You are on page 1of 24

BETWEEN NIHILISM AND ANTI-ESSENTIALISM: A CONCEPTUALIST INTERPRETATION OF

NĀGĀRJUNA
Author(s): John Spackman
Source: Philosophy East and West , JANUARY 2014, Vol. 64, No. 1 (JANUARY 2014), pp.
151-173
Published by: University of Hawai'i Press

Stable URL: https://www.jstor.org/stable/43285884

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

University of Hawai'i Press is collaborating with JSTOR to digitize, preserve and extend access
to Philosophy East and West

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
BETWEEN NIHILISM AND ANTI-ESSENTIAL I SM:

A CONCEPTUA LI ST INTERPRETATION OF NĀGĀRJUNA

John Spackman
Department of Philosophy, Middlebury College
jspackma@middlebury.edu

Introduction

Rumors of the death of the nihilist interpretation of Nāgārjuna are, it appears, g


exaggerated. The nihilist reading is the radical view that the central c
Nāgārjuna's writings is that nothing whatsoever exists in any sense, either
level of the ultimate truth (paramärthasatya) or at the level of the conventional t
(samvrtisatya). Such a view is to be contrasted with what is arguably the pre
contemporary account - or better, family of accounts - of Nāgārjuna, whic
call the anti-essential ist interpretation, some form of which is maintained,
stance, by Jay Garfield, David Seyfort Ruegg, David Kalupahana, and Fr
Streng.1 Though the views of these authors differ considerably in important resp
what they have broadly in common is the idea that what Nāgārjuna rejects is
existence of things altogether, but only the existence of things as svabhãva,
pendently or intrinsically existent entities. In this view, while Nāgārjuna re
existence of things at the ultimate level - their existence as svabhãva - he do
that they exist conventionally, where this conventional existence is conceiv
different way of existing, namely existing dependently on other things and on h
knowers. The nihilist and the anti-essentialist thus agree that for Nāgārjuna ultim
nothing exists; where they differ is on the question of whether things may be sa
exist conventionally.
The nihilist reading has been summarily dismissed by most recent interpr
of Nāgārjuna. And yet there has been a perennial attraction toward it among
of Nāgārjuna, both in ancient India and in the modern West. Among contem
interpretations, in addition to Thomas Wood's detailed and provocative def
the nihilist reading in his Nāgārjunian Disputations , recent years have also s
publication of David Burton's Emptiness Appraised, which argues that Nāgā
views are nihilistic even if he didn't think so.2 Several other interpreters, even if
do not advocate the nihilist account, have worried that it poses an unackno
threat to the anti-essentialist reading.3
There are, I would suggest, two main factors that pull interpreters tow
nihilist account. The first is textual evidence from the Mūlamadhyamakakār
Nāgārjuna's other works that seems to support the nihilist reading. At a num
points in his works Nāgārjuna makes statements that seem, prima facie, to p
that whatever exists must either exist intrinsically or depend on somethin
that exists intrinsically - in Nāgārjuna's terms, that whatever exists must b

Philosophy East & West Volume 64, Number 1 January 201 4 1 51-1 73 151
© 2014 by University of Hawai'i Press

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
svabhāva or parabhāva. At least, this is how the nihilist interprets these statements. I
will call these statements ESP statements, for "existents are svabhāva or parabhāva "
If such statements are to be interpreted as the nihilist maintains, and if it is also true
as Nāgārjuna claims that in fact nothing exists intrinsically, then the nihilist claim
that nothing exists at all seems to follow. The second source of the nihilist reading,
I believe, is an alleged incoherence in the anti-essentialist interpretation. The anti-
essentialist holds that, for Nāgārjuna, nothing is svabhāva , and that at the con-
ventional level things exist only in dependence on other existents and on human
knowers. It has sometimes been alleged, however, that it makes no sense for all exis-
tence to be dependent existence, since there would then be no independently exis-
tent things for the dependent things to depend on. Paul Williams, for instance, has
recently argued that most Buddhists in India viewed Nāgārjuna's position as tanta-
mount to nihilism, because if all things were dependent, then, as he puts it, "all
things would be constructs with nothing for them to be constructed out of."4
These two sources of the nihilist reading represent, I would argue, important
challenges to the anti-essentialist interpretation, challenges that the anti-essentialist
view has not properly accounted for and that will continue to raise nihilist doubts
until they are properly accounted for. But it is wrong to think that in order to accom-
modate these challenges we have no option but to accept the nihilist account. What
I propose here is an alternative reading of Nāgārjuna, what I will call the conceptual-
ist interpretation, which, I will argue, can accommodate the nihilist challenges while
also doing greater justice to Nāgārjuna's writings than the anti-essentialist can. As we
shall see, the conceptualist account has very different implications from both the
anti-essentialist and the nihilist views for how we should think about the nature of
the two truths.
My plan will be first to clarify what the conceptualist view holds, then to con-
sider what exactly is problematic about the nihilist interpretation, and finally to argue
that the conceptualist account does a better job of responding to the two nihilist chal-
lenges than can the anti-essentialist.

The Conceptualist Interpretation

The conceptualist account finds an element of truth in both the anti-essentialist and
the nihilist interpretations. The conceptualist agrees with the anti-essentialist, contra
the nihilist, that for Nāgārjuna there is a sense in which things may be said to exist at
the conventional level. But this interpretation understands conventional existence in
a way very different from the anti-essentialist. The central move of the conceptualist
account is to interpret the ESP statements as statements not directly about existent
things, but rather about the concept of existence, or about the meaning of terms like
"exists," and about existents only indirectly. That is, this account views the ESP state-
ments as expressing what I will refer to as the conceptual thesis:

The Conceptual Thesis. It is a core part of the meaning of the concept of existence (bhāva)
that whatever exists either exists independently (is svabhāva) or is dependent for its exis-
tence on the independent existence of something else (is parabhāva).

1 52 Philosophy East & West

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
This view interprets Nāgārjuna's claim as being directly about the concept of exis-
tence, or about the meaning of terms like "exists," rather than being directly about all
existent things, as the nihilist assumes. This is not, I should emphasize, to view it as
saying nothing about existent things themselves at all; obviously, a claim about the
meaning of "exists" is still indirectly a claim about the nature of existent things as
well. But, as we shall see, it is the fact that it interprets the ESP statements as being
directly about concepts that allows the conceptualist account to offer an alternative
route for avoiding the nihilist view.
The second key component of the conceptualist account is its interpretation of
the nature of the conventional truth. The conceptualist view of the nature of dis-
course at the level of conventional truth is quite different from the anti-essentialist
account of this discourse. As we shall see more fully later, according to the anti-
essentialist, discourse at the level of the ultimate truth is discourse that operates with
a concept of existence as svabhãva, as independent existence; it is because in fact
nothing is svabhãva that at the level of ultimate truth nothing can be said to exist. By
contrast, discourse at the level of conventional truth, for the anti-essentialist, operates
with a concept of existence that is distinct from the concept employed at the ultimate
level, namely a concept of "dependent existence." And since things are indeed inter-
dependent, they can thus be said to exist at the conventional level. The conceptualist,
however, finds truth in the nihilist's critique of the notion of universal interdependent
existence. This account views Nāgārjuna as holding that we cannot make sense of
the idea of a world in which all existence is dependent existence, without any inde-
pendently existing things to serve as a foundation; this idea is at root incoherent. This
does not mean that the conceptualist holds that for Nāgārjuna things really do exist
as svabhãva. Clearly for Nāgārjuna all things are empty of svabhãva. The conceptual-
ist simply rejects the analysis of emptiness in terms of interdependent existence. For
the conceptualist, then, Nāgārjuna's view is not that we have a distinct concept of
dependent existence at the conventional level. We have only one concept of exis-
tence, used at both the ultimate and conventional levels, namely the concept de-
scribed by the conceptual thesis, which includes the belief that whatever exists is
either svabhãva or parabhāva. What account does the conceptualist offer, then, of the
difference between discourse at the ultimate and the conventional level?
According to the conceptualist reading, discourse at the ultimate level is dis-
course that in a certain sense endorses the full content of the concepts it employs -
specifically, in the case of the concept of existence and derivative concepts, the
presupposition that whatever exists is either svabhãva or parabhāva. What I mean by
"endorse" here is not that speakers at this level of discourse necessarily believe that
whatever exists is svabhãva or parabhāva ; enlightened speakers at the ultimate level
in fact do not believe this. Rather, the point is that this level of discourse is character-
ized by the fact that speakers take it to be part of the meaning or truth conditions of
existence claims that existents be either svabhãva or parabhāva. Deluded speakers
wrongly believe that these truth conditions are sometimes satisfied, that some things
do exist as svabhãva ; enlightened speakers rightly believe that the truth conditions
are never satisfied, that nothing exists in this sense because nothing is svabhãva. But

John Spackman 153

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
either way, to speak at this level is to treat existence claims as having the specified
truth conditions. More generally, since all assertorie statements at the ultimate level
make claims about the existence or reality of certain objects, events, or properties, it
will be part of the truth conditions of all statements at this level, whether they actu-
ally use the concept "exists" or not, that whatever they claim to exist is svabhāva
or parabhāva. It is speech of this kind, according to the conceptualist view, that
Nāgārjuna has in mind when he refers to language that purports to make assertions
(vaktavya).
By contrast, in the conceptualist account, discourse at the conventional level is
discourse in which speakers do not endorse the full content of the concepts they
employ, in the sense described above. Specifically, they do not endorse part of the
meaning of the concept of existence and derivative concepts, namely the stipulation
that whatever exists is either svabhāva or parabhāva , in the sense that they do not
treat this as part of the truth conditions for existence claims at the conventional level.
Speakers at this level can thus treat some existence claims as true, in the conven-
tional sense, even if there are, and they believe that there are, nothings that are either
svabhāva or parabhāva.
This way of interpreting conventional discourse presupposes that for Nāgārjuna
it is possible to apply terms or concepts truly without endorsing all aspects of their
meaning, though presumably one would have to endorse enough of their meaning
that one could still be said to be using the concepts in question. It presupposes, that
is, what might be called a "pragmatic thesis" about language use and meaning:

The Pragmatic Thesis. It is possible truly to apply a concept without endorsing all features
of its meaning - without treating all features of its meaning as truth conditions for applica-
tions of the concept - as long as one endorses enough of its meaning that the concept
retains its identity.

I call this thesis pragmatic because language used in this way - language at the con-
ventional level - might be described as functioning pragmatically in the broad sense
that among the features associated with the meaning of any given concept, the spe-
cific features by virtue of which it applies in any case might vary with the context, so
that the meaning of statements might potentially be determined in part by the spe-
cific communicative purposes of the speaker in the context of use.5
If we adopt this conception of conventional language, and also interpret the ESP
statements along the lines of the conceptual thesis, it becomes possible to avoid the
nihilist challenges we noted above. If we view Nāgārjuna as holding that ESP state-
ments of the form "whatever exists is either svabhāva or parabhāva " apply directly to
all things that exist, then this, together with Nagarjuna's denial that any thing is
svabhāva , does entail nihilism. But if we interpret these statements as making a point
about the concept of existence or about the meaning of terms like "exist," nihilism
does not follow. It does not follow because on the conceptualist view, even though it
is part of the meaning of the concept of existence that whatever exists is svabhāva
or parabhāva , it is possible in conventional discourse to truly apply this concept - to
say that things exist - without endorsing this part of the concept's meaning. The ESP

1 54 Philosophy East & West

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
statements are thus compatible with saying that things exist at the conventional level,
so they pose no threat of nihilism.
The second nihilist challenge derived from the alleged incoherence of the anti-
essentialisťs notion of interdependent existence. The conceptualist interpretation ac-
cepts the incoherence of this idea, but because of its interpretation of the nature of
conventional discourse it can simultaneously affirm that things can be truly said to
exist at the conventional level. This is possible because to say that things exist at the
conventional level is not to apply a concept of interdependent existence, but rather
to apply the same concept of existence we use at the ultimate level, but without
endorsing its svabhāva- related content. As we shall see, the implication is that for
the conceptualist not only the ultimate truth but the conventional truth becomes
inconceivable, for what it means to say conventionally that things exist is not fully
intelligible.

What Is Problematic about the Nihilist Account?

Before turning to my case for the conceptualist view, it will be helpful to consider
what exactly is problematic about the nihilist interpretation. The nihilist reading has
faced many objections, but I will focus here on what I take to be the most important
of these. Remember that the nihilist maintains that for Nāgārjuna nothing exists at
all in any sense, either at the ultimate or at the conventional level. The main obsta-
cles to this account are textual ones, for there are many statements in Nāgārjuna's
writings that do seem to affirm the existence of things in some sense. Consider, for
instance, the so-called positive catuskoti or tetralemma from chapter 18 of the
Mülamadhyamakakärika ( MMK ), on the self. Here Nāgārjuna states that the buddhas
have taught that there is a self, that there is no self, and that there is neither self nor
not self (1 8 : 6), and then goes on to make a parallel claim concerning all things quite
generally:

Everything is real; and is not real;


Both real and unreal;
Neither unreal nor real.
This is the Lord Buddha's teaching.6 (18:8)

The four statements made here ("Everything is real," "Everything is not real," "Every-
thing is both real and not real," and "Everything is neither real nor not real") are the
affirmative correlates of the statements of the more usual negative form of the
catuskoti and the first of them appears to affirm that there is nothing objectionable
in saying, in some sense, that things are real.7 CandrakTrti, we should note, interprets
these positive statements as upãya, expedient means intended for beings at different
levels of awakening, and so as true at the level of conventional truth, though not
ultimately.8
Of course, the nihilist will not accept that such statements are intended to be true
even at the conventional level. Thomas Wood, perhaps the most prominent contem-
porary proponent of the nihilist reading, avoids this implication of the present lines

John Spackman 155

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
by maintaining that teachings at the conventional level are not in fact "truths" at all,
but in fact proclamations of falsehoods that can lead others to the real, ultimate
truth.9 As he puts it, "since this lower or provisional truth contradicts the absolute
(paramārthika) truth, it must be, from a logical standpoint, simply false. And this, I
think, is exactly how the Mädhyamikas viewed the matter."10The distinction between
paramãrthasatya and samvrtisatya, for him, is in fact not a distinction between two
kinds of truth, but rather a distinction between two kinds of teaching that the Buddha
used on different occasions. In this account, the Buddha sometimes taught as //Things
were real in order to lead people to realize the ultimate truth, but this teaching was
not in fact true. Wood makes a similar point about the kind of discourse he takes to
be characteristic of the conventional level, what is called, at MMK 24 : 1 8, prajñaptir
upãdãya , which Wood translates as "conventional, dependent designation." Prajñapti
he takes to mean a merely conventional or "purely verbal" notion used to make
something known, and so he regards prajñaptir upãdãya as "an expression or propo-
sition that does not express or state the way things really are."11
There is good reason to think, however, that this nihilist account does not do
justice to the complexity of the relation between paramãrthasatya and samvrtisatya
as Nāgārjuna conceived them. Perhaps the main obstacle standing in the way of this
view of conventional truth is the apparent identification of emptiness and dependent
co-arising at MMK 24 : 1 8:

That which is dependent origination


Is explained to be emptiness.
That, being a dependent designation,
Is itself the middle way.

This stanza is often taken as establishing an identity between the ultimate truth that
things are empty and the conventional truth concerning the various dependently
arisen things. But if, as Wood suggests, the conventional truth is nothing but a false
overlay over the ultimate truth of emptiness, it is hard to see how the two could be
identical. Wood in fact does not dispute that the verse states an identity between
paramãrthasatya and samvrtisatya , but he suggests that he can avoid this objection to
his view by regarding this equation as itself being merely a conventional rather than
an ultimate one. Though conventionally things may be said to be empty, in order to
promote realization, ultimately there are no things to be called empty. Wood's main
support for this reading of the stanza is a general philosophical one.12 If two things
are identical, they must share all of their properties in common; but emptiness is
beyond dependence and relativity, while things that are dependently co-arisen are
dependent and relative; thus, he argues, it cannot be true at the ultimate level - that
is, it cannot be true - that emptiness and dependent co-arising are identical.
It is questionable, though, whether Wood's proposed response to this objection
is adequate. Even if we grant that the claim of the identity of emptiness and depend-
ently arisen things is not true at the ultimate level, but only conventionally, Wood's
account faces a dilemma concerning the nature of such conventional identity claims.
Wood regards these identity claims, like all conventional statements, as strictly false,

1 56 Philosophy East & West

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
but what does it mean to say that they are false? Consider, for instance, a statement
like "this desk, as a dependently arisen thing, is empty/' If to say that this statement
is false means that it does not pertain in any way to the true nature of things, it is hard
to see why it should be any different from ordinary falsehoods like "the sun orbits
around the earth/' and in particular why it should promote realization any better than
such ordinary falsehoods. In order to show how such identities are better suited to
promote realization than other falsehoods, Wood might say that these identities do
pertain to the nature of things in some way - perhaps, for instance, by being causally
related to states of realization - but still without being true. But if such statements do
pertain to the nature of things in some way, it is hard to see how this differs from say-
ing that they are conventionally true. What exactly it means to say that a statement is
conventionally true is a matter of considerable debate. But as long as Wood would
allow that the identity statements in question pertain to reality in some respect, even
just causally, it would seem that what he is offering is really an account of the nature
of conventional truth, rather than an account that rejects the truth of conventional
statements entirely. And if conventional statements that things exist must be true in
some sense, then the nihilist account cannot be right.

The Case for the Conceptual ist Account

If, then, for Nāgārjuna there is some sense in which things may be said to exist at the
conventional level, the important question becomes in what sense this is so. It is on
this question that the conceptualist interpretation differs from the anti-essentialist
view. So why should we prefer the conceptualist view on this issue to the anti-
essentialist account? The main grounds for adopting the conceptualist account stem
from the need to avoid the two nihilist challenges discussed earlier: the alleged in-
coherence of the anti-essential isťs notion of dependent existence and the need to
accommodate ESP statements in Nāgārjuna's writings. We will consider these two
challenges in turn, but first it will be helpful to clarify more fully the anti-essentialisťs
conceptions of ultimate and conventional truth.
The anti-essentialist perspective seems to depend, implicitly or explicitly, on our
having a coherent concept of conventional existence distinct from the concept we
employ at the level of ultimate truth. Consider, for instance, Garfield's view. The ul-
timate truth, for Garfield, is that all things are empty of svabhāva, of independent
existence. Things may still be said to exist at the conventional level, however, and
conventional existence for Garfield is dependent existence: the existence of things
dependent on their parts, their causes, their materials, and, above all, on our concep-
tualization and language. Part of what it means to reject the idea that things are
svabhāva is that it is not possible to establish criteria of identity for them in a prin-
cipled way independently of the words we use. Thus, to say that something is de-
pendently arisen is to say that its identity is finally nothing more than its being the
referent of a word.13 The conventional truth, for Garfield, is thus "the truth about
things as they appear to accurate ordinary investigation," which is dependent on
tacit human agreement.14

John Spackman 157

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
It is especially important for our concerns to note how Garfield clarifies his view
of the relation between the two truths in terms of the functioning of language at
the ultimate and conventional levels. Following Gadjin Nagao, he views the role of
language at the ultimate level as captured by the "negative catuskoď that appear at
several points in the Mûlamadhyamakakàrikâ, most importantly at 22 : 1 1 :15

We do not assert "Empty/'


We do not assert "Nonempty/'
We neither assert both nor neither.
They are asserted only for the purpose of designation.

Garfield views this passage as showing that, for Nāgārjuna, at the ultimate level noth-
ing at all - not even that all things are empty - can be asserted about reality.16 The
notion of "assertion" (vaktavya), which Nāgārjuna employs here, he takes to be the
ascription of a property to an entity.17 But for Nāgārjuna, ultimately there are no in-
dependent entities to which we could ascribe properties, nor indeed are there inde-
pendent properties to ascribe.
However, in contrast to its functioning at the ultimate level, from the conven-
tional standpoint language does make assertions. Conventionally, entities and prop-
erties do exist, and so from the conventional standpoint language functions by
ascribing conventional properties to conventional entities. Garfield holds that "cor-
responding to these conventional assertions are real propositions that make them
true or false - entities with or without the ascribed properties."18 He does not, of
course, regard these entities and properties as existing as svabhāva ; they exist only
dependently. But it seems to be part of this picture, since the constituents of the
thoughts involved in assertion are generally viewed as concepts, that at the conven-
tional level we possess a concept of existence - dependent existence - that is dis-
tinct from the concept of existence at the ultimate level.19 This represents a central
difference from the conceptualist account. Garfield holds, as does the conceptual ist,
that in conventional discourse speakers do not commit themselves to the belief that
entities are svabhāva. But for Garfield they do so by employing a perfectly coherent
concept of existence different from that used at the ultimate level. The conceptualist
doubts the coherence of such a concept; in this view, conventional speakers avoid
commitment to things being svabhāva by giving up part of the concept of existence
used at the ultimate level.
Thus, in response to the first nihilist challenge that it is not possible to make sense
of the idea of all things being interdependent with no independent things for them
to depend on, the anti-essentialist maintains that we do possess a coherent concept
of such existence, while the conceptualist holds we do not. It will be helpful at this
point to spend a bit of time considering whether this notion of universal ontological
interdependence is in fact intelligible. Whether this idea is coherent is of course dis-
tinct from the question of whether Nāgārjuna held that it is. But seeing why it might
be regarded as incoherent may make it more plausible that Nāgārjuna doubted its
coherence as well. Though I cannot here present a fully developed argument against
the intelligibility of universal ontological interdependence, I do want to suggest at

1 58 Philosophy East & West

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
least that there are strong reasons for doubting its intelligibility. To this end, I will
consider the difficulties that face several contemporary defenses of this sort of view.
I will address below what reasons we might have for attributing such a position to
Nāgārjuna.
What I take to be the fundamental difficulty for the notion of universal ontologi-
ca! interdependence and related notions, a difficulty that has been raised both his-
torically and in contemporary discussions, concerns whether such views can account
for the world having a determinate character, being one way rather than another.
We can appreciate the basic problem by considering recent debates over the related
question of whether it would be possible for there to be a world in which objects
possessed only dispositional properties, powers to affect other objects, without any
categorical (i.e., non-dispositional) properties underlying them.20 Several contempo-
rary philosophers have defended the coherence of such a world without categorical
properties.21 A number of others, however, have argued that the idea of such a world
is incoherent, one reason being precisely that it would be hard to see how such a
world could have any determinate nature at all.22 To see why, consider the model of
such a world presented by Richard Holton.23 The constituents of this world are wholly
characterized by four dispositional sentences: P, Q, R, and S. For instance, P is de-
fined wholly in terms of R and S, R in terms of P and Q, and so on. Holton suggests
both that this world is coherent and that while the definitions of the sentences are
circular this is not problematic, since even though we cannot know what P says in-
dependently of what R and S say, we can "solve" all of the definitions simultane-
ously. But this seems simply to skirt the fundamental problem. The real problem is
that if the objects in this world have no categorical properties underlying their dispo-
sitions, it is hard to see what determines these dispositions to be one way rather than
another - why, for instance, P should be defined in terms of R and S rather than in
terms of any other sentences. In the absence of categorical properties, there is noth-
ing to explain why the world has one determinate nature rather than another.
Let us turn from this discussion of dispositions back to the idea of universal on-
tologica! interdependence itself. One recent defender of the coherence of this idea is
Ross Cameron, who argues that the widespread intuition against "metaphysical de-
pendency all the way down" cannot be justified by any more basic metaphysical
principle.24 Cameron, we should note, does actually end up supporting this intuition
by appeal to its theoretical utility, but more important for our concerns is that he
views the rejection of it as a coherent possibility. One argument against the possibil-
ity of a world of universal interdependence, familiar from Leibniz, is that if there
were an infinite chain of entities e1r e2, e3, . . . where e ^ is ontologically dependent
on e2, e2 on e3, and so on, then even though each link in the chain would have a
ground (metaphysical explanation) of its existence, there would be nothing to ex-
plain why the chain as a whole had the nature that it did.25 Cameron maintains,
however, that even if we grant that the chain as a whole must have a metaphysical
explanation, this argument does not entail that there must be an independent entity.
For perhaps the chain ev e2, e3, . . . depends on and is explained by another entity
eaļ, which is itself part of an infinite chain of dependence eav ea2, ea3, ... ; and the

John Spackman 159

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
chain eau ea2, ea3, ... as a whole depends on an entity eò1r which belongs to a chain
eòi/ eb2 f eò3, • • • / ar|d so on- If this were the case, he argues, then each entity and
each chain would have a metaphysical explanation, without there being any inde-
pendent entity.
This response, however, just seems to postpone the problem posed by the origi-
nal Leibnizian argument, for in the envisaged state of affairs there would still be no
explanation for why the world as a whole has the nature that it does. Just as the Leib-
nizian argument points out concerning the first chain e1r e2, e3, . . . that there is no
explanation of why it has the nature that it does, so we can ask, concerning the total-
ity of interlocking chains of dependence, why it has the character that it does. If it is
suggested that what determines the nature of this totality is yet another entity outside
it, itself part of an infinite chain of dependence, we can then ask of this new totality
what determines its nature. And so on. The upshot is that it is hard to see how a world
consisting solely of dependent entities could have any nature at all - how there could
be, as Blackburn puts it, any truth about such a world.26
As noted earlier, what the conceptualist concludes from this line of thought is
not that we must hold that some things do, in fact, exist independently, as svabhãva,
nor that Nāgārjuna holds this. Nāgārjuna's view is clearly that all things are empty of
svabhãva. The conceptualistas point is rather a critique of the anti-essentialist account
of what this emptiness consists in. The anti-essentialist notion that conventional exis-
tence is interdependent existence is not ultimately a coherent way of thinking of a
world without anything that is svabhãva. The conceptualist proposes instead a way of
thinking about the ESP statements that both acknowledges the incoherence of the
idea of interdependent existence and allows us to avoid the apparently nihilistic im-
plications of this incoherence. Such a view is not burdened with the problems facing
the notion of universal interdependent existence. It avoids nihilism not by equating
conventional existence with interdependent existence, as the anti-essentialist does,
but by viewing conventional existence statements as applying the standard concept
of existence without endorsing its svabhãva- related content. This view of conven-
tional discourse provides the foundation, as we shall see in the final section, of an
alternative way of thinking about the conventional existence of a world with nothing
svabhãva.

Can we really view Nāgārjuna as holding that the idea of universal ontological
interdependence is incoherent? It may seem that there are several insuperable ob-
stacles to doing so. First, MMK 24 'A 8 apparently identifies emptiness and dependent
origination, and this might be seen as tantamount to equating emptiness with the
interdependence of all things. Similarly, Nāgārjuna's apparent strategy in many chap-
ters of the MMK is to exhibit the reciprocal dependence of various concepts, for in-
stance motion, the mover, and the moved (chapter 2); seeing and the seer (chapter 3);
characteristics and the characterized (chapter 5); and so on. What must be recalled
here, however, as commentators of many stripes (including Garfield and Wood)
agree, is that the identity of 24:18 and the other statements of reciprocal depen-
dence must be taken as truths at the conventional level. The conceptualist can per-
fectly well admit that the identity of emptiness and universal interdependence is true

1 60 Philosophy East & West

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
at the conventional level, even though this notion of interdependence is not com-
pletely intelligible. To say that the notion of universal interdependence is not com-
pletely coherent is not to say that it is without utility. The conceptualist can certainly
allow that we have good reasons for affirming conventionally that everyday things
exist interdependently. What she denies is the anti-essentialisťs view that this notion
gives us a completely consistent, intelligible way of thinking of a world without
svabhāva.

It is worth noting that the idea of universal interdependence was likely not as
familiar in Nāgārjuna's time as it is now. Joseph Walser argues that there was only
one Abhidharma school in the era of Nāgārjuna, the Prajnaptivādins, that adopted a
doctrine of "reciprocal designation" (anyonya prajñapti ), according to which all con-
ditioned things are conceptually constructed dependent on other conceptually con-
structed things.27 It is certainly possible that Nāgārjuna held such a position. But it
is also not implausible historically that he did not view the idea of universal inter-
dependence as coherent.
The main positive grounds for thinking that Nāgārjuna in fact doubted the coher-
ence of universal ontological interdependence are textual ones, and the most direct
textual support comes from the ESP statements. Let us turn, then, to the question of
how to understand these statements, which represent the second nihilist challenge
to the anti-essential ist view. Anti-essential ists do in fact offer interpretations of the
ESP statements that would dispel their apparent nihilist implications. What I would
suggest, however, is that the anti-essentialist accounts of these statements inevitably
resort to ad hoc readings of terms in the texts in order to preserve their view of con-
ventional existence. Interpreting the ESP statements as expressing the conceptual
thesis allows us both to avoid nihilism and to be more faithful to Nāgārjuna's texts
than the anti-essentialist account.
There are a number of examples of ESP statements both in the Mdlamadhya-
makakārikā and in Nāgārjuna's other writings. I will focus here on three of these.
What is perhaps the clearest passage involving an ESP statement occurs in the chap-
ter on svabhāva itself. Verse 1 5 :4 of the Mûlamadhyamakakârikâ reads:

Without having either essence or the nature of being other,


How could things exist?
If essence or the nature of being other existed,
Entities would be established.

How are we to interpret this passage? In the Sanskrit original, the term here translated
as "essence" is svãbhava, "the nature of being other" translates parabhāva , and "enti-
ties" is bhāva. The word bhāva ordinarily refers simply to existent things. If we read
the passage taking bhāva in this way, the first two lines appear to say that it is not
possible for things to exist except as either svabhāva or parabhāva. This is the nihilist
reading of these lines. For if the text is read in this way, then given the dependence of
what is parabhāva on what is svabhāva , and given Nāgārjuna's clear denial that any-
thing exists as svabhāva , the lines would entail that nothing exists at all, in any other
sense.

John Spackman 161

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
In order to avoid this nihilist result, anti-essential ist interpreters reject the literal
reading of bhāva as it appears in this passage. Garfield, for instance, reads lines b
and d as referring not simply to existent things, but to "inherently existing entities/'
that is, entities conceived as svabhāva.28 Interpreted in this way the passage has no
nihilist implications: it says only that it is not possible for inherently existing things
to exist except as svabhāva or parabhãva. This reading is consistent with the anti-
essentialisťs view of Nāgārjuna as rejecting not the existence of things per se but
only their essential existence. And yet there seems to be no strictly textual basis for
interpreting bhāva in this context as restricted in meaning to "inherent existence"
rather than the broader "existence." After all, Nāgārjuna has a word for "inherent
existence," namely svabhāva , which he could have chosen to use here, but did not.
Notice, too, that Garfield's reading has the unfortunate effect of portraying each of
the two statements in 15:4 as peculiarly elaborate tautologies. What the first two
lines say, in this account, is: how could there be things with essence that do not have
essence or the nature of being other?
One might, of course, view the anti-essentialist reading as forced on us by
broader interpretive requirements, and in particular by the need to avoid the nihilist
view. But we can avoid the nihilist account, while at the same time avoiding the anti-
essentialisťs arbitrary reading of bhāva , by understanding the lines in accordance
with the conceptual thesis. In this account, bhāva would indicate existence in gen-
eral and not just inherent existence, but the lines would make a point not directly
about existent things but about the concept of existence, or the meaning of the term
bhāva. What these lines say, in this view, is that it is part of the very concept of an
existent thing (bhāva) that existent things must be either svabhāva or parabhãva. This
must be, of course, a claim not merely about concepts, but also about things insofar
as concepts apply to them with their full sense - if no things are svabhāva , the claim
will entail that no things, not even conventional things, exist in the full sense of the
term "exists." But if we accept the conceptualist account of conventional discourse,
that in such discourse concepts are applied without endorsing these concepts'
svabhāva- related content, the conceptual thesis will not have nihilist implications. It
will be compatible with saying that conventional things exist in a looser sense.
As a second example of an ESP statement, consider, for instance, Mdlamadhya-
makakārikā 1:10, where Nāgārjuna expresses what I take to be the conceptual thesis
in terms of words involving the root sat , existence or being, as well as in terms of
bhāva. Kalupahana translates the verse as follows:

Since the existence (sattā) of existents (bhāvānām) devoid of self-nature (nihsvabhāvānām)


is not evident, the statement: "When that exists, this comes to be" will not be appropriate.29

Garfield and Samten's rendering is similar:

Since there is no existence


Of essenceless things,
The statement "When this exists, this will arise"
Is not tenable.30

1 62 Philosophy East & West

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
There is controversy about how to interpret this verse on several levels, most promi-
nently concerning the attitude it expresses toward the traditional Buddhist formula
for causation, "when that exists, this comes to be." In its most literal reading, the
stanza seems to reject this traditional formula altogether, and with it, perhaps, depen-
dent co-arising itself. This would be the nihilistic interpretation of the verse. Accord-
ing to Wood, for instance, just as Nāgārjuna denies the existence of things in every
sense, he rejects the arising and ceasing of things in every sense. But most recent
interpreters have, in an anti-essentialist vein, rejected this nihilistic reading in favor
of the view that what Nāgārjuna calls "inappropriate" here is only a svabhāva-
oriented understanding of the formula, that is, one which takes the things that arise
and cease to be svabhãva.
What is most important for our purposes, however, is the first half of the verse.
Taken most literally, these lines appear to say simply that there are no beings or exis-
tents (sattã) that are not svabhãva. And just as in the case of MMK 15:4, since
Nāgārjuna clearly repudiates things being svabhãva , if we interpreted the lines in this
way, the result would be a nihilist reading. Kalupahana and Garfield, however, seek
to avoid the potentially nihilist implications of this passage in their translations. Kalu-
pahana reads sattā in this context not simply as having its literal meaning of "being"
or "existence," but rather as meaning "substantial existence," that is, existence as
svabhãva .31 And Garfield offers a similar rendering of the Tibetan text. Once again it
seems arbitrary to impose such a restriction of meaning when there is no textual basis
for doing so. Furthermore, as in the case of Garfield's reading of MMK 15:4, this
reading makes the first half of the verse tautological. But we can both avoid the nihil-
ist interpretation and simultaneously do greater justice to the text by reading sattā
here as meaning simply "being" or "existence," and then viewing the passage as
expressing the conceptual thesis. In this reading, the first two lines proclaim a con-
ceptual link between sattā and svabhãva ; what they claim is that it is part of the
meaning of the concept of sattā , being or existence, that what is said to be or exist is
svabhavā.
As a final example of textual evidence, consider a passage from the Mdlamadhya-
makakārikā that makes a claim related to the conceptual thesis, and which I take in
fact to entail it. Chapter 5 of the work is devoted to a consideration of the notion of
a "characteristic" (laksaņa), and in particular to the application of this notion to one
of the six elements (dhātus), space. The chapter is usually taken to be directed against
an understanding of the notion of a characteristic such as that adopted by some
Abhidharma schools, such as the Sarvāstivādins. In such a view, a characteristic is
a defining or essential property of a thing, in the specific sense that it defines the in-
herent existence or essence (svabhãva) of the thing, and might perhaps be viewed
as inherently existing itself. In the Abhidharmakosa, for instance, space is said to be
one of the unconditioned dharmas that "has for its nature, not hindering matter"
(anāvarana).32
The conclusion for which Nāgārjuna argues in chapter 5, against this view, is that
space is neither an existent (bhāva) nor a non-existent (abhāva) thing. The argument
proceeds as follows. The second verse of the chapter reads:

John Spackman 163

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
A thing without a characteristic
Has never existed.
If nothing lacks a characteristic,
What do characteristics characterize? (5:2)

The aim of this verse is to establish the interdependence of existent things (bhāva)
and characteristics (laksaņa). Nāgārjuna argues on this basis that since characteristics
cannot be said to arise either in the uncharacterized or in the characterized (5:3),
there are no characteristics, and hence no existent things (5:5); and given the inter-
dependence of the existent and the non-existent (5:6), nothing, including space, is
either existent (bhāva) or non-existent (abhāva) (5 : 7).
Now how one interprets this argument will depend crucially on how one inter-
prets the term laksaņa. If one views it as meaning an essential property, as in the
Sarvāstivādin account, 5:2 might seem to suggest a nihilistic interpretation of the
chapter as a whole. For if laksaņa here means a property that defines the inherent
existence (svabhāva) of a thing, the claim that there are no things without character-
istics will be tantamount to the claim that there are no things (bhāva) that are not
svabhāva. And as we have seen before, given Nagarjuna's denial that anything is
svabhāva , this would mean that there are no things at all. The chapter's ultimate
claim that nothing is either a bhāva or an abhāva would then be viewed as an ex-
pression of this nihilism. Garfield, as usual, resists the nihilist account, and instead
interprets the main claim of the chapter as being that space, like other things, neither
exists as svabhāva nor fails to exist entirely. In order to support this reading, though,
he must once again resort to what seem to be ad hoc interpretive hypotheses, this
time by interpreting the same term, laksaņa , as meaning different things in different
contexts, without any strictly textual basis for doing so. In his general commentary on
the chapter, Garfield says he subscribes to the standard interpretation of laksaņa as
meaning "distinguishing characteristic" or "essence." But in fact he reads the term as
it appears in 5:2ab as simply meaning "properties," which allows him to preserve
the anti-essentialist interpretation by viewing these lines as claiming not that there
are no things without essential properties, but simply that there are no things without
properties.33 And yet when Nāgārjuna concludes at 5 :5 that "there is no character-
ized and no existing characteristic," in this case he reads "characteristic" as mean-
ing "inherently existing characteristic," where this reading accords with his anti-
essentialist account.34
These tensions in the anti-essentialist interpretation can be avoided, though, if
we read MMK 5 :2 in light of the conceptual thesis. It is possible both to read laksaņa
consistently as meaning a defining or essential characteristic and to resist the pull of
the nihilist interpretation. In such an account, what 5 : 2 claims is that it is part of the
concept of an existent thing (bhāva) that it be viewed as having essential characteris-
tics (laksaņam), characteristics that define it as svabhāva. This claim entails the con-
ceptual thesis, and might be viewed as a corollary of it. But, once again, such an
account can avoid the potentially nihilistic implications of the claim by holding that

1 64 Philosophy East & West

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
to use the concept of existence at the conventional level is to use it without endorsing
the conceptual thesis or this corollary.
In each of the three examples of ESP statements we have examined, we can avoid
the apparent nihilism involved in these statements, and also give a more consistent
and faithful account of the textual evidence, if we view the statements as expressing
the conceptual thesis rather than in an anti-essentialist vein. Note finally that in ad-
dition to this textual support for the conceptual thesis, there are several statements in
the MMK that seem, prima facie, to cast doubt on the idea of reciprocal dependence
that underlies the anti-essentialist notion of universal ontological interdependence
itself. For instance,

Fire is not dependent on fuel


Fire is not independent of fuel.
Fuel is not dependent on fire.
Fuel is not independent of fire. (1 0 : 12)

Since it appears to state that fire is neither independent of nor dependent on fuel, and
so, too, for fuel on fire, this verse has sometimes been viewed as expressing a critique
not only of the idea of independent existence but also of the notion of reciprocal
dependence itself.35 The anti-essentialist reading of such lines avoids this interpreta-
tion by restricting their reference to entities conceived as svabhāva. What Nāgārjuna
means by saying "Fire is not dependent on fuel/' in this view, is that if fire and fuel
are conceived as svabhava, fire cannot depend on fuel, since fire would by definition
be independent.36 But there is no explicit indication in the text that "fire" and "fuel"
should be understood in this restricted way, and the conceptualist account offers a
coherent way of reading them as referring to fire and fuel in an unrestricted sense.
In this reading, clearly the implication could not be that there is no sense in which
entities are mutually interdependent, since, as we have seen, Nāgārjuna's strategy is
often to demonstrate the mutual dependence of various concepts. Rather, the impli-
cation would be that although the notion of reciprocal dependence can usefully be
employed at the conventional level, there is even at this level also a sense in which
it cannot properly be affirmed of things. In the conceptualist account, this is because,
for Nāgārjuna, in the last analysis universal ontological interdependence is not a
coherent concept.
I take the textual evidence we have considered, together with the fact that con-
ceptualism also allows us to avoid the objection that the anti-essentialist account of
interdependent existence is incoherent, to provide strong reasons for adopting the
conceptualist reading rather than the anti-essentialist one. Before considering some
of the implications of the conceptualist view, though, I will consider two possible
objections that might be raised against it.
Given that I have claimed that the conceptualist account of the ESP statements is
superior to the anti-essentialist account in part because the anti-essentialist view re-
quires arbitrary or ad hoc readings of the text, one objection that might be leveled
against the conceptualist account is that in fact it is equally or more arbitrary. After

John Spackman 165

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
all, Nāgārjuna does not explicitly say that the ESP statements concern not existent
things themselves but the concepts associated with terms like bhāva, sattā, and
sadbhūta, or the meanings of these terms. Nor does he speak of the possibility of us-
ing a concept without endorsing all of its content. It may also seem anachronistic to
attribute to Nāgārjuna a somewhat technical contemporary view about word mean-
ings and concepts. But the conceptualist reading seems to me neither to be arbitrary
nor to impose technicalities beyond what is required to understand the text. The con-
ceptualist reading of the ESP statements draws on a pattern of usage that is common
in everyday speech. We not uncommonly use statements - often in a universally
quantified form - that appear to be about things as a way of talking implicitly about
the meanings of concepts or terms. Statements such as "Bachelors are unmarried
men" or "How could there be a brother that was not male?" are in part about what
we mean by "bachelor" and "brother."
It seems to me quite natural to read the ESP statements as making claims about
concepts or meanings in just this way. Compare the second example just given with
the lines from MMK 15:4: "Without having either essence or the nature of being
other, how could things exist?" It is true that Nāgārjuna does not speak explicitly of
using concepts with or without all of their contents. But he does draw the distinction
between ultimate and conventional levels of discourse, and it is as an analysis of this
distinction that the conceptualist introduces the contrast between two ways of using
concepts. The conceptualist reading is not forced or arbitrary; on the contrary, as we
have seen, it allows us to make better sense of the way words are used in the texts
than does the anti-essential ist account.
A second worry that might be raised against the conceptualist view is that it is
in fact committed, as Nāgārjuna clearly would not be, to certain things having an
essential nature, being svabhāva - namely concepts and meanings themselves. For
it might be thought that what the conceptual thesis holds is precisely that the be-
lief that what exists is either svabhāva or parabhāva is an essential part of the con-
cept of existence - which might seem to entail that concepts have an essential nature.
But this is not what the conceptual thesis holds. In the sense in which I intend it,
what the thesis says is not that the belief in question is an essential part of the mean-
ing of the concept, but that it is a core part of that meaning. As I use this notion, a
core part of the meaning of a concept is not a necessary condition for its applica-
tion; the statement "whatever exists is svabhāva or parabhāva" is neither necessarily
true nor analytic. Used at the conventional level, in fact, it would be false. More
generally, it is open to the conceptualist to follow Quine's critique of analyticity in
holding that there are no beliefs that capture part of the meaning of a term that
could not turn out to be false.37 To say that a certain belief is a core part of the mean-
ing of a concept is to say, rather, that the belief is given considerable weight in deter-
mining whether the concept applies or not, and that when the belief does not hold
true we begin to lose our grip on the meaning of the concept.38 Such a view of the
structure of concepts and meanings is not committed to their having an essential
nature.

1 66 Philosophy East & West

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
Reconceptualizing the Two Truths

I will conclude by considering some of the implications of the conceptualist inter-


pretation for how we should understand the nature of the two truths, and how this
conception of the two truths differs from that espoused by anti-essential ist inter-
preters. Recall for a moment Garfield's view of the two truths and of the relation
between them. The ultimate truth is that things are empty of svabhāva. Nāgārjuna
holds - with a slight qualification that we will consider in a moment - that nothing
can be asserted about the ultimate truth. The conventional truth is identified with
pratîtyasamutpâda , the dependent co-arising of things. At the conventional level, we
can make assertions about things, because entities and properties do exist - not as
svabhāva but interdependently - that make these assertions true. And because, as
Garfield puts it, "to say of something that it is empty is another way of saying that it
arises dependency," the two truths are, ontically speaking, identical.39
From the conceptualist perspective, this anti-essentialist outlook misrepresents
the nature of the two truths, most importantly by masking their inconceivability. As is
widely accepted, there is substantial evidence that Nāgārjuna viewed the ultimate
truth, at any rate, as being inconceivable, ungraspable by concepts. Both in the
dedicatory verse of the MMK and in 1 8 : 9, he emphasizes that the ultimate nature of
things is "without distinction" (anānārtham); assuming that something that contains
no distinctions cannot be conceptualized, then, ultimate reality must be inconceiv-
able. In 18:9 he characterizes ultimate reality as nirvikalpam as well, which is often
translated as "not conceptualized":

Not conceptualized, without distinctions:


that is the characteristic of things as they really are.

The non-conceptual ity of ultimate reality has considerable importance both for a
proper understanding of the soteriological goal of Buddhist practitioners and for their
means of achieving that goal. Many anti-essentialist interpreters affirm the inconceiv-
ability of the ultimate truth for Nāgārjuna, at least in their explicit statements. And
yet it is arguable that the anti-essentialist reliance on the concept of interdependent
existence as a model for understanding the nature of a conventional world that is
empty of intrinsic existence represents both the conventional truth and emptiness as
coherently graspable by concepts. On the conceptualist view, by contrast, the ulti-
mate truth is strictly inconceivable, and indeed, because it is identical to the ultimate
truth, so is the conventional truth.
Consider first the nature of the ultimate truth. Though, as noted, most anti-
essentialist interpreters ostensibly affirm the inconceivability of the ultimate truth,
Garfield for one has recently argued that Nāgārjuna takes it to be in a limited way
conceivable and expressible. In an article with Graham Priest, "Nāgārjuna and the
Limits of Thought," Garfield has argued that statements such as "Something that is not
dependency arisen, such a thing does not exist" and "a non-empty thing does not
exist," from MMK 24:19, are positive statements of the ultimate truth. As noted

John Spackman 167

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
above, Garfield follows CandrakTrti, and more recently Nagao, in holding that the
real content of the doctrine of emptiness is given by the negative catuskoti of MMK
22 : 1 1 , which says that one should not assert any of four statements about reality, that
it is empty, that it is not empty, that it is both, and that it is neither.40 He thus holds
that for Nāgārjuna nothing expressed by language can capture ultimate reality. And
yet he also maintains that in fact Nāgārjuna does state positive ultimate truths that go
beyond the negative catuskoti .41 To affirm both that there are no ultimate truths and
that there are ultimate truths is obviously contradictory, but Garfield and Priest inter-
pret Nāgārjuna in accord with a dialetheic logic in which there are true contradic-
tions. Nāgārjuna in fact affirms certain contradictions because of the contradictory
nature of the ultimate truth itself.
For the conceptual ist, however, while statements of these kinds might appear to
be positive statements of ultimate truth, they are not what they seem to be. The con-
ceptual ist accepts that the real content of the notion of emptiness is given by the
negations of the negative catuskoti. The statement "all things are empty" is thus
equivalent in meaning to these negations, and makes no positive statement about
ultimate reality. Other statements, such as "all things are dependently arisen," should
be taken as truths only at the conventional level. Garfield and Priest resist the latter
suggestion, since such statements seem so clearly to pertain to the ultimate nature of
things.42 But the conceptual ist view illuminates why statements of this kind should be
taken as true only conventionally. The reason they cannot be true at the ultimate
level is that to speak at that level is to use concepts with their full commitment to
things being svabhāva. Since nothing is svabhãva, to apply concepts at the ultimate
level - even in such apparently ultimate statements as "all things are dependently
arisen" - is inevitably to distort the fluid and impermanent nature of reality. The only
level at which such statements can be counted as true is thus the conventional level,
where concepts are used without endorsing their svabhāva- related content. The ulti-
mate truth is thus strictly inconceivable, for the conceptual ist. This is not to say, of
course, that it cannot be perceived by the kind of direct apprehension available to the
advanced Buddhist practitioner; it just cannot be understood conceptually.
As to the conventional truth, anti-essentialist interpreters treat the nature of con-
ventionally existent things as in a straightforward way conceivable. Indeed, despite
the affirmation by most anti-essentialist interpreters of the inconceivability of the
ultimate truth, it is arguable that they treat the emptiness of conventional things as in
a certain sense conceivable as well. The anti-essentialist treats statements that things
exist at the conventional level as applications of a concept of existence, interdepen-
dent existence, which is distinct from the concept of independent existence used at
the ultimate level. The conventional existence of things is thus in principle graspable
by this conventional concept, and indeed this concept provides a means of grasping
what it is for conventional things to be empty of intrinsic existence as well.
As we have seen, the conceptualist views this concept as ultimately incoherent.
The conceptualist holds that we have only one concept of existence, the one that
contains a commitment to whatever exists being either svabhāva or parabhāva ;
conventional statements that things exist employ this same concept, but without en-

1 68 Philosophy East & West

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
dorsi ng its svabhāva- related content. For the conceptual ist, then, the only concept of
existent things on which speakers at the conventional level can draw represents ex-
istents as either svabhāva or parabhāva. They recognize, if they are enlightened, that
in fact nothing is svabhāva , and hence nothing is parabhāva. Nonetheless, their only
tools for thinking about things conceptually are tools that represent them as svabhāva ;
when they affirm conventionally that things exist, they do so only by leaving out part
of the content of their concept of existence. What speakers refer to conventionally as
existent is thus not adequately conceived of by their concept of existence. The real
nature of conventional existence, of everyday things, is in this sense, strictly speak-
ing, inconceivable. The point might also be put by saying that for the conceptual ist,
strictly speaking, things do not exist - where what this means is that in the only sense
of "exists" that we can fully understand, nothing exists. We can indeed say conven-
tionally that things exist, but when we do, we cannot fully understand what this ex-
istence consists in. From this perspective, to view everyday reality as adequately
graspable by any concept of existence, as the anti-essential ist does, is to mask the
evanescent character of that reality that Nāgārjuna sought to convey.
If the nature of conventional things, and the emptiness of these things, is not
coherently conceivable in terms of a model of interdependent existence, what alter-
native model can the conceptualist account offer? There is, perhaps, nothing wrong
with the model of interdependent existence, as long as it is recognized for what it is,
a conventional model that contains at its heart an incoherent picture. A more coher-
ent picture of conventional reality, however, suggested by the pragmatic thesis noted
above, might be a broadly pragmatist one. The pragmatic thesis holds that at the
conventional level concepts are applied not on the basis of their svabhāva- related
content, but rather on the basis of features that may vary with context according to
the communicative needs of speakers. A pragmatist account in accord with this thesis
might thus look as follows: the ultimate nature of things is wholly inconceivable and
ineffable - it can be accurately described neither as existent nor as non-existent,
neither as many nor as one. To say that something exists, at the conventional level, is
to use the concept of existence and other concepts as tools to convey various things
about that ineffable reality, without having any ultimately coherent conception of
what it is for that thing to exist. A conventional statement such as "snow is white" is
thus grounded not, contra Garfield, in conventional entities and properties of which
we can have an adequate conception, but rather, in part anyway, in conventional
practices that may vary with the circumstances and the practical and theoretical
needs of the speaker. Neither, however, can we say that such a statement is grounded
solely in arbitrary conventions independently of the ultimate nature of things, since
at the ultimate level, while we cannot assert that snow is white, neither can we assert
that it is not the case that snow is white. What, then, are the ultimate grounds of our
conventional statements? We can characterize them only negatively, by saying that
they are neither the independently existing natures of things nor wholly arbitrary
conventional practices.
Finally, on the conceptualist view not only are each of the two truths inconceiv-
able, but it is also inconceivable how the two could be identical, as Nāgārjuna holds

John Spackman 169

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
that they are. This, again, is something the anti-essentialist view fails to capture. We
can appreciate the ungraspable character of this identity by considering a problem
that confronts any account of its nature. Nāgārjuna does not speak of any entities or
properties existing, either as svabhāva or in any other sense, at the level of ultimate
truth. But the ultimate truth is supposed to be ontically identical to the conventional
truth; the entities and properties posited at the level of conventional truth are not
merely a false overlay on top of the real truth of emptiness, but are identical with
emptiness. So what is it, at the level of the ultimate truth, that corresponds to the
entities and properties that our everyday speech says exist at the conventional level?
It cannot be entities and properties at the ultimate level, since there are none. Con-
ventional entities and properties must thus in some sense be aspects of emptiness
itself. The anti-essentialist might say that the book on my table, for instance, exists in
interdependence on all other things, and that its interdependent existence is one
aspect of emptiness. I have suggested, however, that this notion of interdependent
existence is something we do not really understand. If this is so, we cannot use the
notion of interdependent existence as a kind of bridge principle to explain how my
book can be identical to emptiness. And since we do not have any other adequate
conception of the nature of conventional existence, either, for the conceptualist the
nature of this identity must itself be inconceivable. From the conceptualist perspec-
tive, if everything is empty of independent existence, it is not possible to understand
conceptually how things can appear as they do to our conventional minds at all.

Notes

For very helpful comments on earlier versions of this essay, I would like to thank Jay
Garfield, Mark Siderits, Charles Goodman, Bill Waldron, and my colleagues in the
Middlebury College Philosophy Department. I am also indebted to the suggestions of
two anonymous referees for Philosophy East and West. I would in addition like to
express gratitude for a grant from the National Endowment for the Humanities for
participation in the 2006 Summer Seminar on Mind and Metaphysics, organized by
John Heil, which contributed to some of the ideas in this essay.

1 - See Jay L. Garfield, The Fundamental Wisdom of the Middle Way: Nāgārjuna's
Mūlamadhyamakakārikā (New York and Oxford: Oxford University Press, 1 995);
David Seyfort Ruegg, "The Uses of the Four Positions of the Catuskoti and the
Problem of the Description of Reality in Mahayana Buddhism," Journal of In-
dian Philosophy 5 (1977); David J. Kalupahana, Nāgārjuna: The Philosophy of
the Middle Way (Albany: State University of New York Press, 1986); Frederick
J. Streng, Emptiness: A Study in Religious Meaning (Nashville and New York:
Abingdon Press, 1967).

2 - Thomas E. Wood, Nagarjunian Disputations: A Philosophical Journey through


an Indian Looking-glass (Honolulu: University of Hawai'i Press, 1994); David

170 Philosophy East & West

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07an 1976 12:34:56 UTC
All use subject to https://about.jstor.org/terms
Burton, Emptiness Appraised: A Critical Study of Nāgārjuna's Philosophy (Lon-
don and New York: Routledge, 1999). See also David Burton, "Is Madhyamaka
Buddhism Really the Middle Way? Emptiness and the Problem of Nihilism," in
Contemporary Buddhism 2, no. 2 (2001 ): 1 77-1 90.

3 - See, for instance, David E. Cooper, "Emptiness: Interpretation and Metaphor,"


in Contemporary Buddhism 3, no. 1 (May 2002): 7-20, and Paul Williams with
Anthony Tribe, Buddhist Thought: A Complete Introduction to the Indian Tradi-
tion (London and New York: Routledge, 2000), p. 150.

4 - Williams, Buddhist Thought, p. 150.

5 - In calling this conception of language pragmatic I do not mean to associate it in


any close sense with the classical American pragmatic conceptions of language
of, for instance, James and Dewey.

6 - Here and throughout, unless otherwise noted, I will use the translation of
the Mūlamadhyamakakārikā that appears in Jay L. Garfield and Geshe Nga-
wang Samten's Ocean of Reasoning: A Great Commentary on Nāgārjuna's
Mūlamadhyamakakārikā, which is a translation of Tsong Khapa's commentary
on the work (New York: Oxford University Press, 2006). This translation super-
sedes Garfield's earlier rendering of Nāgārjuna's text in The Fundamental Wis-
dom of the Middle Way, though I will also draw here on the interpretation
Garfield sets forth in the latter work.

7 - The Sankrit term tathyam, here translated as "real," might be more directly
translated as "such" (as Kalupahana translates; see Kalupahana, Nāgārjuna: The
Philosophy of the Middle Way, p. 269) or "just so" (as Ruegg translates; see
Ruegg, "The Uses of the Four Positions," p. 6). But translating it as "real" seems
to do justice to the gist of the passage. As Ruegg points out, the antonym
of tathyam is mrsā, "illusory" or "false" (Ruegg, ibid., p. 60 n. 17). And at
Müiamadhyamakakärikä 13:1, Nāgārjuna characterizes all things that have a
deceptive nature (mosa-dharma) as mrsā. There is at least a connection, then,
between what is tathyam and what is non-deceptive or real.

8 - See Ruegg, "The Uses of the Four Positions," p. 6.

9 - Wood, Nagarjunian Disputations, pp. 223-224.


10 -Ibid., p. 224.
11 - Ibid., p. 195.
12 - Wood first makes this argument concerning the identification of sarņsāra and
nirvāna in MMK 25:19 (see Nāgārjunian Disputations, pp. 1 85-1 86), and then
repeats it in relation to the stanza we are considering here (ibid., pp. 191-194).
1 3 - Garfield, The Fundamental Wisdom of the Middle Way, p. 305.

14 - Ibid., p. 212.

John Spackman 171

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
1 5 - See esp. ibid., pp. 280-282 and pp. 330-331 .

1 6 - As we shall see in the final section, this statement requires some qualification,
since in recent writings Garfield also maintains that Nāgārjuna actually does
make assertions about ultimate reality.

17 -Ibid., p. 213.
18 - Ibid.

1 9 - In "Emptiness and Positionlessness: Do the Mädhyamika Relinquish All Views?"


Garfield writes that dependent existence is "simply the only way in which any-
thing can exist." Dependent existence is a different way of existing from existing
independently, and so presumably our concepts of the two differ as well. See
Jay L. Garfield, Empty Words: Buddhist Philosophy and Cross-Cultural Interpre-
tation (Oxford: Oxford University Press, 2002), p. 51.

20 - The notion of a categorical property, as it is used in these discussions, is not


simply equivalent to the notion of an intrinsic property. Categorical properties
are non-dispositional, but while some regard dispositions as relations, others
hold that they are best viewed as intrinsic properties. For a discussion of these
issues, see John Heil, From an Ontologica! Point of View (Oxford: Clarendon
Press, 2003), chap. 8.

21 - See, for instance, Richard Holton, "Dispositions All the Way Around," Analysis
59 (1999): 9-14; Randall R. Dipert, "The Mathematical Structure of the World:
The World as Graph," Journal of Philosophy 94, no. 7 (1997): 329-358; and
Mary Clayton Coleman, "Could There Be a Power World?" American Philo-
sophical Quarterly 47, no. 2 (2010): 161-1 70.

22 - For versions of this objection, see Heil, From an Ontological Point of View,
pp. 76, 98, 109; Simon Blackburn, "Filling in Space," Analysis 50, no. 2 (1990):
64; Keith Campbell, Metaphysics: An Introduction (Encino, CA: Dickenson
Publishing Company, 1976), p. 93; D. M. Armstrong, Perception and the Physi-
cal World (New York: Humanities Press, 1961), p. 189.

23 - Holton, "Dispositions All the Way Around."

24 - Ross P. Cameron, "Turtles All the Way Down: Regress, Priority, and Fundamen-
tal ity," Philosophical Quarterly 58 (2008): 1-14. For another recent argument
for a similar position, see Michael Esfeld, "Do Relations Require Underlying
Intrinsic Properties? A Physical Argument for a Metaphysics of Relations," Meta-
physica: International Journal for Ontology and Metaphysics 4, no. 1 (2003):
5-25.

25 - See G. W. Leibniz, On the Ultimate Origination of Things, in G. W. Leibniz,


Philosophical Essays, ed. and trans. Roger Ariew and Daniel Garber (Indianapo-
lis: Hackett, 1989).

26 - Blackburn, "Filling in Space," p. 64.

1 72 Philosophy East & West

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms
27 - Joseph Walser, Nagarjuna in Context : Mahayana Buddhism and Early Indian
Culture (New York: Columbia University Press, 2005), p. 234.

28 - Garfield, The Fundamental Wisdom of the Middle Way p. 221 .

29 - Kalupahana, Nagarjuna : The Philosophy of the Middle Way p. 113.

30 - Note that Garfield and Samten's translation represents a shift from Garfield's
earlier translation in The Fundamental Wisdom of the Middle Way where
the lines are rendered in the form of a conditional: "If things did not exist /
Without essence /The phrase 'When this exists, so this will be/ / Would not be
acceptable/'
31 - Kalupahana, Nagarjuna : The Philosophy of the Middle Way pp. 1 1 3-1 1 4.

32 - Abhidharmakosabhasyam, trans. Louis de La Vallée Poussin, English trans. Leo


M. Pruden (Berkeley: Asian Humanities Press, 1988), p. 59.

33 - Garfield, The Fundamental Wisdom of the Middle Way pp. 1 49-1 50.

34 -Ibid., p. 151.
35 - E.g., Walser, Nagarjuna in Context , p. 244.

36 - Garfield, for instance, follows Tsong Khapa in reading the passage in question
thus (Garfield, The Fundamental Wisdom of the Middle Way p. 1 94).

37 - W. V. Quine, "Two Dogmas of Empiricism/' in From a Logical Point of View: 9


Logico-Philosophical Essays (Cambridge, MA: Harvard University Press, 1953).

38 - For an early expression of something akin to this conception of meaning, see


Hilary Putnam, "The Analytic and the Synthetic," in H. Feigl and G. Maxwell,
eds., Minnesota Studies in the Philosophy of Science 3 (1966): 358-397. For a
more recent development of a similar notion of "core" aspects of the meaning
of a concept, see the work of Mark Johnston, for instance "How to Speak of the
Colors," Philosophical Studies 68, no. 3 (1992): 221-263.
39 - Garfield, The Fundamental Wisdom of the Middle Way p. 305.

40 - CandrakTrti: "All this [namely the four positions] is not to be said. However, if
unspoken, knowers are unable to understand own being as it is
(yathãvadavasthitam). So, relying on transactional truth (vyavahârasatya), by im-
putation (āropatah) we say 'empty' with a view to transactional usage by ac-
commodation to people who are to be trained" (Ruegg, "The Uses of the Four
Positions," p. 14).
41 - Graham Priest with Jay Garfield, "Nāgārjuna and the Limits of Thought," in
Graham Priest, Beyond the Limits of Thought (Oxford and New York: Oxford
University Press, 2002), pp. 249-270.
42 - Garfield and Priest, "Nagarjuna and the Limits of Thought," p. 263.

John Spackman 173

This content downloaded from


130.102.42.98 on Sat, 26 Sep 2020 16:07:30 UTC
All use subject to https://about.jstor.org/terms

You might also like