjuna’s analysis not only makes good sense of the first truth—that of suffering—and by implication of the second—that of the cause of suffering—but also opens the door for an analysis of the third and fourth truths—those of cessation and of the means to cessation.


Chapter XIII

Examination of Compounded Phenomena


In this chapter, Nagarjuna begins to develop the idea of emptiness more explicitly. Up to this point, he has been arguing that phenomena are empty, but has not been characterizing emptiness itself, or its relation to entitihood or to conventional reality, except by example and by implication. At this point, through a general discussion of all compounded phenomena—that is, all phenomena constituted of parts or brought into being dependent upon causes—he argues explicitly both that emptiness is the lack of essence and that emptiness itself is wholly negative in character. It is not an essence that things have instead of whatever essence naive common sense or sophisticated reification might have thought they had—rather, it is the total lack of essence or inherent existence. This is, hence, an anticipation of the explicit discussions of the emptiness of emptiness to follow.

1. The Victorious Conqueror has said that whatever
Is deceptive is false.
Compounded phenomena are all deceptive.
Therefore they are all false.


This is an important verse for any understanding of the relation of the two truths—the conventional and the ultimate—to one another. That relation i�ހs vexed because the conventional truth is sometimes referred to as a truth and sometimes as wholly false. Conventional phenomena are sometimes referred to as empirically real and not imaginary and sometimes as wholly imaginary.70 So it is important to see that the sense of “falsehood” in play when the conventional is characterized as false is “deceptive.” That is, insofar as conventional phenomena present themselves as more than conventional—as inherently existent—they deceive us. We take them to be what they are not—to be intrinsically identified, inherently existent entities. In that sense, they are false. But to the extent that we understand them as dependently arisen, empty, interdependent phenomena, they constitute a conventional truth. Yet one must bear in mind that, according to Nagarjuna, perception untutored by Madhyamika philosophy and rigorous practice delivers objects to consciousness as inherently existent. In this sense, the things that we see are wholly false. For most of us, the best that we can do is reason our way into knowing, but not seeing, their true nature. The goal of meditation on emptiness is to bring this knowledge into perceptual experience and, hence, to see things as they are.

2. If whatever is deceptive is false,
What deceives?
The Victorious Conqueror has said about this
That emptiness is completely true.


The opponent then asks what we are deceived about. Here is what motivates the question: If there are no real tables, for instance, then when I believe that there is a table in front of me and am therefore deceived, what is deceiving me? We don’t want to say that a nonexistent phenomenon is pretending to be existent since it would have to exist in order to pretend. Nagarjuna replies that what actually exists is an empty table. (That is not to say, however, that that empty table is inherently existent—only that the correct way to characterize the entity that exists conventionally is as an “empty table.”) That empty table is misperceived by an ordinary mind as a truly existent table. To the extent that it appears as empty, it appears as it truly is. In the first two lines of the next verse, Nagarjuna notes that it is the absence of essence that permits change:

3. All things lack entity (hood),
Since change is perceived.
There is nothing without entity
Because all things have emptiness.


It is emptiness that makes change possible. If things had essences, they would be incapable of real change. But since they are seen to change, Nagarjuna argues, they must be empty of essence. The opponent, though, rejoins: Since according to Nagarjuna all things are empty and since this is their ultimate nature, all things in fact do have a kind of entitihood, namely, existence as empty phenomena. Nagarjuna is here anticipating the charge that he has rejected other essences only to posit emptiness as an essence, subject to all of the problems he has already adumbrated for essentialist metaphysics.

The opponent then asks (XIII: 4), “If everything lacks being, and is therefore empty, what could change?” Change would seem to have to be change of something, and the doctrine of emptiness seems to rob us of those somethings. Nagarjuna, hence, presents himself, in the voice of the opponent, with a dilemma: He seems to have propounded, his protestations to the contrary notwithstanding, a theory of the essence of all phenomena. That theory, according to this hypothetical objection, is that emptiness just is the essence of all phenomena. He could deny having propounded such a theory, of course. But the consequence of such a denial, the opponent charges, would be no better. For then, the very basis of the argument here offered for emptiness—the reality of change—would have to be rejected. This is because without real entities there would no longer be a possible subject of change. Nagarjuna replies in the third and fourth lines of XIII: 4 that the opponent has things backward: If there was entitihood—if things were nonempty—change would be impossible. It is emptiness itself that makes change comprehensible:

4. If there is no entity (hood),
What changes?
If there were entity,
How could it be correct that something changes?


Now Nagarjuna begins a brief explanation of how to understand change in the context of emptiness and of why entitihood would preclude change. This discussion is certainly grounded in the analysis in Chapter II, but is more explicitly tied to the doctrine of emptiness at this point in the text:

5. A thing itself does not change.
Something different does not change.
Because a young man doesn’t grow old,
And because and an old man doesn’t grow old either.


When we imagine change, we imagine one thing retaining its identity, but changing its properties. But if identity is understood strictly, it is only possible as an internal relation that a thing bears to itself. To the extent that a thing changes, it becomes, strictly speaking, a different thing. But the relation between two things is not t���he change of a thing—it is simply the difference between two nonchanging entities. A young man does not grow old. When he is old he is no longer a young man. The relation between the young man and the old man is simply the difference of two things. But an old man doesn’t grow old either. He is already old. So if change and things that change are thought of nonrelationally, we can make no sense of change at all.

6. If a thing itself changed,
Milk itself would be curd.
Or curd would have come to be
An entity different from milk.


If we think of identity persisting through change, there is a single thing that changes as conventionally, milk becomes curd. Since that thing is identical to milk and to curd, by transitivity we would have to say that curd and milk are identical. But no one would want to put curd in his/her tea! The only way to avoid this result while retaining the idea that milk and curd are entities would be to consider them to be wholly different entities. In that case, there is still no change in an entity—only the difference between two unrelated phenomena.72

7. If there were even a trifle nonempty,
Emptiness itself would be but a trifle.
But not even a trifle is nonempty.
How could emptiness be an entity?


 
Verses 7 and 8 are critical for any understanding of the subtle doctrine Nagarjuna is developing of the emptiness of emptiness. In XIII: 7, Nagarjuna is emphasizing that emptiness is not one of the many properties that a thing might or might not have. It is not that some things are empty and some are nonempty, or that all things happen to be empty although they might have been otherwise. Emptiness is important because it is the only way that things can exist. Moreover, emptiness is not an entity. It is not a distinct phenomenon to which other phenomena are related. It is exactly the emptiness of all phenomena.73 The conventional character of conventional entities and their emptiness are one and the same.

8. The victorious ones have said
That emptiness is the relinquishing of all views.
For whomever emptiness is a view,
That one will accomplish nothing.


The sense of “view” (Tib: lta-ba, Skt: di) at work in verse 8 is crucial. By a view, Nagarjuna here means a theory on the same level of discourse at which reificationist-nihilist debates proceed. A view in this sense is a view about what does or does not exist when existence is taken to mean inherent existence, or about the nature of phenomena, presupposing that the idea of a nature is coherent. So both the theory that compounded phenomena exist in virtue of having natures and identities and the theory that since they don’t have such natures and identities they don’t exist at all are views in this sense. Both presuppose that things exist at all if and only if they do so inherently. But the analysis in terms of emptiness is not a view at all in this sense. For the claim is not that things exist in virtue of having the property of emptiness as an essence. Rather it is the claim that they are empty because they have no essence.

It is also very important to see that this understanding of what a view is is closely bound up with Nagarjuna’s account of assertion and of the role of language in Madhyamika dialectic. For Nagarjuna, assertion in the literal sense is always the ascription of a property to an entity. As long as we are talking from the conventional standpoint, there is no problem here. There are plenty of conventional entities and conventional properties to go around and, so, lots of available conventionally true assertions. That is the basis of conventional truth. It is also important to note here that corresponding to these conventional assertions are real propositions that make them true or false—entities with or without the ascribed properties. Again, as long as we remain and are aware that we remain within the framework of conventional designation and conventional assertion, this poses no problems.

But, when we start to do metaphysics, it is easy to slip into nonsense: For now, when we want to characterize the essence of a thing, we take ourselves to be positing a non-conventional thing and ascribing to it an essential property. And there not only are no such things, but there are not even possibly such things. There is no ultimate way the world is that we are characterizing, truly or falsely.

The danger to which Nagarjuna is here adverting with respect to Madhyamika philosophy (of treating Madhyamika as a view) is then connected to assertion in the following way: If one were to think that in asserting that things are empty that one is positing entities and ascribing to those independent entities the property of emptiness, one would be treating the language of Madhyamika as making literal assertions. But from the standpoint from which these would be true, there are no entities and no characteristics, and a fortiori, there are no entities having the characteristic of being empty. The language must hence be understood, from the ultimate perspective, not as making assertions, but rather as ostending—indicating that which cannot be literally asserted without falling into nonsense—as Wittgenstein puts it in the Tractatus, showing that which cannot be said.

Nagarjuna makes this much more explicit in his discussion of positionlessness in Vigrahavyvartanī XXI-XXVIII, where he explicitly denies that the Madhyamika assert any propositions, in virtue of there being no entities or properties presupposed by their use of language existing independently and corresponding to the words used. Āryadeva makes the same point at Catuśtaka XVI: 21. Candrakīrti in his comments on these verses compares one who treats emptiness as an essential property—as opposed to the lack of any essential property, thus treating Madhyamika language as assertoric in the sense of asserting the view that all things have the essential nature of emptiness—to one who, upon entering a shop and learning that there are no wares for sale, asks the shopkeeper to sell him the “no wares.”74, 75

To hold a view of emptiness—to reify it and then attribute it to phenomena—would then involve simultaneously reifying those phenomena as having a fixed nature and denying their existence at all, in virtue of disparaging their conventional reality as unreality by contrast with the reality of emptiness. It is this incoherence, so characteristic of essentialist philosophies, that leads Nagarjuna to assert that one holding such a view is completely hopeless—incapable of accomplishing anything, philosophically or soteriologically.76, 77

This argument against the coherence of any understanding of emptiness as itself an essence is tied very tightly to the analysis in Chapter XXIV: 18-40 of the emptiness of emptiness and of the connection between emptiness, dependent arising, and convention and tied most directly to the concluding verse of the text, XXVII: 30. (The commentaries on XXIV: 36 and XXVII: 30 below may be useful in elucidating this verse as well.) It is clearly an early anticipation of the powerful and climactic conclusions drawn in those two discussions.


Chapter XIV

Examination of Connection


The word here translated as “connection” (phrad-pa) is the term denoting the relation between the components that are compounded in any compounded phenomenon. It can also describe the relation between two things coming together in space and time or colliding, or two things fitting together, and while this can be taken fairly literally in the context of physical objects when they are understood as compounded of their parts, the relation is actually much more general than that. In fact, the example that Nagarjuna takes as central, and one that is used by some earlier Buddhist theorists as an example of a case of connection in this sense, is visual perception. In such a case, according to the proponent of the reality of meeting, or compounding, the subject, the sensory organs, the sensory faculty, and t���he object join together, or “connect,” not in a literal physical sense of spatiotemporal coincidence, but rather in the sense of forming an ensemble. Sense perception is, on this view, the entire compound ensemble.

So, dialectically, this chapter follows quite naturally on the heels of the examination of compounded entities. For we can imagine an opponent might reason as follows: Nagarjuna may be right in denying the inherent existence of compounded entities in virtue of their dependence upon their parts and upon their parts being compounded, but surely since these phenomena depend upon being compounded that relation—the connection—exists. This chapter is aimed at replying to this position.

1. The seen, seeing, and the seer:
These three—pairwise or
All together—
Do not connect to one another.


First, he claims, these things simply don’t occur in the same place at the same time. There is no literal sense in which they connect.

2. Similarly desire, the desirous one, the object of desire,
And the remaining afflictions
And the remaining sources of perception
Are understood in this threefold way.


In the various chapters on the relation between characteristic and characterized, Ngrjuna has argued that it makes no sense to think of the relation between individuals and their properties or between entities as any kind of relation between independent entities at all, and that these phenomena cannot be understood as the same, as different, or as neither.

3. Since different things connect to one another,
But in seeing, etc.,
There is no difference,
They cannot connect.


In order to have things that connect in the relevant sense, they must be different from one another, but as we saw in the chapters on characteristics, on desire, on seeing, on action, on motion, and on the self, the differences of the relevant kind are not found on analysis.

4. Not only in seeing, etc . . .
Is there no such difference:
When one thing and another are simultaneous,
It is also not tenable that there is difference.


This problem emerges not only in the analysis of intuitively unitary phenomena like vision, but is perfectly general. Things that are separate from one another cannot be coherently thought of as inherently different entities either. For without any inherent identity, there is no basis for inherent difference. This recalls the argument of Chapter I.

5. A different thing depends on a different thing for its difference.
Without a different thing, a different thing wouldn’t be different.
It is not tenable for that which depends on something else
To be different from it.


For there to be substantial difference, it must be possible to independently establish the identity and natures of the relata. But this, Nagarjuna has argued repeatedly, is impossible.

6. If a different thing were different from a different thing,
Without a different thing, a different thing could exist.
But without that different thing, that different thing does not exist.
It follows that it doesn’t exist.


That is, the only way that difference or the identity of a different thing as different could be shown to exist inherently would be for that difference to be present independently of the existence of another different thing. But that is not so. The only alternative would be to argue that difference is present independently in single things. But this ignores the relational character of difference.

7. Difference is not in a different thing.
Nor is it in a nondifferent thing.
If difference does not exist,
Neither different nor identical things exist.


So difference cannot be located either as a relation between things or as a unary property of individual things. So there is no inherently existent difference. But it is the existence of inherent difference that grounds the problem of connection. So there is no such relation, and no problem to be solved.

8. That does not connect to itself.
Nor do different things connect to one another.
Neither connection nor
Connected nor connector exist.


The conclusion is a powerful one and, especially when conjoined with the conclusion of the previous chapter, goes to the heart of any Buddhist (or non-Buddhist, for that matter) ontology that seeks to reify the entities that appear at any stage of ontological analysis. It is quite tempting when examining dependent, compound phenomena to think that while they themselves might not be inherently existent, and might not be the ultimate entities of the empirical world, it must at least be a fundamental fact that their being constituted of parts, or dependent upon their location in a causal and mereological nexus, exists as a fact. That would seem, in fact, to be the natural way to interpret the doctrine of dependent origination and the emptiness of macroscopic entities. But Nagarjuna here pulls the rug out from any such analysis, pointing again to the emptiness of emptiness: Not only are compounded phenomena empty of inherent existence, but so is the relation among their constituents and determinants in virtue of which they are compounded.


Chapter XV

Examination of Essence


This chapter continues the discussion begun in Chapter XIII and carried on in Chapter XIV of the fundamental nature of things and the relation between emptiness and existence. Here Nagarjuna rejects the coherence of the concept of essence and explores its ramifications for the concept of inherent existence, the concept of an entity, and the concept of a nonentity. This chapter is also aimed at dispelling any nihilistic interpretation of the Madhyamika philosophical orientation and in explaining the deep connection between the analysis of phenomena as empty of essence and the demonstration of the possibility of empirical reality.

1. Essence arising from
Causes and conditions makes no sense.
If essence came from causes and conditions,
Then it would be fabricated.


Essence by definition is eternal and independent. So it can’t arise dependently. Chapter XV: 1, 2 develop this point directly. But since all entities arise dependently, it follows that none of them have essence.78

2. How could it be appropriate
For fabricated essence to come to be?
Essence itself is not artificial
And does not depend on another.


In these first two verses, Nagarjuna indicates the three cardinal characteristics of an essence: An essence (or an entity that exists in virtue of possessing an essence) is uncaused, independent of other phenomena, and not fabricated from other things. It is important to bear this in mind in any Madhyamika analysis of emptiness. For when Nagarjuna argues that phenomena are all empty, it is of essence in this sense that they are empty. Hence, when Nagarjuna argues that all phenomena originate in dependence upon conditions, that all phenomena are interdependent, and that all phenomena are fabricated (both in virtue of being compounded from parts and in virtue of acquiring their identity as particulars through conceptual imputation), he is thereby arguing quite directly for their emptiness.

3. If there is no essence,
How can there be difference in entities?
The essence of difference in entities
Is what is called the entity of difference.


This is an echo of the argument about difference presented in Chapter I. Essential difference presupposes essences of individuals. So any argument against individual essence will count as an argument against essential difference.

4. Without having essence or otherness-essence,
How can there be entities?
If there are essences and entities
Entities are established.


The concept of an inherently existent entity is the concept of an entity with an essence. So without essence, there are no inherently existing entities.

5. If the entity is not established,
A nonentity is not established.
An entity that has become different.
Is a nonentity, people say.


By a nonentity, Nagarjuna means something inherently different from some existing entity. A nontable in this sense would be inherently different from a table. But a nonexistent in general would be a Meinongian subsistent which is available as a basis of predication but is intrinsically different from what it is to be an existent—a real thing possessed of the property of being nonexistent. Just as a table must be established as a determinate entity in order to establish the nature of nontables, existence must be established as an inherently existent property in order to establish the parallel status of nonexistence. But neither tables nor existence can be so established. By the same token, then, there are no inherently established nontables, nor any inherently established nonexistents in their stead. So even though it might appear that an analysis through emptiness would leave us only with nontables and nonexistent phenomena, it doesn’t even leave us with that (inherently), though it leaves us with plenty of tables, nontables, existents, and nonexistents (conventionally).

6. Those who see essence and essential difference
And entities and nonentities,
They do not see
The truth taught by the Buddha.


If the only way that one can think about phenomena is to think of them as things with inherent natures and to think of things without such natures as thereby nonexistent, none of the Buddhist doctrines of impermanence, emptiness, or liberation will make any sense.

7. The Victorious One, through knowledge
Of reality and unreality,
In the Discourse to Ktyyana,
Refuted both “it is” and “it is not.”


In the Discourse to Ktyyana, the Buddha argues that to assert that things exist inherently is to fall into the extreme of reification, to argue that things do not exist at all is to fall into the extreme of nihilism, and to follow the middle way is neither to assert in an unqualified way that things exist nor in an unqualified way that things do not exist. It represents one of the fundamental suttas of the Pali canon for Mahyna philosophy. In the sutta, the Buddha claims that reification derives from the failure to note impermanence and leads to grasping, craving, and the attendant suffering. Nihilism, he claims, is motivated by the failure to note the empirical reality of arising phenomena. It leads to suffering from failure to take life, others, and morality seriously enough. The middle path of conventional existence leads to engagement in the world without attachment.79

8. If existence were through essence,
Then there would be no nonexistence.
A change in essence
Could never be tenable.


If for a thing to exist were for it to be a determinate entity with an essence, then no thing would ever cease to exist or change in any way. For an essential property is a necessary property, and it is incoherent to say that a thing loses a necessary property.

9. If there is no essence,
What could become other?
If there is essence,
What could become other?


In the first half of this verse, the opponent replies that since the argument in the previous verse presupposes the reality of change, it must presuppose the reality of the changer. If it presupposes the reality of change, it presupposes the reality of things that change and, hence, that persist through time. In order to remain the same, there must be some essence that accounts for this identity. Nagarjuna replies, however, that if this persistence through time were determined by essence, the change it putatively explains would be impossible. Only conventional existence over time can explain change. Nagarjuna summarizes, paraphrasing the Discourse to Ktyyana:

10. To say “it is” is to grasp for permanence.
To say “it is not” is to adopt the view of nihilism.
Therefore a wise person
Does not say “exists” or “does not exist.”


11. “Whatever exists through its essence
Cannot be nonexistent” is eternalism.
“It existed before but doesn’t now”
Entails the error of nihilism.80


To say that if something exists, it does so in virtue of having an essence and hence cannot change or pass out of existence would entail the absurd position that everything is eternal. To say of something that it existed in this strong sense—with an essence—in the past, but does not do so now, is absurd. For if for something to exist is for it to do so inherently, and if it is not now existent, it could never have been. So since everything we observe is impermanent, if the only existence that there could be were inherent existence, nothing could exist at all. That would be nihilism. The upshot of this chapter is that the very concept of an essence, and hence the very concept of an inherently existent entity at all, is incoherent. No coherent conception of the phenomenal world can be one in which things are posited other than conventionally.81


Chapter XVI

Examination of Bondage


So there are no entities. But still, from a Buddhist perspective, we are bound: bound to our conceptions of entities and essence, bound to our selves, bound to objects, and principally, bound to cyclic existence itself. Surely, the opponent might ask, mustn’t the bondage that accounts for the illusions so ruthlessly analyzed in the previous chapter be intrinsically real? If not, what is the causal basis for all of these illusions and all of this suffering? In a Buddhist framework, this bondage to cyclic existence is instantiated in endless transmigration in sasra, and freedom from bondage would be liberation from cyclic existence into nirva. We will postpone a discussion of the precise nature of that liberation and of nirva until we reach the chapters where that topic is discussed, namely, XXII and XXV. Nagarjuna begins with an examination of transmigration and the entity that transmigrates:

1. If compounded phenomena transmigrate,82
They do not transmigrate as permanent.
If they are impermanent they do not transmigrate.
The same approach applies to sentient beings.


Nagarjuna sets up a by now familiar destructive dilemma: Either compounded phenomena—of which sentient beings, the beings who are bound, are instances—are permanent or impermanent. Let us just consider the compounded phenomena who are sentient and hence who transmigrate: If they are thought of as permanent, they cannot transmigrate because transmigration involves, by definition, change. And what is permanent, as we have seen, cannot change. But if they are impermanent, then they do not endure through time and, hence, cannot transmigrate. So no sentient being considered as an inherent entity can be conceived of as a transmigrator in cyclic existence.

2. If someone transmigrates,
Then if, when sought in the fivefold way
In the aggregates and in the sense spheres and in the elements,
He is not there, what transmigrates?


Given that no inherently existent person can be found upon analysis as the bearer of the aggregates, as identical to the aggregates, as different from the aggregates, as the collection of the aggregates, or as the arrangement of the aggregates, and mutatis mutandis for other possible modes of analysis in terms of domains of knowledge or experience and in terms of basic elements, it follows that there is no inherently existent subject of transmigration. If the transmigrator cannot be identified on analysis, though, neither can the transmigration itself. It will follow that there is no inherently existent transmigration and, hence, no inherently existent bondage to cyclic existence.

3. If one transmigrates from grasping to grasping, then
One would be nonexistent.
Neither existent nor grasping,
Who could this transmigrator be?


Grasping” here refers primarily to grasping the aggregates as one’s self. Transmigration—or for that matter continuation within one life, which from the Madhyamika perspective is exactly the same kind of process—involves moving from grasping one set of phenomena as one’s self to grasping another in the same way. That is one of the most fundamental delusions from a Buddhist standpoint. But grasping can also be the grasping of an object as an object, or the clinging to possessions. Life in sasra, Nagarjuna would insist, can equally well be characterized in any of these ways. But if in order to exist as an individual one would have to retain one’s identity over time since on this view it is of the very nature of cyclic existence that one constantly changes from one moment to another, then it would follow that no subject exists. But if there is no subject of grasping, there can be no grasping. So, on the supposition that to exist and to transmigrate is to exist as a continuing entity, there is no way to make sense of the phenomenal world. So an inherently existent grasper, posited in order to guarantee the reality of cyclic existence, in fact makes the reality of cyclic existence incoherent.

4. How could compounded phenomena pass into nirva?
That would not be tenable.
How could a sentient being pass into nirva?
That would not be tenable.


If compounded phenomena are permanent, grasping is permanent. And if grasping is permanent, sasra is permanent. And if sasra is permanent, then nirva is impossible. But the philosopher who is positing inherently existent bondage is doing so in order to defend a Buddhist perspective on cyclic existence and nirva. This is precisely the motivation for the reification—the worry that sasra and nirva are, if not inherently existent, nonexistent. So this conclusion is inadmissible for such an opponent.

5. All compounded phenomena are arising and ceasing things:
Not bound, not released.
For this reason a sentient being
Is not bound, not released.


Neither bondage nor release can be seen as inherently existent, nor as inherent properties of sentient beings. This is the conclusion of the argument that follows. Nagarjuna first considers bondage as an inherent property, and then liberation:

6. If grasping were bondage,
Then the one who is grasping would not be bound.
But one who is not grasping is not bound.
In what circumstances will one be bound?


If grasping is identified with the property of bondage, then the continuity of bondage across transmigration is inexplicable: The problem is that grasping is not only the cause, but is also the effect of bondage. Delusion by which we are bound, from a Buddhist perspective, leads us to grasp at things; that grasping perpetuates delusion and bondage. To the extent that we grasp onto external phenomena or onto the self as inherently existent, we are bound to the delusions that constitute and ground sasra. To the extent that we are bound in delusion, we continue to grasp. The bondage is hence not only conditioned by, but overarches, particular instances of grasping. But we don’t want to infer from the fact that grasping and bondage are not identical that the relinquishing of all grasping would not free one. The task is then to figure out the nature of bondage, which must be conceived as relational.

7. If prior to binding
There is a bound one,
There would be bondage, but there isn’t.
The rest has been explained by the gone, the not-gone, and the goer.


The only way that bondage itself could be an inherently existent phenomenon would be if it could exist prior to and independently of a bound sentient being. But then the case would be strictly analogous to motion (as well as to several other analysands we have considered so far). That is, just as there is no motion apart from the mover, there is no bondage apart from the bound. The argument can be applied in a strictly parallel way.

8. Whoever is bound is not released.
Whoever is not bound does not get released.
If a bound one were being released,
Bondage and release would occur simultaneously.


Nagarjuna then recalls another argument from Chapter II, the argument against the possibility of the beginning of motion. There, Nagarjuna argued that motion could not begin in a stationary object since it is not moving, nor in a moving object since it is already in motion. And there can be no moment when a thing is both moving and stationary, nor any moment when an entity is neither. Similarly, nirva cannot arise in one in sasra, nor in one already in nirva. One cannot be simultaneously in sasra and nirva. Nor is there any third option.

9. “I, without grasping, will pass beyond sorrow,
And I will attain nirvana,” one says.
Whoever grasps like this
Has a great grasping.


There is a stylistic feature in this verse that deserves note: The pronoun “I” (bdag) is uncharacteristically fronted in the sentence and is emphasized with the focus particle (ni). Nagarjuna is hence drawing attention to the fact that the individual in whose mouth this verse is put is grasping to his own identity as an agent and as a continuing subject both through sasra and into nirva. This grasping onto self, he suggests, precludes the nirva the speaker craves. But Nagarjuna presents another argument as well: It is also possible to grasp after nirva—to reify it as a state and to crave it as a phenomenon inherently different from sasra and as highly desirable since it is indeed characterized as liberation from suffering. But this grasping onto the end of grasping is itself a grasping and so precludes the attainment of nirva. nirva requires, according to Nagarjuna, a complete cessation of grasping, including that onto nirva itself. While that might seem paradoxical, it is not: To grasp onto something in this sense requires, inter alia, that one reify it. By refusing to reify liberation, in virtue of seeing it as the corelative of bondage, which itself is not inherently existent, it is possible to pursue the path to liberation without creating at the same time a huge obstacle on that path—the root delusion with regard to nirva itself. Possible, that is, but perhaps not that easy.84, 85

10. When you can’t bring about nirva,
Nor the purification of cyclic existence,
What is cyclic existence,
And what is the nirva you examine?


Anyone who is subject to either of these pathologies—grasping to one’s self or grasping for nirva—is incapable of attaining that peace. So, Nagarjuna urges, in order to make such progress possible, one should reexamine one’s conception of the nature of phenomena in cyclic existence (both oneself and external phenomena) and nirva itself. By coming to see their ultimate emptiness, he suggests, one can relinquish that grasping and attain that liberation.

Neither nirva nor samsara are inherently existent. Ultimately both are nonexistent. So, what, Nagarjuna asks rhetorically, are they? The answer is that they are conventionally real, dependently arisen phenomena that are empty of inherent existence. In virtue of that fact, it is possible to escape the former and to attain the latter. But that escape would be impossible were they inherently existent and is impossible for anyone who takes them to be so.


Chapter XVII

Examination of Actions and Their Fruits


Arguing for the emptiness of bondage and liberation, however, raises a further question that demands an answer: If there is no real bondage and no real release, what are the effects of our actions? For it would appear, at least given standard Buddhist moral theory and the doctrine of karma on which it is grounded,86 that meritorious actions conduce to liberation and that morally wrong actions increase bondage. Given the emptiness of these latter, an analysis of the consequences of action is in order. Nagarjuna begins with Buddhist moral truisms, accepted by the Madhyamika as well as by members of other Buddhist schools. It is important to note that the first nineteen verses of this chapter represent the views of four distinct opponents in order of increasing similitude to the Madhyamika understanding. Despite the fact that Nagarjuna sets these views up as targets, however, some of the views the opponents put on the table are, suitably interpreted, shared by Nagarjuna. Each

can be seen as, despite being inadmissible as a characterization of a nonconventional basis for the relation between action and its effects, a reasonable empirical assessment of at least part of the conventional reality in this domain.

1. Self-restraint and benefiting others
With a compassionate mind is the Dharma.
This is the seed for
Fruits in this and future lives.


2. The Unsurpassed Sage has said
That actions are either intention or intentional.
The varieties of these actions
Have been announced in many ways.


The classification to which Nagarjuna refers is a partition of actions into mental and physical. Mental actions are mere intentions on this view; physical actions and speech (generally distinguished in Buddhist psychology and action theory) are properly intentional. That is, the latter two involve a mental and a nonmental component; the mental actions only involve a mental component. Verse 3 clarifies this:

3. Of these, what is called “intention
Is mental desire.
What is called “intentional”
Comprises the physical and verbal.


In the next verse, an opponent uses these truisms as a platform for the defense of the view that actions themselves must remain in existence until their consequences are observed. Actions that derive from renouncing the world are different from those that derive from worldly concerns. This difference in nature, he argues, must explain the difference in their consequences:

4. Speech and action and all
Kinds of unabandoned and abandoned actions
And resolve
As well as …


5. Virtuous and nonvirtuous actions
Derived from pleasure,
As well as intention and morality:
These seven are the kinds of action.


The kinds of actions to which Nagarjuna’s imaginary opponent refers are simply the various kinds of virtuous and nonvirtuous actions. In general, morally good actions are done for the sake of pleasure for others; morally bad actions sacrifice others’ good for one’s own pleasure. The opponent, however, goes further, pointing out that these actions have diverse long-term consequences that must be explained:

6. If until the time of ripening
Action had to remain in place, it would have to be permanent.
If it has ceased, then having ceased,
How will a fruit arise?


The problem is this: Given that the consequence of an action may be far in the future, something must persist to connect the action to the result. This is a kind of karmic analog of doubts about action at a distance. It is the same kind of move that lies behind trace theories of memory in recent philosophy of mind. So this first position is that there must be some permanent entity that remains in existence until the consequences of an action occur.

A second possibility is that some third thing mediates the relation between action and consequence—a kind of karmic link that is generated by the action and remains in the psychophysical continuum until the consequence is produced. The interlocutor then offers an analogy popular in Buddhist philosophy:

7. As for a continuum, such as the sprout,
It comes from a seed.
From that arises the fruit. Without a seed,
It would not come into being.


That is, just as every actual fruit requires an actual seed as its predecessor and a sprout to mediate between them, the opponent reasons, every consequence of action requires an actual action and an actual karmic link between the action and the consequence. The next three verses extend this analogy:

8. Since from the seed comes the continuum,
and from the continuum comes the fruit,
The seed precedes the fruit.
Therefore there is neither nonexistence nor permanence.


That is, this interlocutor points out, the position developed in XVII: 5–6 requires that actions either be permanent or nonexistent. His own view, on the other hand, allows actions to exist as impermanent and is, hence, more plausible:

9. So, in a mental continuum,
From a preceding intention
A consequent mental state arises.
Without this, it would not arise.


10. Since from the intention comes the continuum,
And from the continuum the fruit arises,
Action precedes the fruit.
Therefore there is neither nonexistence nor permanence.


In the next verse, another opponent offers an orthodox formulation from a substantialist Buddhist school, arguing that particular kinds of action are described as the methods of attaining realization and that particular rewards for the practicioner are mentioned as consequences of realization. The implication is that, since these are specified in stras as real, they must be inherently existent:

11. The ten pure paths of action
Are the method of realizing the Dharma.
These fruits of the Dharma in this and other lives
Are the five pleasures.


“The ten paths” simply denotes the totality of virtuous actions as characterized by one of the Buddhist botanies of morally worthy action. The five pleasures are the pleasures appropriate to the various sense faculties. According to the opponent, all we need to do in order to reach enlightenment and to lead good lives is to act virtuously. The principal consequence of this is that we will enjoy temporal happiness.

Yet another interlocutor replies that this wholly misunderstands the Buddha’s explanation of the relation between action and its consequences. While it is the case that acting well is an important ingredient in Buddhist practice and in any account of what it is to lead a good life, and while it is true that when one lives well, one in general is rewarded with material happiness, this hardly indicates that action, the agent, or the consequences of action are inherently existent. Rather, this more sophisticated opponent suggests, the nature of the link is completely abstract, like a legal obligation:90

12. If such an analysis were advanced,
There would be many great errors.
Therefore, this analysis
Is not tenable here.


13. I will the n explain what is tenable here:
The analysis propounded by all
Buddhas, self-conquerors
And disciples according to which …


14. Action is like an uncancelled promissory note
And like a debt.
Of the realms it is fourfold.
Moreover, its nature is neutral.


Using the metaphor of a promissory note, the defender of this view compares action and its consequences to a document attesting to a particular debt or other legal action: Though the act to which the document attests was in one sense momentary, its consequences, and the evidence of its reality, are unlimited in duration. So the consequences of any action—however local that action might appear to be—reverberate through all realms of existence. 91 Moreover, the fundamental nature of action and its consequences is neutral. That is, simply considered as such, on this view, neither action nor its consequent trace is either positive or negative. Any particular action or trace may of course be so—but action itself is equally capable of being positive or negative in character. We now turn to specific advice to enable one to realize the nature of reality and to abandon the mundane concerns and attachments that lead to binding actions (advice with which Ngrjuna would not take issue):

15. By abandoning, that is not abandoned.
Abandonment occurs through meditation.
Therefore, through the nonexpired,
The fruit of action arises.


Simply by resolving to abandon attachment one cannot thereby succeed in shedding it. It is difficult to accomplish this. Attachment arises as a consequence of the persistent, pervasive psychological, verbal, and physical habits that together constitute what Buddhist philosophers call the “root delusion,” the ignorance of the true nature of things. That delusion consists in confusing existence with inherent existence and issues inevitably in one of the two extreme views—reification or nihilism. Only through extensive meditation on the nature of phenomena and on the nature of emptiness can these habits be abandoned, and only through an understanding of the ultimate nature of things can the fruit of actions done through abandonment—that is, liberation from the suffering of cyclic existence—be attained. The promissory note metaphor is at work here as well. The idea is that one cannot simply cancel a promissory note on one’s own without paying the debt. One must do something more substantial to discharge one’s obligation to one’s creditor.

16. If abandonment occurred through abandoning, and
If action were destroyed through transformation,
The destruction of action, etc.,
And other errors would arise.


If one thought that one could just resolve to abandon attachment and delusion and succeed, that would be to treat attachment and attached action as trivial entities—even as illusory in the full sense. Just as when one sees a mirage, one can, knowing that it is a mirage, stop seeing it as water. That is possible for illusory things, but not so for empirically real ones. It takes effort to see an actual puddle as empty—not of conventional water, but of nonconventional inherent existence—and it takes effort to stop reifying habits. Again, though this is articulated in defense of the opponent’s view, this is a sophisticated opponent, and Ngrjuna in fact agrees with much of this.

17. From all these actions in a realm,
Whether similar or dissimilar,
At the moment of birth
Only one will arise.


One performs countless various actions in one’s life. And the confluence of the karmic consequences of all of them, on this view, are realized in the beginning of a single individual at the moment of rebirth (the one who arises). This comment is, of course, most directly about rebirth and the mechanism of karma in transmigration. Here is a way to understand that explicit point: The mechanism by which karma operates in rebirth is not that each individual action in a continuum designated as an individual remains permanently in place or leaves a substantial trace that lies dormant until it produces its consequence. This is indeed how karma is often conceived by substantialist Buddhist schools. Rather, each moment of such a continuum, including the moment of rebirth, is a consequence, through the mechanism of dependent arising, of all of the previous moments of that continuum (and, of course, of much else besides). Those karmic consequences are, as it were, “summed up” in the total state of the individual at birth.

But of course the implications of this are more general and concern every moment of any life. They can hence be made independently of any discussion of transmigration, though of course they help to demystify that Buddhist doctrine, at least as it is conceived in Mahayna philosophy. The point is this: Every moment of our lives represents the causal consequences of, inter alia, all of our prior actions. No action “lies dormant” waiting for its consequences to emerge. Nor does any action somehow become “canceled” when some salient consequence is noticed. There is no accounting kept, and no debit and credit system, either from the causal or the moral point of view in the continuum of human action and experience. Rather, at each moment we are the total consequence of what we have done and of what we have experienced. And the only sense in which some past action may determine some future reward is one in which that past action, as well as other conditions, have determined a state now that, together with other future conditions, will determine that reward. Mutatis mutandis, of course, for negative consequences. This sober empiricist account of these matters forms the basis for Mahyna moral theory and its account of the nature of soteriological practice.

18. In this visible world,
All actions of the two kinds,
Each comprising action and the unexpired separately,
Will remain while ripening.


But here the opponent slides over into the substantialism that Nagarjuna will criticize. For although he has characterized actions as impermanent, he has retained the seed-and-sprout metaphor that has the actions identifiable over time and, hence, as having an independent existence and identity. Moreover, he suggests, their consequences are determinate in time, delimited by death or nirva:

19. That fruit, if extinction or death
Occurs, ceases.
Regarding this, a distinction between the stainless
And the stained is drawn.


Ngrjuna now mounts a reply against all of these positions collectively:

20. Emptiness and nonannihilation;
Cyclic existence and nonpermanence:
That action is nonexpiring
Is taught by the Buddha.


All phenomena, including action, its result, and the connection between them, will come out to be empty of inherent existence, yet conventionally real; they will be part of cyclic existence, but will be impermanent. This is not surprising. But Nagarjuna also says that no action expires (retaining the promissory note metaphor). Obviously, he cannot mean that actions are permanent. Rather, we should understand this to assert two related theses: First, it indicates that the consequences of actions do not cease at some point. All actions have ramifications into the indefinite future, due to dependent arising. Second, actions themselves, being empty of inherent existence are not entities capable of passing out of existence, when passing out of existence is interpreted to mean the cessation entirely of something that once existed inherently. Since actions are not inherently existent, they are not suitable bases for inherent cessation. And this resolves the final apparent paradox: The tension between the assertion that nothing is permanent and that all action is nonexpiring. All phenomena are indeed impermanent, but that entails both that they do not inherently cease and that their effects are indefinite in scope.

21. Because action does not arise,
It is seen to be without essence.
Because it is not arisen,
It follows that it is nonexpiring.


This verse emphasizes the second reading of the thesis of the nonexpiration of action and echoes the arguments from Chapter VII.

22. If action had an essence,
It would, without doubt, be eternal.
Action would be uncreated.
Because there can be no creation of what is eternal.


Moreover, Nagarjuna reminds us, again drawing heavily on the arguments reviewed and redeployed in Chapter VII, things with essences don’t arise and cease, and can’t be related causally to other things. If action existed inherently, it couldn’t be initiated. So, if one were trying to preserve the reality of action and karma against the analysis in terms of emptiness (because one viewed that analysis as undermining their genuine existence), it would be pointless to defend the existence of action and karma as inherent existence.

23. If an action were uncreated,
Fear would arise of encountering something not done.
And the error of not preserving
One’s vows would arise.


Nagarjuna here and in XVII: 24 draws some of the moral consequences of the nihilistic view of action that seems to follow from the conditions set on its existence by the reificationist: Actions would not come into being through agency and so would have no regular relation to any agents. And so one might find oneself experiencing the consequences of some action one had not performed, or find that it was, in some sense, one’s own action. One would not take action seriously as one’s own responsibility and would not worry about moral infractions. Monks and nuns would break their vows. Since morality depends on a distinction between morally positive and morally negative acts, if there were no actions, or if actions could not be thought of as initiated by their agents, there would be no morality. From another perspective, the preservation of vows would be an impossibility anyway since preserving the vows requires taking action, which would be impossible if action were uncreated.

24. All conventions would then
Be contradicted, without doubt.
It would be impossible to draw a distinction
Between virtue and evil.


Moreover, Nagarjuna argues in the next verse, if actions had essences, they could not cease, and if their karmic consequences had essences since they would need no conditions to arise, they would just keep arising:

25. Whatever is mature would mature
Time and time again.
If there were essence, this would follow,
Because action would remain in place.


26. While this action has affliction as its nature
This affliction is not real in itself.
If affliction is not in itself,
How can action be real in itself?


Moreover, Nagarjuna continues, afflicted action is, for the opponent, done essentially in affliction. But given that affliction has already been shown to be empty in the chapter on suffering (XII), how could it serve as an essence for action?

27. Action and affliction
Are taught to be the conditions that produce bodies.
If action and affliction
Are empty, what would one say about bodies?


The opponent replies, however, that action and affliction are referred to in stras as the causes of different kinds of rebirth and of different characteristics in rebirths. And since beings are indeed reborn and do indeed have characteristics, how, from the standpoint of a Buddhist view of rebirth, could empty actions and empty karmic consequences explain this?

28. Obstructed by ignorance,
And consumed by passion, the experiencer
Is neither different from the agent
Nor identical with it.


Nagarjuna focuses in his reply on the nature of the individual who is the putative agent of these actions and experiencer of their consequences. The present objection rests on the presupposition that they exist inherently. That is why the problem arises about how empty actions and empty karmic links could be sufficient to link their properties. So Nagarjuna emphasizes that neither an analysis in terms of inherent identity nor one in terms of inherent difference between agent and action will suffice. Both presupposes, incoherently, the inherent existence and hence the possession of an essence, of each term in the putative relation. But this of course recalls the problem posed near the end of Chapter I: How can actual effects arise from empty conditions? And Nagarjuna’s reply echoes the reply developed there:

29. Since this action
Is not arisen from a condition,
Nor arisen causelessly,
It follows that there is no agent.


Since the action does not arise inherently, it lacks inherent existence. Since, as per the discussion of agent and action in Chapter VIII, empty actions entail empty agents, there is no inherently existing agent of the kind presupposed by the objector. But the objector continues:

30. If there is no action and agent,
Where could the fruit of action be?
Without a fruit,
Where is there an experiencer?


That is, if we deny the reality of the action and the agent, we seem to deny the reality of the consequences of the action and, hence, the experiencer, whether “without understanding and consumed by passion” or not. But Nagarjuna’s view is not that these things are non-existent, as he emphasized in XVII: 20—only that they are empty. So it does follow that the consequences are empty—but that does not entail in any way that they are nonexistent. And it follows that the consequence and the karmic link are empty. From this it follows that the reborn individual whose existence and characteristics are determined by this causal sequence is also empty of inherent existence. And if so, there is no problem about how his/her genesis is dependent upon an empty sequence. Nagarjuna introduces an analogy to explain this situation:

31. Just as the teacher, by magic,
Makes a magical illusion, and
By that illusion
Another illusion is created,


32. In that way are an agent and his action:
The agent is like the illusion.
The action
Is like the illusion’s illusion.


That is, we can understand the entire sequence of agent, action, consequences of action, and arising of new agent, whether within a single lifetime or—in the context of Buddhist ont����ology and doctrine—across lifetimes, as an entirely empty sequence with entirely empty stages. But that does not prevent its being perceived, or its reality for those who participate therein.

33. Afflictions, actions, bodies,
Agents, and fruits are
Like a city of Gandharvas and
Like a mirage or a dream.


Again, it is important to emphasize that emptiness, rather than being a kind of non actuality contrasting with empirical reality, is in fact the very condition of empirical reality and hence the only kind of genuine actuality. Mirages and dreams are actual phenomena, which actually appear and which have consequences. But that does not mean that they appear to us in a nondeceptive way. Mirages are not water and do not quench thirst, and dream-elephants carry no loads. By analogy, sasra, action, karmic link, and consequence, Nagarjuna argues, are real empirical phenomena, but are empty of anything more than conventional existence. While they may appear to exist inherently, either as persistent phenomena, as processes or elements of processes, or as abstract phenomena—as per the various opposing views considered in this chapter—they do not so exist. For to exist in those ways would in fact be incoherent. This analysis hence does not entail the nonexistence of agent and action, except from the ultimate point of view. Rather it explains how it is possible for them to exist at all.


Chapter XVIII

Examination of Self and Entities


A good deal of the confusion Nagarjuna diagnoses in the previous two chapters concerns the presupposition that the self, as an afflicted being capable of liberation from suffering, must be thought of as an inherently real entity. In this chapter, therefore, Nagarjuna turns to an examination of the self, per se, apart from its relation to such things as perception, action, suffering, affliction, and so forth, as he has examined it in prior chapters.

1. If the self were the aggregates,
It would have arising and ceasing (as properties).
If it were different from the aggregates,
It would not have the characteristics of the aggregates.


If there is an inherently existent self, it must either be identical to or different from the aggregates. The aggregates are the more basic components into which the individual divides upon analysis. In standard Buddhist analysis, they include the