Paradox of Causality in Mādhyamika
By David Loy
International Philosophical Quarterly v25 n1. p63-72
Copyright 1985 by Fordham University Press
PROBLEM of causality is central to all schools of Buddhism, and this is
especially true of
Mādhyamika. But at first glance there seems to be a contradiction in the
Mādhyamika analysis. On the one hand, causal interdependence is clearly a
crucial concept, so important that Nāgārjuna identifies it with the most
important concept, śunyatā: "We interpret the dependent arising of all things (pratītyasamutpāda)
as the absence of being in them (śunyatā)." 
This emphasis on interdependence develops to completion the early
Buddhist doctrine of impermanence: there are no unconditioned elements of
existence (dharmas), for all things arise and pass away according to conditions.
The undeniable relativity of everything is the means by which self-existence (svabhāva)
At the same time, Nāgārjuna redefines pratītyasamutpāda in such a way as to
negate causality altogether. This is apparent even in the prefatory dedication
of the Mūlamādhyamikakārikās,
in the eight negations which Nāgārjuna attributes to the Buddha:
perishing nor arising in time,
terminable nor eternal,
self-identical nor variant in
coming nor going;
pratītyasamutpāda.. . .
with this, the first
and most important chapter of the Kārikās concludes that the causal relation is
inexplicable, and later chapters go further to claim that causation is like māyā.
"Origination, existence, and destruction are of the nature of māyā, dreams, or a
fairy castle." 
The last chapters seize on this issue as one way to crystallize the difference
and nirvāna. The nirvāna chapter distinguishes between them by attributing
causal relations only to saṁsāra:
which, taken as causal or dependent, is the process of being born and passing
on, is, taken non-causally
and beyond all dependence, declared to be nirvāna."  In
his commentary on the previous chapter, Candrakirti
and duḥkha in
the same way: ".
dependent in existence, that is, for things to be based on
(hereafter "MMK') XXIV 18, in Lucid Exposition of the Middle Way: The Essential
Chapters from [he Prasannapada of andrakirti, trans. Mervyn Sprung (Boulder:
Press, 1979), p. 238. This identification must be kept in mind to avoid
error of misinterpreting sunyaia as non-being and making Mādhyamika into a
I bid., pp. 32-33, 35. For
pratītyasamutpāda Sprung gives "the true way of things."
MMK VII 34, as quoted in T. R. V. Murti, The Central Philosophy of Buddhism
(London: Allen and Unwin, 1960), p. 177.
MMK XXV 9, in Sprung, p. 255. In my
opinion this is the most important verse in the Kārikās.
each other in utter reciprocity, is saṁvṛti." ". . it is precisely what arises in dependence that constitutes duḥkha, not what does not arise in dependence." 
How are we to understand this obvious contradiction? That is, how do we get from interpreting pratītyasamutpāda as "dependent origination" to "non-dependent non-origination," and, what is more, reconcile the two? Explanations which differentiate between two different types of sutras (exoteric neyārtha and esoteric nītārtha), or which refer to the two-truths theory, just raise the same question in a different form, for we need to know how such contradictory truths can both be true, i.e., how these two are related to each other.
This paper will offer one way (perhaps not the only way) of resolving this problem, which may more properly be said to be a paradox. I shall argue that complete conditionality is phenomenologically equivalent to a denial of all causal conditions. That is, a view which is so radical as to analyze things away into "their" conditions is offering an interpretation of experience which becomes indistinguishable from a view that negates causality altogether. If this is true, we have another instance where it becomes very difficult to distinguish the Mādhyamika nonduality from that of Śaṅkara's Advaita Vedānta.  The argument will be made in two steps. We shall see that there is a dialectic inherent in the Mādhyamika analysis. The first stage (discussed in Part I) is apparent: looking at the common-sense distinction between things and their cause-and-effect relationships, Nāgārjuna uses the latter to "dissolve" the former and deny that there are things. Less obvious is the second stage (Part II), which reverses the analysis: as we shall see, the lack of "thingness" in things implies a way of experiencing in which there is no awareness of cause and effect. Things and their causal relations stand and fall together, because our notion of cause-and-effect is dependent on that of things, which cause and are affected. If one collapses, so does the other. The basic problem is that any dualism between them is untenable. It is the delusive bifurcation between them that Nāgārjuna is concerned to negate, and the two ways to do this are to use each pole to "deconstruct" the other. Consistent with the general Mādhyamika project, this is criticism without affirming any philosophical position: having conflated the duality, Nāgārjuna does not offer any view about causality, because nothing remains to be related to anything else. Part III compares this conclusion with the Advaitic position regarding causality.
In order to understand the Mādhyamika critique, we must begin with a clear sense of what it is that is being criticized. This is our common-sense understanding of the world, which sees it as a collection of discrete entities (including myself) interacting causally "in" space and time. Just as space and time, if they are to function as "containers," require something understood as nonspatial and nontemporal to "contain," so the causal relation is normally used to explain the interaction between things which
 Sprung,pp. 230, 236.
I have argued for this equivalence in two other papers. "Enlightenment in
Buddhism and Advaita Vedanta: Are Nirvāna and Moksha the Same?"
International Philosophical Quarterly, 22 (1982), 65-74, claims that the no-self
modal-view of Buddhism is indistinguishable from the all-Self substance-view of
Advaita. See my paper "The Mahayana Deconstruction of Time"
(unpublished) which maintains that the nonduality between things and time
amounts to making the same claim regarding temporality: if there is only time,
this is phenemenologically equivalent to a nunc stans ("eternal now").
are distinct from each other. Nāgārjuna attacks more than the philosophical
fancies of Indian metaphysicians, for there is a metaphysics inherent in our
common-sense view. This common-sense understanding (one or the other aspect of
which is absolutized in systematic metaphysics) is what makes the everyday world
saṁsara for us, and it is this saṁsara that Nāgārjuna is concerned to
"deconstruct." This is why one must beware of making Mādhyamika into an
"ordinary language" philosophy by interpreting
merely as a "meta-system" term denying a correspondence theory of truth. By no
means does the end of philosophical language-games "leave everything as it is"
for Nāgārjuna, except in the sense that saṁsara has always really been nirvāṇa.
Why do we experience
the world as saṁsara,
if that is delusive? Why don't we experience it as it is? The traditional
Buddhist answer points to craving and ignorance, but
Nāgārjuna focuses on a particular type of mental attachment, that which makes
all other attachment possible: prapañca.
chapter of the Kārikās concludes by characterizing nirvāṇa,
negatively, as "the coming to rest of all ways of taking things, the repose
of named things (prapañcopaśama)The
precise meaning of prapañca
is, unfortunately, unclear. Sprung defines it as "the world of named
This point about prapañca
is important because without it one might conclude that Nāgārjuna's critique of
self-existence (svabhāva) is a refutation of something that no one believes in
anyway. But one does not escape his critique by defining entities in a more
common-sense fashion as coming-into and passing-out-of existence. There is no
tenable middle ground between self-existence independent of all conditions—an
empty set, for there are no such entities—and the complete conditionality of
The implication of Nāgārjuna's arguments against self-existence (e.g., MMK
chapters I, XV) is to point out the inconsistency in our everyday way of
"taking" the world: we accept that things change, and at the same time we assume
that somehow they also remain the same—which is necessary if there are to be
"things" at all. Other philosophers, recognizing this inconsistency, have tried
to solve it by absolutizing one of these at the expense of the other; so the
satkāryavāda substance-view of Samkhya emphasizes permanence at the price of not
being able to account for change, and the asatkāryavāda modal-view of early
Buddhism has the opposite problem of not being able to account for continuity.
Nāgārjuna arranges these and the other solutions that have been proposed into a "tetralemma"
which exhausts the possible philosophical alternatives; then he proceeds to
refute them all. The basic difficulty
is that any understanding of cause and effect which tries to relate two separate
things together can be reduced to the contradiction of both asserting and
denying identity. Nor can one respond to this simply by denying causality, for
that is likewise contradicted by our experience. So
Nāgārjuna concludes that the "relationship" between cause and effect is
''MMKXXV 24, in Sprung p. 264.
The role of prapañca
in our experience cannot be fully understood without relating it to our
intentions. The relations among craving, conceptualizing, and causality have
been discussed in "The Difference between saṁsara
Philosophy East and West, 33 (1983), 355-365.
the problem is not resolved simply by criticizing such positions, for the difficulty
is fundamentally not abstract and philosophical but very personal: it is our
lives, not just our theories, which are inconsistent in "taking"
the world as a collection of discrete "self-existing"
things which yet change. Although during, a
philosophy seminar I may accept the complete conditionality
and contingency of all experience, as soon as the seminar ends I unconsciously
assume that the colleague I join for lunch is the same person whom I spoke with
before the seminar—although ever-so-slightly
different, due to a relatively extraneous change of mood, etc. This constitutes saṁsara
because it is by reifying such "thingness"
out of the flux of experience that we become attached to things. (Of course,
other hypostatizations of self-existence—
my wife, my car and, most of all, myself—tend to be more problematic loci for
attachment, but the problem is the same in each case: we cling to things which
dissolve as we try to grasp them.)
does not suffice to answer this Humean
critique of identity  with an
"ordinary language" rejoinder that
we should become more sensitive to the ways we use our permanence-and-change
vocabulary, for the
position is that our usual everyday experience is
deluded and this ordinary use of language is deluding. As the first
prong of his attack,
Nāgārjuna refutes our common-sense distinction between things and their causal
relations simply by sharpening the distinction to absurdity: if things are to be
"self-existent" then they must be separable from their conditions, but their
existence is clearly contingent upon the conditions that bring them into being
and eventually (when those conditions no longer operate) cause them to
disappear. If it is objected that one cannot live without reifying such
fictitious entities, at least to some extent, then
response is to agree. The
"lower truth", saṁvṛti,
is not negated altogether—it is a truth—but it must not be taken as
"the higher truth," as a correct understanding of the way things
the first stage of the
critique uses the complete interdependence of
all things to refute their "thingness." The distinction between things
and their causal relations is negated by adopting the latter as a means to
"deconstruct" the former. This completes the early Buddhist attack on
substance. But this absolutizing of conditions
is only the first step.
Now the critique dialectically reverses to
use the perspective of the deconstructed thing in order to deny the reality
of causal conditions. In the delusive bifurcation between things and their
causal relations, the category of causality turns out to be just as
dependent on things as things are on their causal
conditions. Our concept of causality presupposes a set of discrete entities,
whose interrelation we explain as causation. Cause-and-effect
requires something to cause and something to
be effected. If this is so, then a complete conditionality which is so radical
that it "dissolves" all things must also dissolve itself.
order to make this point, it is helpful to transpose the argument from the
too-general category of causal conditions to the more specific one of
motion-and-rest. Nāgārjuna analyzes motion and rest in chapter two of the Kārikās,
immediately after his initial treatment of conditions, and it is clear that the
second chapter is meant to apply the general conclusions of chapter one to a
particular case. The additional advantage of shifting to motion-and-rest is
that we illuminate what is otherwise a very puzzling chapter. The basic problem
is that it is not clear what
is actually doing in chapter two. Like Zeno,
he denies the reality of motion, but this is not done
See Hume's Enquiry
Concerning Human Understanding, Section IV, Parts I and II.
assert an unchanging Parmenidean
"block-universe," for Vedantic
permanence (i.e., "rest") is also denied. As a result,
Nāgārjuna has been criticized for an arid play on words which "resembles the
shell game" in being a logical sleight-of-hand" —
that is, for basing his argument on a subtle distinction between words which are
meaningless because they have no empirical referent—and
for committing the fallacy of composition in arguing that what is true for the
parts (in this case, traversed, traversing, and to-be-traversed) must be
true for the whole.  But I think
such criticisms miss the point of Nāgārjuna's arguments. Their import is that
our usual way of understanding motion, which distinguishes the "mover" from "the act of moving,"
simply does not make sense, because the interdependence of "mover" and
"motion" shows that the hypostatization
of either is delusive. (Here "mover" means, not "that which
causes motion", but "that thing which moves"; and
"motion" is "the movement that happens to the thing")
Nāgārjuna's logic in stanzas two through eleven demonstrates that once we have
reified a distinction between them, it becomes impossible to relate them back
together again—a quandary familiar to students of the Western mind-body problem,
which is the result of another reified bifurcation. The difficulty is shown by
isolating this hypostatized mover and inquiring into its status.
asks: In itself, is it a mover, or not? That is, is the predicate
"moves" intrinsic to this mover, or contingent? The dilemma is that
neither way of understanding the situation is satisfactory. If the
"mover" in and of itself already
moves, then there is no need to add an "act of motion" later; the predication of
such a "second motion" would be redundant. But the other alternative—that the
"mover" by itself is a non-mover—does not work either because we cannot
thereafter add the predicate: it is a contradiction for a non-mover to move. In
neither way can we make sense out of the relation between them. It follows that
the "mover" cannot have an existence of its own apart from the "moving," which
means that our usual dualistic way of understanding
motion is untenable. To summarize this in contemporary terms,
is pointing out a flaw in the everyday language we use to describe (and
hence our ways of thinking about) motion: our ascription of motion predicates to
substantive objects is unintelligible.
first encounter the above argument may be
unconvincing. The options seem so extreme that we suspect there must be some
middle ground between them. Of course we can't accept a double movement, but is
it really a contradiction for a non-mover to move? What else could
move? But, again, no such appeal to our everyday intuitions, or to the ordinary
language which shapes and embodies them, is successful against the
Mādhyamika critique of those intuitions, which seizes on the
inconsistency that is ignored (and to some extent must be ignored) in daily
life. We can elaborate upon this by applying the logic that was used earlier to
deconstruct the difference between things and their causal relations. Just as
(general rule) complete interdependence dissolves the thing into its rational
conditions, with no residue of substance remaining, so (specific
case) the completeness of movement—the fact that no part of "me"
stays unmoved in the chair when "I" go to lunch—means that no
self-existing and hence unchanging "thing" remains to move. As with
time and with space, we think of the relation between mover and moving with the
metaphor of container and contained, and in all three instances the bifurcation
is delusive. In order to expose the absurdity
Nāgārjuna needs only to sharpen the dichotomy. Despite our intuitions, which
 "Richard Robinson, "Did Nāgārjuna Really Refute All Philosophical Views?"
Philosophy East and West, 22 (1972), 325.
Cheng, "Motion and Rest
in the Middle Treatise," J. of Chinese
Philosophy, 1 (1980) 235 ff.
want to postulate some "unchanging core" in order to save the mover,
there is no middle ground between a self-existent and hence unmoving
thing, and the complete dissolution of the thing that does the moving.
Understood in this way, it becomes obvious why his arguments also work just as
well against the intelligibility of rest: the bifurcation between the thing and
its being at rest is just as delusive, and for precisely the same reason.
far, we have effected only the first stage of
the dialectic, both in the general analysis of causal conditions and this
more specific case of
motion-and-rest. We have dissolved the thing which moves / is caused
and are left with a changing world of causal conditions. The second stage of the
dialectic is easy to state but harder to understand:
nonetheless, unless we can get a sense of what such a way of experiencing would
be like, the above argument will be at best philosophically persuasive yet will
seem irrelevant to daily life. What consequences does all this have for the way
we actually experience the world? In particular, it is still unclear how,
except by some "logical sleight of hand," all-conditionality
can be phenomenologically identified
with no-conditionality, as we claimed at the
beginning of this paper.
me try to satisfy these questions with the help of a well-known Ch'an
(Zen) story. The following
example discusses the causal relations of a physical action, but what is said
may be applied just as well to sense-perception (e.g., to the nondual sound of a pebble striking bamboo, which awakened Hsiang-yen,
or to the nondual pain when Yün-mên broke
his ankle) and to thought (Hui Neng:
"If we allow our thoughts, past, present and future, to link up in a series, we
put ourselves under restraint. On the other hand, if we never let our mind
attach to anything, we shall gain liberation.").
was a monk in the monastery of Huang-po.
Three times Lin-chi asked the Master:
"What is the real meaning of Bodhidharma
coming from the West?" and each time Huang-po immediately struck him.
Thereupon, discouraged, he decided to leave
direct relevance of Ch'an experience to this issue cannot be questioned. While
it is true that Ch'an is not. and does not have, any philosophy, yet it is also
the case that Mādhyamika, as a philosophical exposition of the Prajñāparamīta,
may be said to be the philosophy that most made Ch'an possible.
was advised to go to Master Ta-yü. Arriving at his monastery, Lin-chi told Ta-yü of his encounters with Huang-po,
adding that he didn't know where he was at fault.
exclaimed; "Your master treated you entirely with grandmotherly kindness, and
yet you say you don't know your fault." Hearing this, Lin-chi was suddenly
awakened and said; "After all, there isn't much in Huang-po's Buddhism!" 
did Lin-chi realize that awakened him? If (rushing in where Ch'an
masters will not tread) we distort his experience into an idea in order to gloss
this story, we may say that Lin-chi must have realized that Huang-po had been
answering his question. The blows he received were not punishment but a
demonstration of why Bodhidharma came from
the West. On the common-sense level, the answer to Lin-chi's question is obvious: Bodhidharma was
bringing Buddhism to China. But this is a relative, "lower truth"
explanation. As a standard Ch'an question designed to initiate a dialogue, it
goes without saying that what is sought is the "higher truth," and on
this level there is no "why." For the enlightened person, each
experience is complete in itself, the only thing in the whole universe; for each
action is tathata. Nothing changes because
without prapañca-reification everything is
perceived afresh, for the first time. As Bodhidharma walked from India there was
no thought of "why" in his head; "he" was each step. In the
same way, there was no "why" to Huang-po's
The paradox which makes the above story relevant to this paper is the fact that, at
the same time, Bodhidharma's and Huang-po's
actions are intentional. Huang-po's blow may be immediate and spontaneous, but
there is also a "reason" for it; it is not a random or irrelevant
gesture, but a very appropriate response to that particular question, drawn
forth by that situation. If we translate this point about intention back into
our more general category of causality, here we have a case of an act which is
both completely caused (perfect upāya: glove
fitting hand tightly, to use the Ch'an
analogy) and yet is also uncaused. This paradox is a contradiction only
according to our usual understanding of causality, which uses that category-of-thought
to relate together the supposedly discrete objects into which prapañca
carves the world. The first and most
important of these hypostatized
"things" is me, the subject who
craves some of these objects and thus needs an understanding of cause-and-effect
relationships in order to manipulate circumstances and obtain them. (It has been
argued that the desire for such manipulation is the very root of our concept of
) This would apply to
the story in question if Huang-po, prapañca-deluded,
were to perceive Lin-chi dualistically:
Lin-chi is sitting there, a person-object that needs to be enlightened, and I,
version of the story, from the Transmission of the Lamp, is given in Chang
Chung-yuan, Original Teachings of Ch'an
Buddhism (New York; Vintage, 1971), pp.
idea of cause has its roots in purposive
activity and is employed in the first
instance when we are concerned to produce or to
prevent something. To discover the cause of
something is to discover what has to be
attested by our activity in order to produce or to
prevent that thing; but once the word 'cause'
comes to be applied to natural events, the
notion of altering the course of events lends to be dropped. 'Cause'
is then used in a nonpractical, purely diagnostic way in cases where we have no
interest in altering events or power to alter them." (P. H. Nowell-Smith, "Causality or
Causation." I have a cyclostyled copy of
this article but have not been able to trace its source.)
sitting here, am the person who will try to enlighten him. Then "my"
blow is reified into a deliberated effect
which I hope will cause Lin-chi's awakening.
if, as all schools of Buddhism agree, there
is no self to do this causal-relating-between-things,
then the above understanding of the situation must be delusive. So Huang-po
must have experienced it differently, and causality too must be understood
differently. It is not denied: on the contrary, without the sense of self and
other prapañca- reified objects
as a counterfoil, it expands to include everything, as Nagar-juna has already
shown. (So the doctrine of karma can be understood simply by applying something
like Newton's third law of physical motion to the mental realm as well.) From
the perspective of Mādhyamika's "all-conditionality"
which deconstructs all things, Huang-po's
blow is part of a seamless web of conditions which can be extended, as Hwa
Yen does, to encompass the entire universe. As one jewel in the infinite
web of lndra, the blow reflects everything
everywhere, at all times. But if every event that happens is interdependent with
everything else in the whole universe, what a different way of experiencing this
involves! It suggests a Spinozistic
acceptance of whatever happens, as a product of the whole, but more than this it
implies the irrelevance of causality as usually understood.
"All-conditionality," in its complete negation of anything to be
attached to, offers no practical utility, because there is no longer any object
to be obtained or any self that craves it; whereas a self that wants to obtain
some thing will need to isolate some discrete object or action as the cause
which leads to obtaining it.
does all this imply about the way Huang-po experienced his own action? Because
he did not perceive the situation dualistically,
the action was not "his." That the blow was appropriate to the
situation was not due to any prior deliberation, however quick. On the contrary,
the action was so appropriate precisely because it was not deliberated, just as
the best responses in "dharma-combat"
are unmediated by any self-conscious
"hindrance in the mind." Then why did Huang-po strike, rather than
shout "ho!" as Ma-tsu often did, or
utter a few soft words, as Chao-chou probably
would have done? This is the crucial point: He
does not know and cannot know.
("Not-knowing is very profound" said Master Lo-han,
precipitating Wên-i's awakening.) His
spontaneous actions are traceless, "like
the tracks of a bird in the sky." 
respond to a situation like a glove
fits on a hand because whatever
"decisions are made" (if this phrase can be used here) are not made by
him. If one nondualistically is the cause (or effect, or
both), rather than being a hypostatized self
that dualistically uses it, then there is not the awareness that it is a
cause (or effect, or both);
Tung-shan told his students to walk "in the bird's
track," which is of course trackless,
having no deliberative
traces before ("Should I do this or that?")
and leaving none after ("Should I have done that?").
For further discussion of the relations among action, intention, and nonduality,
deep—or (what amounts to the same thing) from nowhere.
In this sense Mahayana is not wrong to
śunyatā with the Absolute. 
At first glance, the Advaitic account of causality is very different from the Mādhyamika conclusions and Ch'an experience which have been discussed above. Śaṅkara's position regarding causality constitutes part of his more general māyā doctrine, according to which all phenomena are the indescribable and indefinable ajñana which is superimposed (adhyāsa) upon Brahman. But if we delve beneath the surface of terminology and ask what experience this describes, it becomes difficult to find any phenomenological basis for the distinction between the Mādhyamika and Advaitic accounts.
this similarity should not be
surprising, since Śaṅkara's dialectic was clearly influenced by Nāgārjuna's. In
this regard, it is relevant to compare Śaṅkara's careful critique of Vijñānavāda
with his cursory dismissal of Śunyavāda as nihilistic and unworthy of
repudiation.  Śaṅkara's main
treatment of causality, in Brahma-Sūtra-Bhāsya II.i.14-20, is indebted to the
Mādhyamika dialectic and reaches a similar conclusion, that we cannot derive the
real nature of causal relations from the series of discrete cause-and-effect
phenomena. As a Vedantin, Śaṅkara then leaps to the conclusion that the true
cause of all effects must be Brahman, which provides the permanent substratum
that persists unchanged through all experience. All effect-phenomena are merely
illusory name-and-form superimpositions upon Brahman, the substance-ground.
Since Brahman is the only real (svabhāva), and phenomena existing as distinct from it are illusory, this is a
of satkāryavāda: the
effect pre-exists in the cause. But to distinguish this view from that of
Sāṁkhya, which identifies cause and effect by granting the reality of prakṛti,
Śaṅkara's theory of causality is more precisely labelled vivartavāda, since the
effect (māyā) has a different kind of being from the cause (Brahman).
Expressed in this way, the views of Nāgārjuna and
Śaṅkara seem diametrically op-
This view of Mādhyamika is important
for understanding the trisvabhāva doctrine of Yogācāra Buddhism. The prapañca
world of discrete forms corresponds to parikalpita-svabhāva,
the "imagined nature." "All-conditionality"
corresponds to parinispanna-svabhāva, the
"other-dependent nature." "No-conditionality"
corresponds to parinispanna-svabhāva, the
absolutely-accomplished nondual nature. Read
in this way, Vasubandhu's Trisvabhāvanirdeśa,
for example, is completely consistent with the
Mādhyamika analysis of experience. For both Mādhyamika and Yogācāra, an
understanding of "all-conditionality," with its negation of the self-existence
of discrete things, is the crucial "hinge" by which we turn from avidyā
I think that the
Mādhyamika view of causality—a dialectic which equates complete conditionality
with no-conditionality—also implies a critique of Derrida's deconstruction.
Derrida's use of the open-endedness (différance)
of texts to deconstruct the self-as-writer employs only the First movement of
the dialectic; the second and reverse movement (which Derrida does not make)
uses the lack of a self to deconstruct the dissemination of meaning. One ends up
with something more like the presence of the late Heidegger, where language is
realized to be "the house of Being." The same point can be made by comparing the
Mādhyamika critique of temporal relations with Derrida's critique of "logocentrism."
is not unlikely that Śankara
discovered his own non-dual philosophy in the system of Nāgārjuna and left it
unexplained. His debt to
Śunyata doctrine was so great that he quietly passed
over it." Lal Mani Joshi, Studies in the Buddhistic Culture of India (Delhi:
Motilal Banarsidass, 1977), p. 234.
William Blake: "And every Natural Effect
has a Spiritual Cause, and Not a Natural; for
a Natural Cause only seems: it is a delusion. .. ." (Milton, plate 28, 44-45)
posed. They draw opposite conclusions from the illogicality and
unintelligibility of the causal relationship: Nāgārjuna, as a Buddhist, denies
any substratum-ground to phenomena, and leaves them empty (śunya)
of any Being; Śaṅkara, as a Vedantin, postulates Brahman as an imperceptible but
necessary substratum. Phenomenologically, however, these positions turn out to
be equivalent. For Advaita, Brahman is the real cause of all phenomena, but, in
denying the reality of all changing attributes, Śaṅkara is reduced to defining
the substratum so narrowly that it ceases to have any referent. Absolutely
nothing can be predicated of Nirguna Brahman, and it can be approached only
through the via negatives of neti, neti. Although Śaṅkara would deny it, Brahman
ends up as a completely empty ground,
unchanging only because it is a Nothing from which all phenomena arise as an
ever-changing and hence deceptive appearance. From the perspective of Buddhism,
Vedanta reifies śunyatā into an attributeless substance, which, since it has no
characteristics of its own, cannot really be said to be at all. From the
perspective of Vedanta, however. Buddhism ignores the fact that such a ground is
necessary, for as Parmenides pointed out nothing can arise from nothing and it is meaningless to deny all
substance: something must be real. Despite this family quarrel, the descriptions
converge; what is perhaps more important than the difference is that for both
the emptiness of this "ground"
(however otherwise understood) is also fullness and limitless richness, for it
is lack of any fixed characteristics which makes possible the infinite
diversity of the phenomena which arise from "it" 
What is most significant about their argument is that it is no longer a
disagreement over the nature of the nondual experience. Since Brahman is
qualityless and imperceptible, there is no phenomenological difference between a
Mādhyamika interpretation of Huang-po's blow and an Advaitic one. In both cases, the
arm-movement is experienced nondually,
with no bifurcation
between a self-conscious subject and "his" action. In both cases that action is
mysterious māyā, inexplicable in terms of efficient
causality and having no svabhāva reality of
its own (nor, of course, does Huang-po, or
anything else). The importance of this agreement is great. The only difference
stops here, while Advaita assumes that there must be an
unchanging ground as the source of all the changing phenomena. But since this
source is by definition imperceptible,
the difference is reduced to a far more abstract, although not trivial, one of
emphasis: concluding that phenomena are illusory māyā
seems to devalue them
somewhat more than if phenomena are merely
śunya without any Brahman
"behind" them. In other words, the difference becomes one of attitude
towards the nondual experience rather than
anything in the experience itself: the Advaitin,
with his dualistic distinction between
Brahman and māyā
will be more eager to negate
the phenomenal world than the Buddhist bodhisattva,
for whom there are only empty forms. 
is reflected in the etymology of both words. Most scholars agree that Brahman
comes from the root brḥ, "to burst forth, grow." "To us, it is clear. Brahman
means reality, which grows, breathes or swells." Radhakrishnan, Indian
Philosophy (London: Alien and Unwin, 1962J, I, 164 n.
Śunyatā is from the root
śu, which means "to swell" in two senses: not only "hollow or empty," but also
"to be swollen" in the sense of full, like the womb of a pregnant
woman. It has been unfortunate for Buddhist
studies that the English translation "emptiness" captures only the first
am grateful to the "Singapore Mādhyamika Study Group", especially Peter Della Santina, for their comments on an earlier draft of this paper. For further
discussion of Mādhyamika, see "How Not to Criticize Nāgārjuna," Philosophy East
and West, 34 (Oct. 1984).