Dharmakīrti's Refutation of Theism

Author(s): Roger Jackson

Reviewed work(s):
Source: Philosophy East and West, Vol. 36, No. 4 (Oct., 1986), pp. 315-348
Published by: University of Hawai'i Press
Stable URL: http://www.jstor.org/stable/1398992 .
Accessed: 11/11/2011 17:21

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

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 [email protected].

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

http://www.jstor.org
Roger Jackson Dharmakirti's refutation of theism

I. INTRODUCTION

Indian civilization, no less than that of the West, is haunted by the concept of
God, and Indian philosophical writing, no less than the works of Aquinas,
Descartes, Kant, or Hume, has as one of its important concerns the existence or
nonexistence of an omniscient, eternal, independent, benevolent being who
creates and/or designs the cosmos. Despite Lin Yutang's description of India as a
nation "intoxicated with God," 1 Indian skepticism about such a being goes back
very far indeed,2 and explicit arguments against theism find an important place in
the writings of Buddhism, Jainism, and Mimamsa (as they must have in the lost
writings of Carvaka), while God's importance or even existence for early Sam-
khya, Nyaya, and Vaisesika is at best moot.3 Indeed, the only Indian philosoph-
ical systems that are explicitly theistic are Vedanta, Yoga, and later, Nyaya-
Vaisesika. It undoubtedly is due to the overwhelming preference for Vedanta
among modern exponents of Indian philosophy that Indian tradition so often is
presented through theistically-shaded lenses, and it is not incorrect to assert that,
in general, Indian civilization has become more theistic during the same period in
which the West has become less so. Still, this should not blind us to the fact that as
recently as five hundred years ago thinkers like the Jaina Gunaratna were
adducing sharp and original arguments against theistic assertions, and that even
today the unanimity of Indian belief in God may not be as thoroughgoing as
most swamis and scholars would have us believe.4
As might be expected, arguments for the existence or nonexistence of the being
variously called purusa, brahman,paramatman, or isvara, or by the name of one
or another sectarian deity, increased in sophistication as methods of philosoph-
ical discussion grew more complex and precise. Sometime around the middle of
the first millennium A.D. a philosophical watershed was reached wherein the
various Indian schools arrived at least at a broad consensus on the criteria for
valid and invalid formal inferences, anumdna, and proper and improper argu-
mentative modes, tarka. In principle, at least, this permitted intersystemic debate
on the basis of commonly accepted "logical" canons, and thus prompted the
hope that arguments on fundamental philosophical issues might indeed be
capable of resolution. In general, before the development of these canons, Indian
philosophical arguments that were not simply dogmatic were analogical or
dialectical in form; arguments after the canons were developed still employed
illustrative analogies and dialectical dilemmas, but within the much more care-
fully articulated framework of what is sometimes called the Indian "syllogism."
Among those contributing greatly to the development of generally acceptable

Roger Jackson is an Assistant Professor of Religious Studies at Fairfield University,Connecticut. This

paper was originallypresented at the seventh conference of the InternationalAssociation of Buddhist
Studies, Bologna, Italy, July, 1985.
Philosophy East and West 36, no. 4 (Oct. 1986). O by the University of Hawaii Press. All rights reserved.
316 Jackson

logical canons was the seventh-century Buddhist dcarya Dharmakirti, who

developed the seminal insights of his great predecessor, Dignaga, into an
epistemological and logical system that itself drew the attention of countless
commentators (not to mention opponents5) and has served as the basis of
epistemology and logic in the Tibetan Buddhist tradition right up to the present.
The majority of Dharmakirti's writings6are concerned with epistemological and
logical questions, but he was not uninterested in matters of religious and meta-
physical doctrine, for the chapter titled "Pramdnzasiddhi," or "Establishment of
Authoritativeness," in his masterwork, the Pramdnavdrttika,7is devoted almost
entirely to a rational justification of Buddhist religious doctrines, such as the
authoritativeness of the Buddha, the reality of past and future lives, and the
validity of the Four Noble Truths. In the course of demonstrating these doc-
trines, Dharmakirti attacks the positions of a variety of non-Buddhist oppo-
nents, including the Lokayatas (= Carvakas), Samkhyas, Nyaya-Vaisesikas,
Mimamsakas, and Jainas. Although earlier Buddhist writers had criticized non-
Buddhist systems, and Bhavaviveka had subjected them to systematic scrutiny
nearly a century earlier in his Tarkajvald,Dharmakirti was the first Buddhist to
criticize non-Buddhist doctrines with fully developed methods of inference and
argumentation at his disposal.
Among the non-Buddhist doctrines criticized by Dharmakirti in the "Pramd-
nasiddhi" chapter of the Pramdnavdrttikawas the assertion that an omniscient,
permanent, independent entity, Tsvara,is the creator of the cosmos. Although
George Chemparathy remarks that "the systematic and thoroughgoing attack
on the Isvara doctrine by Dharmakirti" gave a great impetus to the theist-atheist
controversy,8 and Gopimohan Bhattacharyya notes that the "time-honored
cosmological [sic] argument was for the first time subjected to scathing criticism
by Dharmakirti, 'the central figure around whom all the creative minds in India
revolved',"9 Dharmakirti's arguments themselves, pivotal as they may have
been, have received surprisingly little attention; most writers on Buddhist athe-
ism have focused either on the arguments of such earlier sources as the Pali
Nikdyas, Nagarjuna, Asvaghosa, and Vasubandhu, or the later, extended dis-
cussions in Santaraksita's Tattvasamgrahaand the Panjikdupon it by Kamalas-
ia. The earlier arguments are less systematic than Dharmakirti's, and the later
ones are largely based on the discussion of isvara in the Pramdnavdrttika,so it
seems desirable to examine these crucial arguments, for without an understand-
ing of them, our picture of the Indian theist-atheist controversy will be incom-
plete. This essay will sketch the pre-Dharmakirtidevelopment of theism, outline
earlier Buddhist refutations of it, contextualize and analyze Dharmakirti's argu-
ments in some detail, note some of the directions taken in the theist-atheist
debate after Dharmakirti, and conclude by examining problems inherent in
attempting to "decide" the debate and compare it to similar debates in the
Western tradition.
317

II. PHILOSOPHICALTHEISMBEFOREDHARMAKIRTI

Indian speculation about the cosmos, of course, goes back as far as the later
sections of the .Rgveda,where the first cause is said to be, for example, visvakar-
man ("the all-maker"),10 or purusa ("the person"),11 or prajdpati ("the lord of
creatures"),12or tadekam ("the one").13 The divine power, or supreme purusa,
first is referred to as isvara ("the lord") in the Atharvaveda,14while the Brah-
manas and Aranyakas continue Vedic speculations regarding prajapati and
visvakarmanand introduce the concepts of brahmanand brahmd. 5 In most of
these accounts, the discussion of the first cause is couched in mythological
narrative; little real attempt is made to justify the concepts philosophically,
though lurking in the background are unstated assumptions about limiting
principles and simplicity of explanation. Discussions of the first cause in the
various Upanisads focus on the concept of brahman(also referredto as atman or
paramdtman),whose reality as the source and (most often) the substance of the
cosmos is inferredusually through a reductive process that moves from change to
permanence, multiplicity to unity, complexity to simplicity, materiality to spiri-
tuality, and grossness to subtlety.16The earlier, more "monistic" Upanisads tend
to regard brahmanas an impersonal principle that simply becomes the cosmos
(while at the same time remaining in some way transcendent to it); in later
Upanisads, such as the Svetdsvatara, brahmanis personalized-at least to the
point where it has a creative aspect that is responsible for originating the cosmos
and that can be addressed as "lord" (Isa, isvara), or "deity" (deva),17or even as
Rudra. The Svetdsvatara actually lists non-Vedic explanations of the cosmos,
such as svabhdva(nature), kala (time), niyata (fate), yadrccha (chance), and so
forth, but it rejects them out of hand, simply asserting that brahman,rather than
any of these, is the true explanatory principle.18 The BhagavadgTtdfurther
personalizes the first cause by identifying its ultimate nature with the divine
person of Visnu,19 but, again, the fact is asserted rather than argued, and the
appeal is aimed more at the imagination and emotions than at rationality.
It is only with the development of the classical darsanasin the last centuries B.C.
or the first centuries A.D. that theism, widespread as it had become religiously,
began to receive philosophicaljustification. If we take the term "theism" in the
broad sense in which I am using it-comprising any theory that attributes the
creation and/or ordering of the cosmos to one source, whether personal or
impersonal-then there are three darsanas that can be said to be "theistic":
Vedanta, Yoga, and Nyaya.
The Brahmasutrasof Badarayana did not receive their most important advaita
commentaries until Gaudapada and Safikara, both of whom probably postdate
Dharmakirti,20but the sutras themselves have as one of their central concerns to
establish that brahmanis the source and substance of the cosmos.21 Brahmanis
asserted to be that on which the world is dependent,22 the material cause of all
318 Jackson

effects.23 Badarayana's arguments rest primarily on scriptural statements that

would not carry much force for a Buddhist, since the latter admits neither the
validity of the Vedas nor the existence of an independent sabdapramdna.This
Vedantin disinclination to proffer inferentially based arguments for its theistic
beliefs is seen clearly in later commentators such as Safikara, who denies that
brahman'sorigination of the cosmos ever can be established inferentially, since
brahman is imperceptible and inferences must be perceptually based,24 and
Ramanuja, who refutes various rational arguments for theism so as to pave the
way for knowledge of God through scripture and devotion.25
The Yogasutra of Patanijaliquite specifically asserts the existence of a "su-
preme purusa," Tsvara,who is unaffected by affliction, action, or fruition, is
omniscient, the eternal teacher, and the object of the syllable om, and devotion to
whom is one way to achieve samadhi.26The Yoga tradition generally is more
concerned with psychological than philosophical matters, and its literatureis far
from being replete with rational discussions of the existence or nonexistence of
Tsvara,but Patanijali'sfourth-century commentator, Vyasa, does interpret the
assertion that in isvara "the seed of the omniscient is not exceeded" as meaning
that Tsvara'sknowledge and preeminence are knowledge and preeminence other
than which none greatercan be conceived.27 If differentdegrees of knowledge or
preeminencebe admitted, then a supremeinstance of these is not inconceivable-
and that supreme instance is Tsvara.It ought to be noted that the isvara of Yoga
is not a creator-God like that of Vedanta. Yoga arises within the context of
Samkhya, according to which the cosmos is simply a transformationof insentient
nature, prakrti, from which individual purusas-and the supremepurusa-ever
are utterly separate.
Lying midway conceptually between the immanent brahmanof Vedanta and
the detached, inactive supreme purusa of Yoga is the isvara of Nyaya, who is
neither the material cause of the cosmos (like brahman),nor utterly noncausal
(like the supreme purusa), but, rather, the world's shaper and arranger-its
efficient cause, as it were. It is Nyaya (or, later, Nyaya-Vaisesika) alone among
Indian philosophical schools that seriously proposed to offer proof of the
existence and creative activity of tsvara. Ironically, the passage in Gautama's
Nyayasutra that became the basis of later theological elaboration28is not unam-
biguously theistic-a number of scholars believe that, quite to the contrary, its
discussion of the relation between isvara and the results of human action is
intended as a criticism of theism, that is, that if isvara is posited, then human
action is pointless.29 Be that as it may, by the time ofVatsyayana's Nydyabhdsya
(fifth century30), the moot passage in Gautama is interpretedas a demonstration
that Tsvarais the cause of all effects, and that human action could not have results
withoutthe action of Tsvara.Vatsyayana goes on to define isvara as belonging in
general to the category (paddrtha)of substance (dravya),and in particularto the
substance that is self (dtman),of which it is a special instance, powerful, meritori-
ous, benevolent, and in control of both karman and the material elements.31
319

Prasatapada (sixth century) advanced the discussion still further in his Padarth-
adharmasamgrahaby arguing that tsvara is necessary as the conscious impeller
of the unseen (adrsta) force that regulates karman, and as that which impels
atoms to movement and combination at the end of the universe's dormancy
(pralaya).32
The last important pre-Dharmakirti Naiyayika (although by now it is possible
to speak of Nyaya-Vaisesika) was Uddyotakara, who probably flourished in the
period between Dignaga and Dharmakirti (late sixth or early seventh century). In
his Nyayavdrttika, he goes beyond arguing for Tsvaraon grounds peculiar to
Nyaya-Vaisesika and seeks to establish his existence on more general grounds.
Isvara, Uddyotakara asserts, is the instrumental cause (nimittakdrana)of things,
because he assists beings in reaping the fruits of their actions. ITvarafurther is
necessary as an adjunct to material results, because all results must be preceded
by conscious action, as a hatchet requires a wielder in order to function, or the
flow of milk to a calf requires the cow's intention. Further, although Tsvarais a
permanent, unaltering entity, he can cause impermanent entities, because we see
that, for example, spun yarn, though unmoving, is the cause of a movable
garment. Isvara is not the creator of the eternal atoms that comprise the material
world; his "creation," therefore, is neither ex nihilo nor out of himself. Rather, he
fashions the preexistent "material" into a cosmos in response to the necessary
fruition of the dharma and adharma of beings. Finally, tsvara's power and
consciousness are eternal, omnipresent, and unlimited, for events throughout the
entire extent of space and time require a conscious agent as their instrumental
cause; since that cause, Tsvara,can effect all results, his power is unlimited,
and since he is conscious effector of all results, he must, by definition, be
omniscient.33
III. PRE-DHARMAKIRTIBUDDHIST ATTACKSON THEISM

Beginning with the great eleventh-century defender of theism, Udayana,34 any

number of Hindu writers have attempted to argue that Buddhism, with its
worship of an omniscient tathdgata, actually is crypto-theism. The word
"theism" undoubtedly can be twisted in such a way that certain aspects of
Mahayana theory and practice fall under the term, but it is equally clear that
theism in the sense in which I am using it-as the assertion of an omniscient,
permanent, independent, unique cause of the cosmos-is rejected throughout
the length and breadth of the Indian Buddhist tradition. Dharmakirti's anti-
theistic arguments may have taken the Buddhist critique to a new level of sophis-
tication, but he had behind him a millennium of refutations, with many of which
he undoubtedly was familiar, and which ought to be borne in mind when we
consider his discussion.
The Pali Nikdyas contain a number of explicit rejections of theism, and some
important implicit ones, as well. In the Brahmajalasutta, one of the sixty-two
views discussed by the Buddha is the claim that brahmd is the creator of the
320 Jackson

cosmos; this claim is rejected on the grounds that it is based on a mistaken

inference: at the beginning of a world cycle, brahmdis the first being to arise.
Lonely, he wishes for other beings as companions, and they appear. He con-
cludes that he has created them, but is mistaken, for by the Buddhist explanation
the beings simply are arising due to their own karman-karman, rather than the
will of a deity, being the true creative force in the cosmos.35
A second explicit rejection, made on the grounds of theodicy, or the "problem
of evil," occurs at a number of places in the Nikayas,36whereit is claimed that the
postulation of a God as creator of the cosmos and the regulator of karman
undermines human moral responsibility, while at the same time vitiating claims
that the God can be benevolent, since evils are his creation, too. Other explicit
critiques include mockery of the "omnisicient" brahmdfor his ignorance regard-
ing the sphere wherein all elements cease,37 and skepticism regarding the claims
of some brahmins to have seen brahmaface-to-face.38
For the later Buddhist philosophical tradition, however, the most important
early arguments are perhaps the implicit ones: those many passages in the
Nikayas where the concept of a permanent atta or dtmanis rejected, principally
on the grounds that no permanententity is or can be encountered in experienceor
justified by reason. It really is Buddhism's emphasis on universal impermanence
that is at the root of its aversion to the concept of God, as became evident in the
sorts of refutations offered in the post-nikdya period (when the attributes of the
creator, identified by the Buddhists as Tsvara,perhaps had become more clearly
defined).
Poussin remarks that Buddhist refutations of isvara "ont le tort de se
repeter."39 It is true that certain points are stressed again and again, but the
arguments do vary; indeed, their uniformity is more in style than substance:
virtually all are couched in the form of logical dilemmas, in which the predication
of this or that attribute of isvarais shown to lead to unacceptable conclusions, no
matter how it is qualified. Post-nikdya, pre-Dharmakirti arguments are thus
broadly "logical," without being specifically inferential.
One of the earliest post-nikdyarejections of Tsvarais found in the Buddhacdrita
of Asvaghosa (first-second century A.D.), where at one place the rhetorical
question is posed: If Tsvarais the creator, then what point is there in human
effort?40In a second passage, the Buddha is quoted as pointing out that if Tsvara,
the cause, is perfect and unchanging, then the cosmos that is his effect must be
perfect and unchanging, which it manifestly is not. Further, if it be argued that
Tsvaracreates with a purpose in mind, then he has not achieved all purposes, and
his perfection is limited; whereas if he creates without a purpose, then he must be
regarded as no more sensible than a madman or a child.41
A number of works attributed to Nagarjuna-generally believed to belong to
the same period as Asvaghosa-reject the concept of Tsvara.The Suhrllekha
mentions in passing that Tsvarais not to be accepted as the cause of the aggre-
gates.42 The Catuhstava argues that isvara must either originate from another
321

entity, in which case his uncreatedness is violated, or be self-originated, which is

impossible, since an entity cannot at the same time be both the agent and object
of an action.43 The Bodhicittavivarananotes that (a) if isvara is alleged to be
permanent, then he cannot create, either simultaneously or gradually (since
results are impermanent, and so cannot have a permanent cause); (b) if he is said
to be efficient, then he must create the universe unaided all at once (since
efficiency requires the immediate generation of a result); (c) if he requires as-
sistance in creation, then he is not truly eternal or efficient;and (d) if he is alleged
to be an entity (bhdva),then he cannot be permanent, since entities are observed
to be impermanent.44 Finally, the Visnorekakartrtvanirdkaranam,45 which is
entirely devoted to a refutation of isvara, argues that (a) he cannot create the
existent, since it already exists, nor the nonexistent, since it cannot come to be;
and (b) he cannot be self-originated, as that is a contradictory concept; or other-
originated, for that would entail an infinite regress of creators, even one of whom
exist before Tsvara,thereby vitiating his status as creator.46
Arguments against the concept of tsvara also are found in the Abhidharmika
literature of succeeding centuries. The great compendium of Sarvastivadin
thought, the Mahdvibhdsa,notes that (a) if isvara is the cause of everything, then
he must create everything at once (since efficiency implies immediate causation);
(b) if he requireshelp, then he is not the sole cause; (c) if he is undifferentiatedand
eternal, so must his effects be (since effect must resemble cause); and (d) since
effects are known to be impermanent, their alleged permanent cause, isvara, has
no more "existence" than space.47
Vasubandhu's Abhidharmakosaand Abhidharmakosabhdsyareject tsvara at a
number of places, most extensively in the Bhasya to II,, 64d, which asserts that
the various dharmas do not arise from a unique cause like tsvara, because
dharmasare successive and tsvarais not. Among the points made by Vasubandhu
in his dialogue with a theist are that (a) if tsvara is said to will the successive
generation of dharmas,then he must have multiple desires; if he is single, he must
have a single desire, hence create dharmas all at once; (b) if isvara requires
assistance, then he is not the unique cause, and his assistant causes would require
further assistant causes, in infinite regress;(c) if s'varawills the creation of some
dharmasin the present and some later, then he must be incapable of creating the
later dharmas now, and if he cannot create them now, he cannot in the future,
since his nature does not change; and (d) if the observed causes of various effects
are said to be auxiliaries to isvara's causation, then it must be asked whether
tsvara can prevent the effects from arising-he cannot, and therefore is both
impotent and irrelevant, for the observed impermanent causes are perfectly
adequate explanations for effects.48 Vasubandhu also argues that since karmic
results are multiple, their cause cannot be single,49 and that, similarly, suffering
cannot be originated by a cause that is single, nonsuccessive, or guided by
intelligence.50
Among Mahayana Abhidharmika texts that include refutations of isvara, we
322 Jackson

will mention only the Yogacdrabhumiof Asafiga, which argues that (a) if isvara
has a reason for creation, then that reason is the real cause, whereas if he has
none, then he cannot be motivated to become a cause; (b) if isvarais immanent in
the cosmos, then he cannot stand outside as its creator, whereas if he is not
immanent in it, then he has no relation to it, and so cannot create it; (c) if isvara
creates intending some purpose, then it must be admitted that there is a purpose
he has not yet fulfilled; and (d) if creation depends on Tsvara'swill alone, then
everything must arise simultaneously, while if it depends on an isvara who is
assisted, then he is not the unique cause.51
One final pre-Dharmakirti text worthy of brief mention is the Tarkajvaldof
Bhavaviveka, or Bhavya (sixth century), whose discussion of isvara52 shows at
least a rudimentaryawareness of attempts to prove isvarainferentiallyand of the
pitfalls entailed by those attempts. Bhavaviveka recites a number of the standard
refutations, noting that the multiple events we observe in the world cannot be
asserted to arise from a unique cause, but rathermust be explained as proceeding
from a multiplicity of karmic conditions, and that isvaracannot be held to be any
more real than a sky-flower or a barren woman's son. He does note that one
possible argument for Tsvarais the syllogism, "The eye and so forth exist as
accompanied (that is, caused) by a maker, because they are arrangedlike a pot."
To this Bhavaviveka's response is that the syllogism is invalid because it proves
what is already proven for the Buddhist, namely, that events have causes-for
the Buddhist, however, the causes are multiple (karman, the elements, parents,
and so forth), not a single arranger. A second syllogism, namely, "Isvara is the
maker of the eye and so forth because he is permanent, unique and unproduced,"
is rejected as unproved (asiddha) because of the absence of any corroborative
example of such an entity. Finally, the syllogism, "(The eye and so forth have)
isvara (as a maker preceding them) because (they are shaped), just as a pot has a
potter as its maker," is rejected on the grounds that a potter is (a) embodied and
(b) impermanent, neither of which is applicable to Tsvara.53
IV. DHARMAKIRTI'S CRITIQUE OF THEISM: CONTEXT

As noted in the introduction, Dharmakirti's refutation of theism is found in the

Pramdnasiddhichapter of his Pramdnavdrttika.The Pramdnavdrttikais loosely
constructed as a commentary on Dignaga's Pramdnasamuccaya,and the Pramd-
nasiddhichapter-regarded as the first by modern editors and the second by the
Tibetan tradition54-is itself an elaborate gloss on just one verse of the Pramdna-
samuccaya, the first, wherein Dignaga salutes the Buddha as One Who Has
Become Authoritative (pramdn.abhzuta), One Who Desires to Benefit the World
(jagaddhitaisin), the Teacher (sdstr), the Sugata, and the Savior (tdyin).The basic
purpose of the Pramdnasiddhi chapter to demonstrate the Buddha's authorita-
is
tiveness for those who desire spiritual liberation, through demonstrating that it is
reasonable to regardhim as the Benevolent One, the Teacher, the Sugata, and the
Savior.55 These, in turn, are proven through a series of extended philosophical
323

arguments, the most important of which revolve around (a) defining authorita-
tiveness and giving negative and positive examples of beings who embody it, (b)
proving that positive mental qualities such as benevolence can be developed
infinitely, through demonstrating that the mind-body relation is an interactionist
dualism that permits the existence of past and future lives, and (c) showing that
the Four Noble Truths taught by the Buddha are in fact true, and, especially, that
acceptance or rejection of a self (atman) is the key to samsdraand nirvana.As the
nineteenth-century Tibetan commentator Mi pham notes, the proof of past and
future lives paves the way for proving that the Buddha has the causes for being
regarded as authoritative, while the proof that the Four Noble Truths are true
shows us the reason why he is authoritative.56
It is in the first general division of the chapter, that which defines and exem-
plifies authoritativeness, that the rejection of Tsvarais to be found. After defining
pramdna (authoritativeness) in the first six verses57 as uncontradicted, fresh
cognition, Dharmakirti asserts in verse 7 that the Buddha fulfills this definition.
Before demonstrating generally (as he will in verses 29-33) that the Buddha is
authoritative because he knows what is to be rejectedand what accepted (heyopa-
deya) by those intent on liberation, Dharmakirti provides a "nonaccordant
example" for his definition of authoritativeness. This, of course, is Tsvara,whose
authoritativeness, creatorship, and existence are rejected in verses 8-28. The
argument can be broken down into three general sections: (1) verses 8-9 reject
isvara'sauthoritativeness; (2) verses 10-20 are a refutation of a theistic syllogism
purporting to prove that worldly effects must have a conscious cause, and that
cause is Tsvara;and (3) verses 21-28 question the possibility that isvara could be
a causal agent, through a comparison between the characteristics attributed to
Tsvaraand the reality of the causal process. We will consider each of these
arguments in turn, relying primarily upon Dharmakirti's own words. Where
necessary, we will turn for interpretive help to one of the greatest of Tibetan
Pramanavarttika commentators, rGyal tshab dar ma rin chen (1364-1432),58
and, on occasion, to Dharmakirti's own disciple, Devendrabuddhi (or -mati).59
V. DHARMAKIRTI'SREJECTIONOF ISVARA'S AUTHORITATIVENESS

Dharmakirti already has established (in verses 3b-4b) that authoritativeness is

cognitive (dhT),an act of consciousness. Can one then posit the authoritativeness
of a being whose nature is permanent (nitya)?

(8-9a:)
There is no permanent authoritative (being),
Because authoritativeness cognizes (functioning) entities;
Because, since objects of knowledge are impermanent,
That (which cognizes them) is unstable;
Because that which is generated consecutively
Cannot be accepted as generated from a permanent (cause);
(And because) it is unsuitable that (a permanent cause) depend on conditions.60
324 Jackson

Note that the explicit object of refutation here is simply a permanent authorita-
tive cognition, which could define not only isvara, but such other non-Buddhist
concepts as dtman,purusa, and so forth. In fact, the term Isvaradoes not appear
until verse 28. Still, isvarais probably the primaryobject of refutation throughout
this discussion, for (a) isvara is the only "permanent authority" mentioned
anywhere in the verses, (b) isvara is clearly indicated as the object of refutation
by Dharmakirti's commentators, and (c) isvara had been singled out by pre-
Dharmakirti Nyaya-Vaisesikas as a permanent being who was the creator of all
effects, hence, by definition, authoritative regarding all effects (omnisicient).
Taking isvaraas the permanent authoritative being who is being rejected,then,
we see that Dharmakirti's argument is as follows. That which any authoritative
cognition cognizes are the functioning entities (vastu) that are what is "real" in
the world. Functioning entities are known to be impermanent, that is, to exist
only momentarily. Any entity, therefore, actually is a succession of momentary
events, each following the other with inconceivable rapidity, and constituting a
"thing" only insofar as there is a certain similarity from one moment to the next.
Since it is objects that (conventionally, at least) generate cognitions, a cognition
of an object only can arise where an object exists. If an object exists only for a
moment, its cognition must be similarly momentary, generated successively.
Indeed, a permanent authority said to cognize all entities only could cognize them
simultaneously, for it does not change from one moment to the next. This would
mean that all objects in fact exist simultaneously, which manifestly is not so. If
it is maintained that isvara himself remains permanent, but that his cognitions
are impermanent in accordance with the succession of objects, then at least
two consequences ensue: (a) tsvara is being qualified with contradictory prop-
erties (permanence and impermanence) and (b) he is being accepted as depen-
dent on conditions (the succession of objects), which a permanent being cannot
be.61
It ought to be noted, before we continue, that Dharmakirti's argument here
presupposes the validity of the Buddhist doctrine of momentariness, whereby
"existence" only is predicated of efficient(arthakriya)entities, and efficiencyonly
can be predicated of momentary entities-since an entity that is not inherently
and instantly destructible cannot be destroyed, hence is immutable, and what is
immutable cannot interact with what is successive, as entities manifestly are.
Buddhist arguments for momentariness were highly controversial,62being open
to criticism for (a) vitiating causality by denying continuity and (b) begging the
question by defining existence in such a way (as a particulartype of efficacity)that
only momentary entities could fulfill the definition. It is not my intention to enter
into these fundamental arguments here, but simply to point out that the debate
between Dharmakirti and his opponents is not necessarily self-contained, but
constantly opens out onto the broader metaphysical issues dividing them (and
these issues, in turn, are inextricably intertwined with questions of the religious
psychologies of different traditions63).
325

To continue, having shown that a permanent authoritative cognizer is a

contradictory concept, Dharmakirti goes on to rejectthe notion that isvaracould
be regarded as impermanent:

(9b:)
Because (a permanent cognizer) has not been helped in any way,
There cannot be an impermanent authoritative (being).64

According to rGyal tshab, who follows Prajinakaraguptahere,65 "Because (a

permanent cognizer) has not been helped in any way" is intended as a proof that
isvara cannot be impermanent (hence a cognizer, since cognizers must be imper-
manent), for isvara is said to be self-sufficientand eternally liberated, while what
is impermanent may or may not exist, may suffer or be liberated, in accordance
with helping or hindering conditions. Thus, even if isvara is defined as imperma-
nent, other characteristics attributed to him vitiate that definition.66 As rGyal
tshab rightly notes, Devendrabuddhi takes the line in question as further proof
that a permanentauthoritative cognition is impossible; indeed, he takes it as the
reason why a permanent cognizer cannot depend on conditions, that is, because
he cannot be helped in any way (being, by definition, permanent and self-
sufficient).67If isvara'snot being helped is taken in this way, as further proof that
he cannot be a permanent cognizer, then the line, "There cannot be an imperma-
nent authoritative (being)," stands alone, as a simple assertion that a being
defined as permanent simply cannot be impermanent-although impermanence
is the nature of objects, and so of cognitions, too.
VI. DHARMAKIRTI'SREFUTATIONOF A THEISTICSYLLOGISM

Dharmakirti next addresses himself to a specific formal inference that is alleged

to prove the existence of a creator. First, he sets out the syllogism:

(10a:)
(Because of) intermittence, particular shape,
Efficiency, etc., (a creator exists).68
The unnamed opponent here may be the Nyaya-Vaisesika, for, to our knowl-
edge, of all the Hindu theistic schools, only the Nyaya-Vaisesika had, by
Dharmakirti's time, sought to justify the creatorship of isvara through formal
inference. Furthermore, the syllogism here phrased in rather skeletal form by
Dharmakirti bears a close similarity to the argumentsprofferedby Uddyotakara,
who insisted that worldly results required a conscious motivator, like a hatchet,
whose fashioning and use both point to the intervention of a conscious agent.69
The syllogism also recalls Bhavaviveka's unnamed opponent, who argued that
all results require a creator because they have a specific arrangement, like a pot,
whose arrangement informs us of the existence of a potter.70
Commentarial glosses on Dharmakirti's presentation of the theistic syllogism
make it clear that, in fact, threedifferent reasons are being offered as probative of
326 Jackson

a creator. rGyal tshab, thus, restates the syllogism more fully as follows:

"Worldly environs, bodies and enjoyments are preceded by the mind of a maker,
(a) because they act intermittently, like a hatchet, (b) because they have a
particular shape, like a pot and (c) because they are efficient (in fulfilling beings'
aims), like a battle-axe." From these and other such statements, it is proven that
(abodes, bodies and possessions) have a maker whose mind has preceded them,
and also that that (maker) is Tsvara.7'

The argument from intermittence makes the claim that because entities some-
times function and sometimes do not, their existence must be due to action by a
conscious agent. The argument from particularshape makes the claim that, quite
simply, design implies a designer, and there is a design to entities, so there must be
a designer. The argument from efficiency makes the claim that the observed
efficiency of entities requires that they be preceded by an efficient maker who
foresaw the purposes they could fulfill. Dharmakirti does not turn to the argu-
ments from intermittenceand efficiencyuntil later in his discussion, where he will
reject them as part of his refutation of the causal agency attributed to Tsvara.He
will address himself first and in most detail to the argument from particular
shape.
Dharmakirti's first move in refuting the syllogism, however, is to state gener-
ally the problems it entails:
(1Ob:)
(Here,) either (a) the assertion is already proven, or (b) the example is uncertain,
Or (c) the statement issues in doubt.72

According to rGyal tshab, (a) the assertion is already proven because the
syllogism simply states that "environs, bodies and enjoyments are preceded by
the mind of a maker," and this general concomitance will be accepted by the
Buddhist, too, since, according to the Buddhist, environs, bodies, and enjoy-
ments are preceded by mental karman,hence by "the mind of a maker." One of
the requisites for posing a formal inference is that it seek to prove something not
proven before, so the theistic syllogism is, in its general form, redundant. Fur-
thermore, if isvara in particular is posited as the conscious creator, then (b) the
example is uncertain, because all three examples-the hatchet, the pot, and the
battle-axe-are impermanent entities, which must, therefore, be made or em-
ployed by impermanent beings, whereas the entity to whose existence they are
supposed to point, Tsvara,is permanent. The examples, thus, may be probative of
impermanentcauses, but not of a permanentone. Finally, (c) the statement issues
in doubt, because even if environs, bodies, and enjoyments are preceded by a
maker, there is no guarantee that that maker is TIvara,for Isvarais simply one
possible explanation for the way things are-and not a very promising one, given
that, for example, the entities whose explanation is sought are impermanent and
intermittent, while Tsvarais permanent, and so cannot be intermittent.73
Dharmaklrti now turns to an analysis of the argument from particular shape,
327

that is, that environs, bodies, and enjoyments have preceding them the mind of a
maker (Tsvara),because they have particular shape, like a pot, or a mansion:

(11:)
(If) shape, etc., are proved such as to be
Positively and negatively related to a designer,
An inference from that (shape to that designer)
Is reasonable.74
Much of the force of this statement is derived by implication. According to rGyal
tshab, the main point is this: if, and only if, environs, bodies, and enjoyments are
shaped just as a pot is, can we infer that they are preceded by a single conscious
designer, as a pot is. "Shaped as a pot is" can have two different meanings here:
the arrangement of the material of the pot, and the process whereby that
arrangementwas achieved. By either interpretation, "particularshape" is found
to be a reason that is unproven in the subject ("environs, bodies, and enjoy-
ments"). First of all, it is perfectly self-evident that environs, bodies, and enjoy-
ments do not have the same shape as a pot (or a mansion), and so we cannot
necessarily infer that they have a maker in the same way that a pot does, for it is
entirely possible that different particular types of shapes may have different
particular types of causes generating them. Indeed, a Buddhist will argue that
such things as environs, bodies, and enjoyments actually have preceding them a
multiplicity of mental karmans,not a single creator-designer. Secondly, the mere
fact that a particular shape, a pot, arises from a single conscious designer does
not mean that different shapes or different types of shapes need necessarily arise
in the same way; again, the Buddhist will posit mental karmansas the cause and
will claim that, although consciousness may be involved in producing karmic
effects, conscious design is not. Thus, unless the theist wants to claim, absurdly,
that all entities are shaped just as a pot is, he cannot make inferences about them
that are based on the particular circumstances of the pot.75
Dharmakirti turns now from an examination of "particularshape" to "shape
in general," to see whether it may be probative:

(12:)
(A quality) is proven in an entity by a particular (reason),
(But that) a term similar (to the reason is probative) because of its (alleged) non-
difference (from the reason)
Cannot reasonably be inferred;
(That would be) like (inferring) fire from a grey substance.76
Here, the theist's problem is that if he tries to claim the term "shape" in general
(rather than the particular shape of, for example, a pot) as probative, he has
provided a reason that is too general, and thus unproven in the subject. Granted,
we legitimately apply the term "shaped" to environs, bodies, and enjoyments,
but whereas we are able to infer that a pot or a mansion has preceding it the mind
of some person, because we have observed positive and negative concomitance
328 Jackson

between these objects and a maker, we have not observed such concomitance in
the case of, for example, the particularrealms into which we are born, the bodies
we have, and the environment that we share, with its mountains, seas, and
forests. Thus, a particular designer is proven in the cases of some particular
shapes, but one cannot generalize from this that any object to which the word
"shaped" applies necessarily must have a similar designer, for the sources of the
shapes of differing shaped objects may very well differ. Thus, just as the term
"grey substance" (of which smoke is only one type) is too general to be the basis
of a legitimate inference of the presence of fire in a particular place, so the term
"shape" alone is too general to be the basis of a legitimate inference that all
shaped objects must arise in the same way that some shaped objects do.77
Dharmakirti draws out the extreme consequences entailed by the probative
value of "shape" by pointing out that:

(13:)
If that is not the case, then a potter
May be proven to have made an ant-hill,
Because it has some (similarity) to the shape
Of clay in a pot, etc.78

The example is an interesting one, because it can be read as refuting the probative
value of either a particular shape or the general term "shape," the particular
aspects of shape analyzed in the two preceding verses. First, an anthill-at least
of the Indian variety-has the same shape as a pot. We ought, therefore, to
conclude on the basis of this similarity that it was made by a potter, whereas we
know quite well that it was made by ants. Thus, a similarity in shape does not
imply a necessary similarity in origin. Second, even if an anthill were not shaped
like a pot, the general fact that it is "shaped" has no probative value for the
theistic syllogism. Indeed, if anything, the anthill is a counter-example to the pot,
since it is an instance of a shaped object, yet it is one that we know by observation
to be positively and negatively concomitant with causes that are (a) multiple
rather than single and (b) very possibly unconscious rather than conscious.79 In
short, then, the term "shape" cannot be probative, because specific inquiries into
its meaning and relevance show that a particular shape (for example, a pot)
cannot be probative because not all objects have that particular shape, while
shape in general cannot be probative because different shapes may arise under
different circumstances, as an anthill arises in a different manner from a pot.
In the next six verses (numbers 14-20) Dharmakirti digresses in a direction
that is more of logical than theological interest, and I will pass over this dis-
cussion relatively quickly. Against an opponent's suggestion that the analysis to
which he has subjected the reason in the theistic syllogism is an instance of
karyasama, an overly specific refutation that can rebound upon or have "equal
results" for the refuter, Dharmakirti points out that karyasamaoccurs only when
a legitimate general reason is illegitimately undermined by an overly specific
329

analysis of its details,80 as when the generally valid inference "A conch-sound is a
result, because it arises from effort" is undermined by posing a sophistic dilemma
whereby a sound cannot arise before the effort or as a result of a new, unpre-
cedented effort.81 The refutation of the theistic reason, on the other hand, has
rested on the principle that a reason may be probative of a predicate for a
particular class of objects, but that a term lexically similar but semantically
different from that reason cannot thus be probative; one cannot, for instance,
maintain that words have horns simply because there exists a term, gotva, that
denotes both "cowness" and "wordness" 82-any more than one can maintain
that environs, bodies, and enjoyments are preceded by the mind of a creator,
tivara, simply because they have "shape." 83Dharmakirti drives home the conse-
quences of the theistic syllogism by pointing out that if words alone were
probative, then simply by uttering a word one ought to attain its object, and all
goals would be achieved, and all syllogisms proved, simply through the manipu-
lation of words-which, in fact, arise not from the existence of their referents,
but simply from a speaker's or writer's desire to express them.84
This same type of refutation, Dharmakirti adds, can be applied to the syllo-
gisms of other schools. For instance, the Samkhya assertion, "Buddhiis nonsen-
tient, because it is impermanent," is refuted on the grounds that the Samkhya is
using a reason, impermanence, that he himself cannot really accept, since imper-
manence only is admitted of momentary entities, which buddhi and the other
evolutes ofprakrti are not. The Jaina assertion, "A tree is sentient, because it dies
when its bark is stripped," is refuted on the grounds that the definition of death
being applied by the Jaina is too broad to be admissible by his opponents, and
thus cannot be adduced.85 In both cases, words ("impermanence," "death") are
used improperly, and so the syllogism is vitiated.
In the next three verses (and a supplemental verse found in Tibetan but not
Sanskrit), Dharmakirti goes on to draw from these considerations some general
conclusions about logical reasons. In the first place, the validity of a reason
depends on whether it is generally relevant to the subject and positively and
negatively concomitant with the predicate. If the reason is generally correct, then
it cannot be undermined by overly specific critiques, as, for instance, the fact that
"sound is impermanent because it is a product" cannot be refuted by attempting
to show that sound's relationship to space is unaccounted for.86 Further, even if
a particular term is unproven, if the meaning of the term is proven then the
syllogism is valid, as, for instance, the Buddhist syllogism "Atoms are imperma-
nent because they have aspects (murti)" will be accepted for discussion by a
Vaisesika, who does not admit the wording of the reason "have aspects," but can
supply from his own system a term with equivalent meaning, "is tangible."87
Conversely, even if a word is unmistaken, if the meaning is inappropriate, then
the term cannot be probative, for entities are proven from other entities, not from
words.88 The Tibetan version here adds an example illustrating this last point:
one cannot argue that either (a) "A colored cow is a cow because it is a 'goer
330 Jackson

(jagat)'," or (b) "A baby elephant is an elephant because it is an 'hand-possessor'

(hastin)'," for in each instance the reason is merely an expression used colloqui-
ally to refer to the predicate, a cow or an elephant; in fact, there are "goers" that
are not cows and "hand-possessors" that are not elephants.89By the same token,
there are shapes that presuppose a single, conscious shaper, but there are shapes
that may not. Hence, the theistic syllogism is invalid.

VII. DHARMAKIRTI'SREJECTIONOF ISVARAAS CAUSAL AGENT

In the final eight verses of his discussion, Dharmakirti directly attacks the idea
that isvara can be considered a causal entity, exposing the logical difficulties
involved in the theistic belief in a permanent creator-God. In the course of his
analysis, he refutes the two other reasons that formed part of the theistic
syllogism, that is, the argument from intermittence and the argument from
efficiency.He does not refute them in as much detail as he did the argument from
specific shape, but they are central to his concerns. We will signal those passages
in which they are addressed, since the refutation of the arguments from inter-
mittence and efficiencywill complete the refutation of the theistic syllogism posed
in verse 10.
Dharmakirti first attempts to show that the argument from intermittence
entails a logical contradiction:

(21:)
How, if an entity is a cause,
(But is said) sometimes to be
A non-cause, can one assert in any way
That a cause is a non-cause? One cannot so assert.90
The argument from intermittencestates that the fact that entities sometimes arise
and sometimes do not, that is, are occasional or intermittent in nature, requires
the postulation of a conscious being that serves as their cause at those times when
they arise, and that that being is isvara. Dharmakirti points out, however, that a
being that serves as the cause of intermittententities must, by definition, be a
noncause, too, since (a) an intermittent entity has times of nonproduction, when
its eventual cause is actually its noncause, and (b) at the time when the cause is
generating the intermittent entity, there still are other intermittent entities that it
is not generating, so it serves as the noncause of some entities at the same time as
it serves as the cause of others. (a) Successive causality and noncausality poses a
problem because the causal entity posited by the theist, Isvara,is permanent. He
cannot, therefore, change from moment to moment, and if he is asserted to be
causal, then he must always be causal, and can never become noncausal, for that
would entail a change in nature, an impossibility for a permanent entity. (b)
Simultaneous causality and noncausality poses a problem, because isvara is a
single entity, yet is being furnished with contradictory qualities at one and the
same time. Contradictory properties cannot be predicated of a single, partless
331

entity at one and the same time, and if these properties are reaffirmed,then Tsvara
cannot be single, but must be multiple.91 Isvara cannot, thus, be a creator of
intermittent entities.92
Dharmakirti next turns to a series of problems that revolve around the theistic
contention that Tsvarais the actual empowering cause that gives to the causes we
observe the ability-or efficiency-whereby they yield their results. The first
dilemma entailed by this is that:

(22:)
(If Tsvarais an unseen cause, then) when Caitra is healed
By connection with a weapon or medicine,
Why could not an unconnected post,
Although not cognized, be the cause (of healing)?93
Dharmakirti's attack here is directed at the postulation of an extra causal entity
in situations where we already can provide an adequate account of the causal
process. For instance, it is to Dharmaklrti a well-attested fact that a knife wound
can be healed by medicine or by the knife itself, the latter being an instance of
what Nagatomi calls "homeopathic magic." 94 If we are to posit a further unseen
cause behind the observed causes, then why not claim that an unseen, irrelevant
post be involved in the process?95One invisible entity, Dharmakirti implies, is
really no more absurd than another, and the postulation of any such entity tends
to make a mockery of our attempts to understand causality, for the implication is
that anything may be posited as the cause of any result.
Dharmakirti presses the attack, pointing out further problems in the concept
of Tsvara:

(23:)
One whose nature does not vary
Is unsuitable as a creator;
Since a permanent (entity) never is absent,
Even if it has the ability (to be a cause), it is difficult to see.96
The first half of the verse is, in a sense, a reiteration of a fundamental and
recurring argument, namely, that a permanent entity cannot be posited as the
cause of impermanent entities, since (a) the entity is asserted sometimes to be a
noncause, and its nature cannot change, so it cannot become a cause;97 (b)
causality is a process that involves intermittence, and a permanent entity cannot
be intermittent, since intermittence involves a change in nature; and (c) if a
permanent cause is posited, then causality cannot be an intermittent process, but
must occur all at once, since a permanent cause could not alter so as to produce
entities in a second moment. The second half of the verse raises still another basic
objection, namely, that if isvara is the unseen cause of every result, then he must
be ubiquitous and can never be absent. A cause, however, is defined as that in the
absence of which a result does not arise, so an entity that never is absent cannot
meaningfully be described as a cause. Indeed, whether Tsvarais only intermit-
332 Jackson

tently present (as argued earlier) or ubiquitous (as his nature would seem to
dictate) seems to have little actual bearing on our analyses of causality, which, in
fact, turn on the presence or absence of certain observablefactors. One may, if
one wishes, posit an extra entity such as isvara as the cause behind observable
causes, but positive and negative concomitance can only be observed with regard
to the observed causes. Since observedpositive and negative concomitance is an
adequate basis for the explication of any causal situation, the extra "behind-the-
scenes" cause (whether principal or assistant) must be either redundant or
impotent.98
There is a furtherconsequence of the postulation of an invisible cause-behind-
the-scenes, namely:

(24:)
When some (cause) exists, some (result) comes to be;
If some cause other than that
Is supposed, then there will be no end
To the causes of any result.99

rGyal tshab sums up this point succinctly by remarking that:

There would follow an infinite regress of causes for every result, because then it
would be acceptable to think that when some cause assists a result, the cause of
the result is something else, which we do not see as being able to generate some
result.10

Once again, then, the postulation of an unseen cause destabilizes our notion of
causality, for the admission of unseen and unseeable causes opens the door to an
infinity of such causes, which is tantamount to causal chaos. Here, it ought
perhaps to be added in all fairness that the Buddhist notion of karman can be
subjected to the same general critique as isvara. Karman is certainly neither
permanent nor ubiquitous in the way that isvara is, but it is an unseen causal
factor that is operative in virtually every situation in which sentient beings are
involved. In those instances where other causes can be adduced, karman is
superfluous, unless we insist that there be a moral explanation for everything;
while in those instances where we do not have adequate explanations, karman
serves somewhat the way the "God of the gaps" does in Western theology, that is,
as a stop-gap explanation where observable concomitances have not yet been
established. Karman, like isvara, explains so much that it threatens to explain
nothing at all.
In the final four verses of the section, Dharmakirti responds to some possible
objections to his arguments, thereby clarifying his notion of the causal process
and tsvara'sunsuitability for participation in it. The first objection, as supplied
by rGyal tshab, claims that, "... according to you, when soil, etc., do not
generate a sprout, they cannot change their nature, so there will be no generation
of a sprout." 101In other words, if entities cannot change their nature from that
of noncause to cause, then soil, moisture, sunlight, and the seed itself, which are
333

not at this moment generating a sprout, will never be able to. Dharmakirti's
response is that:

(25:)
In the generation of a sprout, the soil and other (conditions)
Do change their nature
And become causes, for when we see that (cultivation)
Is done well, (the harvest) is excellent.102

The implicit point here is that it is only a permanent entity, such as isvara, that
cannot become a cause once it has been a noncause; such causal conditions as
soil, moisture, sunlight, and the seed are all impermanent, momentary entities, so
there is no contradiction in asserting that at one moment they are noncauses and
at another moment they are causes. Indeed, such must be the case, for we observe
that soil and the other conditions do serve as noncauses at one time (early in the
season) and as causes at another, later time (harvest).103Conversely, it might be
added, the fact that it clearly is the case that entities can change from noncauses
into causes is a further demonstration of their necessary impermanence, since a
permanent entity could not thus change.
Dharmakirti next entertains and answers a related objection:

(26:)
If you say, "Just as object and organ,
Meeting without alteration, cause cognition,
So, too, (Isvarais a cause without alteration,)" it is not so.
Because there is alteration (of organ and object) from when (they have not
met).104
The objector here evidently is a Nyaya-Vaisesika, for the account of cognition
being offered derives from the Nydyasutra,l05 where it is said that cognition
results from the contact between an organ and an object. The claim is made that
just as organ and object do not perceptibly change from one moment to the next,
and yet in the first moment there is no cognition while in the second there is, so
isvara, although he does not change, can be a noncause one moment and-our
nonperception notwithstanding-a cause the next. The Buddhist has two pos-
sible responses, one doctrinal, the other logical. The doctrinal is that the Nyaya-
Vaisesika account of the cognitive process is incorrect, and that there is a third
factor that determines a cognition, namely, a previous cognition, whose presence
or absence and particular qualities must be posited to explain the evident fact
that-even if organ and object are admitted not to vary-cognitions do vary.106
Alternatively, if the opponent be taken as accepting the Buddhist postulation of
threeconditions for cognition, then the logical objection can be made that, at the
very least, the organ must vary, for otherwise we could not explain the relative
clarity or dullness of cognitions.107
A further logical objection, of course, is simply that the postulation of an
entity's noncausality at one time and causality at another requiresthat there be
334 Jackson

an alteration, because between a cause and a noncause there is a diference, a

difference that can only be explained by positing an alteration in nature. Thus,
Dharmakirti adds:

(27:)
(Factors) that are individually powerless (as causes),
If they do not change their nature,
Will be powerless even when they meet.
Thus, alteration is proved.108

In the instance of the organ, object, and cognition, the three factors are con-
sidered individuallyunable but collectively able to generate a cognition. Dhar-
makirti's point, however, is that regardless of whether they are functioning
individually or collectively, the three factors cannot be causally potent if it is not
admitted that their nature changes-for the simple reason that previously they
have been a noncause, and in order to be efficacious, they must change in nature
so as to become a cause. Thus, according to rGyal tshab:

... it is proven that the three conditions have different natures when they have
met and when they have not met, because we see the differencethat they generate
or do not generate sense-cognition when they have met or have not met.109

Dharmakirti concludes his refutation with a final observation of the incompat-

ibility between the concept of cause and the concept of isvara-here named for
the first time as the object of refutation:

(28:)
Thus, those (factors) that are individually powerless
(But bring about) the existence of the quality (of the result) when they have met
Are causes; [svara, etc.,
Are not (causes), because they do not alter.110

Causality, then, is a process entailing not only the presence or absence of certain
factors (whereby, as we saw, s'varacould not be considered a cause), but also the
alteration of those factors in such a way that they change from being noncauses
to being causes. Thus, the generation of a sprout requires (a) the presence of
certain factors that might not be present, that is, the seeds, soil, moisture,
sunlight, and so forth, and (b) the alteration of the nature of each of these so that
their individual causal nonefficacity becomes their collective efficacity. Similarly,
a sense-cognition requires (a) the object, organ, and preceding cognition and (b)
an alteration of each of these such that individual nonefficacity can become
collective efficacity. Now, this alteration is not some superadded process beyond
the meeting of the conditions, but it must be specified as part of the causal
process, for without such specification, one might overlook the ontological
difference that is entailed by causality. Difference, in turn, requires imperma-
nence, for entities are known sometimes to be causes and sometimes to be
noncauses of particular events, but it is contradictory that they be both at the
335

same time, while a permanententity, like isvara, cannot alter its nature, and so it
must always be a noncause or always a cause. If it is always a noncause, then the
discussion is academic; if it is always a cause, then it must be ubiquitous, and it
cannot be accepted as a cause, because its presence or absence cannot be
observed to make a difference in the generation of a result.
Before leaving Dharmakirti, we ought to note that he adduces one additional
argument against 7svara,later in the Pramdnasiddhichapter, when he insists
(following the line of reasoning of such predecessors as Vasubandhu111) that the
cause of the various sufferings experienced by beings cannot be a unique cause,
because the variety among results permits us only to infer a variety of causes, and
because, as has been demonstrated, a permanent cause cannot be proven to
exist.112 rGyal tshab, finally, adds his own version of the argument from evil, at
the end of his discussion of the vital and trivial characteristicsof omniscience that
may be attributed to "authoritative" beings: "If someone who can make any-
thing because of his knowledge of the sciences is omniscient, then he also has
made the sufferings of the lower realms...." With this in mind, rGyal tshab
concludes, we should turn not to such a being, but to "someone who, having
accomplished the elimination of every last fetter, is omniscient regarding how all
objects really exist." 113 The latter sort of being is one who truly is authoritative
for those intent on liberation, and is, of course, exemplified by the Buddha, who
has not made the world, but knows it, and knows the way out of it.
VIII. THE THEIST-ATHEISTDEBATEAFTERDHARMAKIRTI

Dharmakirti's attack on theism was a stinging one, but it did not end the debate
between theists and atheists any more than did Hume's critique in the West.
Indeed, as noted earlier, Dharmakirti's discussions had the salutary effect of
raising the discussion to a new level of sophistication, and in the centuries
following him the issue was joined not only by Nyaya-Vaisesikas responding to
his attacks, but by still other Buddhists, as well as by Mimamsakas and Jainas.
These debates have been covered well elsewhere,114and we have neither the need
nor the space to outline them in detail. We will, however, survey them briefly.
The Nyaya-Vaisesika response to Dharmakirti's critique was far from
immediate. Indeed, it was nearly three centuries after Dharmakirti, in the
Nyayamafijarl of Jayanta Bhatta, that a counterattack finally was mounted. In
the meantime, further critiques of theism had been forthcoming, not only from
Buddhists, but from Mlmamsakas and Jainas, as well. The first important post-
Dharmakirti Buddhist attack on theism is that of Santideva (eighth century),
who criticizes a number of non-Buddhist views of causality in the ninth chapter
of his Bodhicaryivatara. Among these is the Naiyayika claim that isvara, a
divine, pure, permanent, single creator, is the source of everything. But, notes
Santideva, if tsvara is identified with the elements that are accepted as the
material causes of material things, there is a contradiction, because these ele-
ments are neither pure nor permanent nor single. On the other hand, if he is said
336 Jackson

to be the creator of the permanentpaddrthasthat constitute the world according

to Nyaya-Vaisesika, then there is a problem, because permanent entities cannot
have an origin, while if worldly phenomena are granted impermanence,then they
cannot be accounted for by a permanent, single entity.115 The remainder of
Santideva's argument recapitulates earlier Buddhist analyses of the problems
entailed by Tsvara'spermanence, his need for assistants, and his creation with or
without a desire to do so.116

A century later, in the Tattvasamgraha, Santaraksita criticized creation

theories centering on both isvara and purusa1 7-though the characteristics of
purusa are not those of the Samkhya purusa, but of the Vedantin brahman.
Together with Kamalasila's Paijika, the Tattvasamgrahaprobably is the most
detailed extant Buddhist critique of theism. Much of the section on isvara
recapitulates and expands upon Dharmakirti's refutation of the theistic syllo-
gism, although Santaraksita does add points of his own. For example, to
Aviddhakarana's claim that the simultaneous functioning of two senses must be
explained by recourse to a conjunctive substratum and that, by analogy, so must
the combinations of the world be explained by the concept of isvara, he replies
that it is unproven either that there can be two simultaneously functioning senses,
or that the category of "conjunction" (samyoga) is admissible.118Santaraksita
also points out that isvara cannot be the source of a verbal revelation, for the
simple reason that he has no body, hence no mouth, and verbal communication is
dependent on the existence of a mouth.119 The critique ofpurusa centers on the
dilemma posed by purusa's (a) motives (if he is motivated by another, he is not
self-sufficient;if he is motivated by compassion, he must create a perfect world,
while if he cannot create a perfect world, he is not powerful; and if he is motivated
by "amusement," then he is both cruel and dependent on the instrument of
amusement, namely, the cosmos)120 and (b) potency (if he is able to create all
things, he must do so immediately, for potency entails immediate generation).121
Attacks on theism also were launched by the two great theoreticians of
MTmamsa,Prabhakara and Kumarila (seventh-eighth centuries). Motivated in
part by their idiosyncratic concern to show that the Vedas are without an author
(which isvara sometimes was said to be), the Mimamsakas adduced some refu-
tations that overlapped those of the Buddhists, and others that were unique. Of
note among the latter were arguments that raised questions of whether isvaracan
be said to have a body or not: we know that creative agency within the world
requiresa body. If isvarais to be proved by analogy to worldly creativity, he must
have a body, yet he is claimed by Nyaya-Vaisesika tradition to be bodiless-
although we know that will alone cannot generate results:some physical agency
is required.If isvarais admitted to have a body, then various consequences ensue:
for instance, if isvarahas a body, whence has that body come? If it is from another
creator, then that creator's body must have a creator, and so on, in infinite
regress;if from himself, then he must have had a body with which to create that
body, which must have had a preceding body-again, there is an infinite regress.
 337

Further, of what could isvara's body be made? It cannot be made of material

elements, because they have not been created yet, while it cannot be immaterial,
because the immaterial cannot be the cause of the material.122 Jaina critiques
of theism, as in the eighth-century Saddarsanasamuccayaof Haribhadra, the
thirteenth-century SyddvddamanjarT of Mallisena and the fifteenth-century Tar-
of
karahasyadFpikd Gunaratna, are easily as sophisticated as those of the Bud-
dhists and Mimamsakas, and open some interesting areas of discussion, but
cannot detain us here.123
As noted above, the first concerted Nyaya-Vaisesika counterattack is con-
tained in the tenth-century Nydyamaijarl of Jayanta Bhatta, who states the
theistic syllogism in the following form: Tsvaraexists because he produces a result
(the cosmos) of a type that presupposes a maker who knows the process and
motive of production, like a potter. Jayanta considers at least twelve different
arguments against the existence of isvara, most of them familiar, such as the
inadequacy of the potter analogy, the problems entailed by Isvara'sembodiment
or bodilessness, questions of motive, and the superiority of "impersonal" expla-
nations, such as karman. Jayanta sets out to demonstrate that his various oppo-
nents' disproofs are themselves riddled with logical flaws. The assertion, for
instance, that natural objects do not necessarily have a conscious designer is itself
uncertain, and thus cannot be adduced as a good logical reason refuting the
theistic reason, while the theistic argument by analogy is valid because in those
instances where we have observed an object's source of design, that source has
been a conscious designer. Thus, all effects can be deduced to arise from a
conscious designer, including the world itself. Jayanta reasserts Tsvara'snon-
corporeality, maintaining that his will can activate physical results in the same
way that the human will can activate the body; in either case, an immaterial
agency does have material effects. Isvara's compassion is justified by explaining
that he creates, for example, hell only as a sort of "holding-cell" for beings until
their karman permits their salvation. Finally, the view, for example, that "col-
lective" karman rather than a single designer is the cause of the natural environ-
ment is rejectedon the grounds that human responses to the environment are too
varied (some people love the mountains, others do not) to enable us to posit such
karmic "cooperation." 124
Other Nyaya-Vaisesika defenses of theism included those of Vyomasiva's
tenth-century VyomavatT,which reiteratesthe point that an effect presupposes an
intelligent designer, and reaffirmsthat the cosmos presupposes a powerful and
omniscient designer,125 and Vacaspatimisra's tenth-century Nydyavdrttikatdt-
paryatlkd, where it is argued that the law of parsimony (ldghava)requiresthat the
creation of the various entities of nature be attributed to one, rather than a
multiplicity of, divinities, and that such a divinity must be inconceivably power-
ful and knowledgeable to be able to effect such a creation.126
The Buddhist position was reaffirmed in the eleventh century by Jfinas-
rimitra, whose Isvaravdda is in part an expanded commentary on some of
338 Jackson

Dharmakirti's discussion, and by Jfianaasrisdisciple, Ratnakirti, in his Isvarasdd-

hanadusana.127 It was in response to Jiananrimitra's attack, described by
Chemopathy128 as the most thorough since Dharmakirti, that the last great
Nyaya-Vaisesika defense of theism, Udayana's Nydyakusumainjali,was com-
posed. Udayana's work is complex and sophisticated enough to have been the
subject of a number of scholarly monographs,129 and I will not discuss it here,
pausing only to note that it includes detailed attempts to refute other schools'
attacks on isvara, and sets out two series of positive proofs, the first (consisting
of nine proofs) demonstrating isvara's existence, and the second (also nine
proofs) demonstrating his authorship of the Vedas. The first series of arguments,
though a brilliant synthesis, does not add a great deal to earlier Nyaya-Vaisesika
discussions; the second series is quite original, but is directed primarily at the
Mlmamsakas, and would be considered irrelevant by a ndstika such as a Bud-
dhist or Jaina. The Nydyakusumdnjaliitself stimulated counterattacks, from the
Vedanta school of Ramanuja and, eventually, in the last great classical work
of the theist-atheist debate, Gunaratna's TarkarahasyadTpik.130
IX. CONCLUDING REMARKS

By way of conclusion, I want to address myself briefly to two somewhat broader

questions that naturally emerge from our considerations of the theological
disputes engaged in by Dharmakirti and other Indian philosophers. The first
question is: To what degree are the arguments of Dharmakirti (or, for that
matter, any of his supporters or antagonists) philosophically conclusive within
an Indian frame of reference?The second question is: To what degree can these
Indian theological discussions be transposed onto the atheist-antitheistdebate as
it has unfolded in the West?
In principle, the various Indian philosophers who argued back and forth about
the existence of savaraaccepted a common set of rules for their discussions, and
so deciding who was right and who was wrong ought to be a simple matter of
seeing who begs the fewest questions and who constructs syllogisms with the
most care. Such decisions only can be simple, however, if (a) the rational
structuresdevised for discussion are themselves foolproof and (b) the disputants
do not import any idiosyncratic doctrinal notions into intersystemicdiscussions.
In point of fact, however, (a) the reliability of formal inference-either in
principle or, at least, for deciding metaphysical questions-was attacked even
from within the Indian tradition, by such thinkers as Nagarjuna, Safikara,
Jayarasi and Purandara, who claimed either that the positing and structuringof
pramdanascould not themselves be supported by any pramdna without begging
the question, or that inference, even if accepted as provisionally valid, could not
inform us on matters forever beyond perceptual ken. Further, (b) very real
differences in the ways in which different schools approached philosophical
problems tended invariably to color even the most carefully "depersonalized"of
339

arguments. Indeed, I think that discussions of the existence or nonexistence of

Tsvaraserve as a good example of the inevitability of such coloration.
If we strip away the almost bewildering variety of arguments we have re-
viewed, we find at bottom two basic issues on which-to take our two main
antagonists-the Nyaya-Vaisesikas and Buddhists have disagreed: (1) the
existence of a permanent entity and its relation to the impermanent and (2) the
requirement that causal action entail a conscious agent. Many complex philo-
sophical discussions turn on these two issues, yet it might be argued that the
attitudes toward each entertained by each of the schools is, in fact, pre-
philosophical, and thus not essentially amenable to revision on the basis of
rational considerations.
(1) The permanence-impermanence issue is one that goes very far back and
very deep in the Indian tradition. Much of the religious and philosophical search
that produced both the Upanisads and Buddhism was geared toward the discov-
ery of an immutable state that was free from the vicissitudes of samsara, yet in
searching for and explicating this state, Hindu and Buddhist schools arrived at
very different conclusions. Hindu schools, of which Nyaya-Vaisesika is one,
concluded on the basis of religious experience and logic that the impermanent
entities we see around us must in some way be subsumed or limited by an eternal
substance that provides their continuity, the continuity that we know to be the
basis of order in the cosmos. Buddhists, on the other hand, concluded on the
basis of theirempirical and logical explorations that there is not, nor could there
possibly be, a permanent substance, for such a substance can neither change itself
nor interact with the impermanent. Thus, the Nyaya-Vaisesika (or Vedantin)
insistence on the necessity of permanence to explain continuity, and the Buddhist
insistence on the necessity of impermanence to explain change are deep-seated
and seemingly irreconcilablepositions, and, much like a Kantian antinomy, each
seems logically to exclude the other and yet, when taken alone, to lead to
insuperable difficulties. To the degree that the dispute over permanence and
impermanence is one of the core issues in discussions of isvara, those discussions
may be impossible to resolve.
(2) The question whether cause-and-effect requires a conscious agent also
seems rooted in prephilosophical decisions that commit the Nyaya-Vaisesikas
and Buddhists to irreconcilable positions. Here, I think the problem may be the
source for the model of causality that each school constructs. Buddhist medi-
tation and Buddhist logic tend to be radically depersonalized, that is, to de-
construct personal notions into impersonal processes much like those we observe
among nonsentient entities. Thus, causality is not even a "personal" process on
the level of the sentient individual, who actually is a nexus of impersonal forces,
some material and some mental; and, needless to say, nonsentient entities do not
requirea personal agent, either. The Nyaya-Vaisesikas, on the other hand-their
"atomism" notwithstanding-tend to draw their model of causality from human
340 Jackson

activities:movement of the body is preceded (usually) by a conscious intention, a

pot by a potter, a house by an architect and builders. By analogy, then, we
conclude that other objects in nature-whose sources we do not know-must
also arise through personal agency, and so, by extension, must the overall
arrangement of the world; an agent responsible for the overall arrangement of
the world must be a vastly powerful and knowledgeable being, such as isvara.
Once again, the antagonists seem to have arrived at completely antithetical
positions by beginning from different places, and it is difficult to see where a
common ground could be found.
Thus, Indian arguments over the existence or nonexistence of [svarahave their
inherent fascinations, and yet we must remain aware that they may not be finally
soluble, for the simple reason that, despite their agreement on the meanings of
many terms, the disputants have vastly different approaches to some basic
problems, and this disparity of approaches threatens to render the arguments on
which they are based forever inconclusive.
Let us turn, then, to the second general question with which we began this
section, that of the applicability of the Indian discussion to the Western debate
over the existence of God. One must, needless to say, be very cautious in
entertaining such comparisons, for concepts that seem identical in two different
cultural-philosophical traditions more often than not are revealed on closer
examination to be quite different,both in denotation and connotation. Certainly,
the frequent translation of isvara as "God" seems at first blush to be legitimate,
for are not the basic characteristics of isvara-permanence, omniscience, inde-
pendence, creatorship, compassion-very much like the attributes of the Chris-
tian God? A closer examination, however, reveals that there are considerable
differences between the Christian God and most of the Indian models. The
brahmanof most Vedantin schools, for instance, transforms itself into the world,
is the world's material cause, whereas the Christian God does not become the
world, but, rather,creates it ex nihilo,and remains forever transcendentto it. The
paramapurusathat is isvara in the Yoga system does not create the world, or
arrange it, or relate to it in any way, whereas the Christian God does all three.
The tsvaraof Nyaya-Vaisesika does not create the eternalpaddrthasthat consti-
tute the world, although he does arrange them into the cosmos that we know,
whereas the Christian God creates both the "raw material" and the arrangement
of the cosmos.
These considerations, in turn, must be weighed when we decide whether or not
to describe the Nyaya-Vaisesika syllogism rejectedby Dharmakirti as simply an
Indian version of the "argument from design" (or teleological argument) for
God's existence. 31 The Nyaya-Vaisesikas, after all, were talking about one type
of "God," while the God asserted by, for example, Aquinas and Newton and
rejected by, for example, Hume are, in fact, very different; furthermore, the
arguments arise in differentcontexts and are conducted in differentphilosophical
languages.
341

All these points are well taken, but they do not totally undermine the com-
parison. To begin with, the "languages" of Indian and Western philosophy are
different, but that does not mean that there is not a fair degree of translatability
across traditions: the inductive and deductive processes generally accepted to be
the basis of sound reasoning are found in both, as are many of the same notions
of the types of flaws that may vitiate arguments. Secondly, even if there are
differences between isvara and God, they are not so great as to obviate all
comparison between their roles and the arguments for their existence. The
argument from design, after all, simply attempts to show generally that the order
we perceive in the cosmos presupposes a single conscious designer and/or sus-
tainer of that order. It really is a secondary matter (pace Kant) whether the being
responsible for the cosmic order creates ex nihilo or arranges preexistent raw
material; in either case, it is not mode of ordering that is at issue, but the existence
of a single eternal being who is the conscious agent of that ordering.
Thus, I think it is fair to call the Nyaya-Vaisesika syllogism rejected by
Dharmakirti an "Indian argument from design," just as I think it is relativelyfair
to call s'vara"God." Therefore, I think that the sorts of arguments proffered by
Dharmakirti and his opponents can be of interest to Western theologians. The
precise ways in which the Indian arguments overlap or deviate from the Western
ones must be the topic of another study, as must detailed considerations of
whether the Indian tradition has arguments that could serve either theists or
antitheists in the West.132 Hume, Kant, and others have given fairly thorough
treatment to the problem of conscious agency, and it is my suspicion that it is on
the permanence-impermanenceissue that the Indian tradition may have the most
to contribute. The Buddhist critique of a God believed to be immutable seems to
me an acute one, and the price of accepting God's mutability a high one, that is,
his susceptibility to conditions, hence loss of omnipotence. There are, of course,
currents in modern Christianity, represented by, for example, Hartshorne, or
Kazantzakis (in his The Saviors of God), that do not require the omnipotence of
God, and admit his dependence on his creatures for the fulfillment of his ends.
These would escape the objections raised against a permanent, independent God,
though whether they could evade criticisms aimed at the concept of divine
teleology (especially those regardingthe admissibility of extraneous causes), I am
not so certain. It also is my suspicion, alas, that cross-cultural debates may in the
end be no more conclusive than intra-cultural ones have been, and that the
arguments, if examined carefully enough, will be seen to rest on prephilosophical
choices and assumptions that cannot really be questioned, and yet which vitiate
the certainty to which philosophers forever aspire.
This final note of uncertainty can, if we permit it, grow into a more general
uncertainty about the "order" we perceive and discover in the cosmos, an order,
incidentally, that was assumed by both theists and atheists in India, their only
disagreement being over how to account for it. Is it not, in fact, possible that this
order simply is not there, that it actually is conceived and invented rather than
342 Jackson

perceived and discovered?This is the possibility entertained in a modern master-

work from Bologna, Umberto Eco's TheName of the Rose. At the end, the monk-
detective protagonist, William of Baskerville, bemoans to the book's young
narrator, Adso, that his "solution" of a series of crimes has been purely acci-
dental, and implies thereby a sort of "argument from no-design":
"I arrived at (the killer) pursuing the plan of a perverse and rational mind, and
there was no plan, or, rather, there began a sequence of causes, and concauses,
and causes contradicting one another, which proceeded on their own, creating
relations that did not stem from any plan. Where is all my wisdom, then? I
behaved stubbornly, pursuing a semblance of order, when I should have known
well that there is no order in the universe."
"But in imagining an erroneous order you still found something...."
"What you say is very fine, Adso, and I thank you. The order that our mind
imagines is like a net, or like a ladder, built to attain something. But afterward
you must throw the ladder away, because you discover that, even if it was useful,
it was meaningless...."
"You have no reason to reproach yourself: you did your best."
"A human best, which is very little. It's hard to accept the idea that there
cannot be an order in the universebecause it would offend the freewill of God and
His omnipotence. So the freedom of God is our condemnation, or at least the
condemnation of our pride."
I dared, for the first and last time in my life, to express a theological conclusion:
"But how can a necessary being exist totally polluted with the possible? What
difference is there, then, between God and primogenial chaos? Isn't affirming
God's absolute omnipotence and His absolute freedom with regard to His own
choices tantamount to demonstrating that God does not exist?"
William looked at me without betraying any feeling in his features, and he said,
"How could a learned man go on communicating his learning if he answered yes
to your question?" 133

NOTES

1. Lin Yutang, ed., The Wisdom of Chinaand India (New York: Modern Library, 1942), p. 11.
2. Compare such "skeptical" Vedic passages as Rg (Rgveda) II, 12, 5; IV, 18, 12; and VIII, 100,
3; and their discussion in Depibrasad Chattopadhyaya, Indian Atheism (Calcutta: Manisha, 1969),
pp. 32-43. Chattopadhyaya's book, while occasionally straining for evidence that one or another
ambiguous passage is atheistic, presents overall a compelling picture of the pervasiveness of atheism
in Indian philosophical (if not religious) thought-especially early in the tradition.
3. Compare Chattopadhyaya, Indian Atheism, chaps. 9 and 16.
4. Compare, for example, ibid., chap. 14; and Narendranath Bhattacharyya, Jain Philosophy:
Historical Outline (New Delhi: Munshiram Manoharlal, 1976), pp. 93-108.
5. Compare, for example, Nagin J. Shah, Akalahka's Criticismof DharmakTrti'sPhilosophy:A
Study, L. D. Series no. 11 (Ahmedabad: L. D. Institute of Indology, 1967);and D. N. Shastri, Critique
of Indian Realism: A Study of the Conflict Between the Nyaya-Vaisesika and the BuddhistDignaga
School (Agra: Agra University, 1964).
6. The Tibetans attribute seven works to him: Pramanavarttika(hereaftercited as PV), Pramd-
n.aviniscaya,Nydyabindu,Hetubindu,SambandhaparTkisd, Samtdndntarasiddhiand Vadanydya.Only
the PVand Nyayabinduare completely extant in Sanskrit;the others exist in Tibetan translation. For
a list of editions and translations, compare A. K. Warder, Indian Buddhism,2d ed. (Delhi: Motilal
Banarsidass, 1980), pp. 539-540.
 343

7. Complete editions include: Dwarikas Shastri, ed., PramdnavdrttikaofAcharya DharmakTrti,

with the Commentary" Vritti" ofAcharya Manorathdnandin(Varanasi: Bauddha Bharati, 1968); and
Y. Miyasaka, ed., Praminavdrttika-Kdrikd(Sanskrit and Tibetan),in Acta Indologica2 (1971-1972),
3 (1973-1975), and 4 (1977). Various parts of the Svdrthdnumdnachapter have been translated;
compare Warder, Indian Buddhism, and Leonard Zwilling, DharmakTrtion Apoha (Unpublished
diss., University of Wisconsin, 1976). The Pramdnasiddhichapter (PS) was translated by Masatoshi
Nagatomi in A Study of Dharmaklrti's Pramdnavdrttika:An English Translationand Annotation of
the Pramdnavdrttika,Book I (Unpublished diss., Harvard University, 1957). I translated rGyal
tshab dar ma rin chen's tibetan commentary on the PS chapter as part II of my dissertation, Is En-
lightenment Possible? An Analysis of Some Argumentsin the BuddhistPhilosophical Tradition, With
Special Attention to the Pramdn.asiddhiChapter of DharmakTrti'sPramdnavdrttika(Unpublished
diss., University of Wisconsin, 1983). This annotated translation, revised, will be issued in 1986 as
Mind, Body, Selflessness, Freedom:DharmakTrti'sDefense of the Buddhist World-View as Expounded
in rGyal tshab's"Elucidatingthe Path to EnlightenmentAccordingto the 'Pramdnavdrttika"'(London:
Wisdom Publications).
8. George Chemparathy, An IndianRational Theology:An Introductionto Udayana'sNydyaku-
sumdnjali,Publications of the DeNobili Research Library, vol. 1 (Vienna, 1972), p. 28.
9. Gopimohan Bhattacharyya, Studies in Nyaya-Vaisesika Theism, Calcutta Sanskrit College
Research Series, no. 14 (Calcutta: Sanskrit College, 1961), p. 44.
10. .Rgveda,X, 82, trans., for example, by Sarvepalli Radhakrishnan and Charles A. Moore, in
Radhakrishnan and Moore, eds., A Source Book in Indian Philosophy (Princeton, New Jersey:
Princeton University Press, 1957), p. 18.
11. Rgveda, X, 90, in Source Book, p. 19.
12. Rgveda, X, 121, in Source Book, pp. 24-25.
13. .Rgveda, X, 129, in Source Book, pp. 23-24.
14. AtharvavedaVII, 102, 1 and XIX, 6, 4. Compare Margaret and James Stutley, Harper's
Dictionary of Hinduism(San Francisco, California: Harper & Row, 1977), p. 120; and M. D. Sastri,
"History of the Word Tsvaraand Its Idea," All India Oriental ConferenceVII (Baroda), pp. 492 ff.
15. Cf. David Kalupahana, Causality: The Central Philosophy of Buddhism(Honolulu, Hawaii:
University Press of Hawaii, 1975), pp. 17-18. Kalupahana's summary of pre-Buddhist causation
theories is a good one.
16. Compare, for example, BrhaddrdnyakaII, 1, 2-13, and 20; III, 6, in Source Book, pp. 79,
85-86.
17. Compare, SvetdsvataraIII, 7-10; IV, 1, 1, and so forth, in Source Book, pp. 90-91.
18. Svetasvatara I, 2, in Source Book, p. 89.
19. BhagavadgTtdX, 21; XI, 9-35, in Source Book, p. 136, 138-141.
20. Ibid., p. 506; Surendranath Dasgupta, A History of Indian Philosophy (Reprint, Delhi:
Motilal Banarsidass, 1975), vol. 1, p. 418. Dasgupta notes (pp. 420-421) that most of the early
commentators on the Brahmasutraswere quasidualisticVaisnavas. Compare also Hajime Nakamura,
A History of Early VeddntaPhilosophy, trans. Trevor Leggett and others, Religions of Asia Series,
no. 1 (Delhi: Motilal Banarsidass, 1983), section IV.
21. Brahmasutras(hereafter cited as BS) I, i, 2, in Source Book, p. 511.
22. BS I, iv, 3, in Source Book, p. 515. Safikara's comments here, and at II, i, 6 (ibid., p. 522)
indicate that this may be a variant of the cosmological argument, with the existence of the cosmos as a
whole pointing to the existence of a cause on which it is contingent.
23. BS I, iv, 23, in Source Book, p. 521.
24. Commentary to BS I, iv, 23, in Source Book.
25. Cf., e.g., Ninian Smart, Doctrine and Argumentin Indian Philosophy (London: George Allen
and Unwin, 1964), pp. 156-158.
26. Yogasutra(hereafter cited as YS) 23-28, in Source Book, pp. 458-459.
27. YS 25 and Bhdsyya,in Source Book, p. 458; Smart (p. 157) argues that this is a modified form of
the ontological argument.
28. Nydyasutra(hereaftercited as NS) IV, 1, 19-21, trans., eg., by Mrinalkanti Gangopadhyaya,
Nydya Philosophy (Calcutta: Indian Studies Past and Present, 1973), IV, pp. 21-26.
29. Compare, for example, Chattopadhyaya, chap. 16; and Karl H. Potter, ed., Encyclopediaof
344 Jackon

Indian Philosophies: Indian Metaphysics and Epistemology:The Traditionof Nyaya-Vaisesika up to

Gahgesa (Princeton, New Jersey: Princeton University Press, 1977), p. 100.
30. Compare Potter, Encyclopedia,p. 239.
31. Compare Gangopadhyaya, Nydya Philosophy, and Potter, Encyclopedia,p. 263.
32. Paddrthadharmasam.graha 40; compare Potter, Encyclopedia,p. 285.
33. NydyavdrttikaIV, 1, 19-21, summarized by Potter, Encyclopedia,pp. 331-333.
34. Nydyakusumdnjali,I, 3; compare, for example, Chattopadhyaya, Indian Atheism p. 21;
Potter, Encyclopedia,p. 558.
35. DTghanikdya(hereafter cited as D) I, 17; compare, for example, Kalupahana, Causality,
pp. 20-21; and Helmuth von Glasenapp, Buddhism-A Non-Theistic Religion, trans. Irmgard
Schloegl (New York: George Braziller, 1966), pp. 40-41, 144.
36. For example, MajjhimanikdyaII, 222; AhguttaraI, 173;Jdtaka V, 238, and so forth;compare,
for example, Kalupahana, Causality, p. 22; Glasenapp, Buddhism,pp. 39-40.
37. D II, 11, 81-83; compare Glasenapp, Buddhism,p. 146.
38. D II, 13,14-20; compare Glasenapp, Buddhism,pp. 146-148. On Nikdya discussions of God,
compare also Gunapala Dharmasiri, A BuddhistCritiqueof the ChristianConceptof God (Colombo:
Lake House Investments, 1974), passim.
39. Louis de la Vallee Poussin, Vijnaptimdtratdsiddhi: La Siddhi de Hiuan-Tsang (Paris: Paul
Geuthner, 1928), I, p. 30.
40. Buddhacarita IX, 63; compare E. H. Johnston, The Buddhacarita or Acts of the Buddha
(Reprint, Delhi: Motilal Banarsidass, 1972), II, p. 136.
41. BuddhacaritaXVI, 18 ff.; compare, for example, Chattopadhyaya, Indian Atheism, p. 103.
42. Suhrillekha 50; compare, for example, Geshe Lobsang Tharchin and Artemus B. Engle,
Nagarjuna's Letter (Dharamasala: Library of Tibetan Works and Archives, 1979), pp. 84-86.
43. CatuhstavaII, 33-34; compare Chr. Lindtner, Nagarjuniana,Indiske Studier4 (Copenhagen:
Akademisk Forlag, 1982), pp. 150-151.
44. Bodhicittavivarana7-9; compare Lindtner, pp. 186-189.
45. Lindtner, in Nagarjuniana (p. 16), maintains that its attribution to Nagarjuna is "most
probably" false.
46. Compare George Chemparathy, "Two Early Buddhist Refutations of ITvaraas the Creator
of the Universe," Wiener Zeitschriftfiir Kunst und OrientalischeStudien, 22-23, pp. 89-94, 97-99;
and Th. Stcherbatsky, Papers of Th. Stcherbatsky, trans. H. C. Gupta, ed. Debiprasad Chattopad-
hyaya, Soviet Indology Series no. 2 (Calcutta: Indian Studies Past and Present, 1969), pp. 1-16.
47. Mahavibhdsa,TT XXVII, 993b, summarized in Nakamura, History, pp. 147-151.
48. Abhidharmakosa(hereafter cited as AK ) and -bhdsya II, 64d; compare Louis de la Vallee
Poussin, L'Abhidharmakosade Vasubandhu(hereafter cited as AV ) (Paris: Paul Geuthner, 1923-
1931), I, pp. 313-315.
49. AKIV, 1.
50. AKVII, 13a;A VVII, pp. 38-39. These characteristicsare the last three of the four aspects of
the truth of origination, namely, samudaya,prabhavaand pratyaya.
51. Yogadcrabhumi,pp. 144-5; compare Chemparathy, "Two Early Buddhist Refutations,"
pp. 86-89, 94-96.
52. Bhavya, Madhyamaka-hrdaya-vritti-tarkajvala, III, 9, in Daisetz T. Suzuki, ed., The Tibetan
Tripitaka, Peking Edition (hereafter cited as PTT ) (Tokyo, Kyoto: Tibetan Tripitaka Research
Institute, 1957), no. 5256, vol. 96, pp. 49/3/6-50/2/2/.
53. Ibid., pp. 49/5/2-7.
54. For a discussion of this much mooted point, compare Th. Stcherbatsky, Buddhist Logic
(Reprint, New York: Dover, 1962), I, pp. 38-39; and Masatoshi Nagatomi, "The Framework of the
Pramanavarttika, Book I," Journal of the American Oriental Society 79 (1959): 263, note 1.
55. For a discussion of the way in which these epithets structurethe chapter, compare Nagatomi,
"The Framework," and compare also Ernst Steinkellner,"The spiritual Place of the Epistemological
Tradition in Buddhism," Nanto Bukkyo 49 (1982): 1-18.
56. Mi pham, Tshadma rnam 'grelgyigzhung gsalpor bshadpa legs bshadsnang ba 'igter (Block-
print, Dehradun: Nyingma Monastery, n.d.), p. 257.
57. I am following Miyasaka's numbering here. Shastri numbers two introductory verses that
345

Miyasaka does not, so the Shastri number is found by adding 2 to the Miyasaka number. Compare
previous note 7, for references.
58. rGyal tshab rje, Rnam-Agrel-Thar-Lam-Gsal-Byed(hereafter cited as GT) (Sarnath: Tibetan
Monastery, 1974), vol. 1, pp. 238-248. The section on 'svarais, according to rGyal tshab (p. 239),
part of Dharmakirti's attempt to show the meaning of the word "became" (bhuta) authoritative: a
permanent entity like Tsvaraalways has been authoritative, and so cannot "become" so.
59. The PVpanjikd,or -vrtti (hereaftercited as PVV) is extant only in Tibetan: PTT no. 5717(b),
vol. 130; and sDe dge no. 4217, found at, for example, TheNyingma Edition of the sDe-dge bka'-'gyur
andbsTan-'gyur(Oakland, California: Dharma Press, 1981), vol. 94, fols. 732-746, pp. 184- 188. The
section on Tsvarais at the end of chapter 11 and the beginning of chap. 12.
60. nityam pramanam naivdsti pramdnydd vastusangateh / jneydnityatayd tasyd adhrauvyat
kramajanmanah// nitydd utpattivislesddapeksaya ayogatah / (tshad ma rtag pa nyid yod min / dngos
yod rtogspa tshadphyirdang / shes bya mi rtagpa nyidkyis / de ni mi brtannyidphyir ro / rimbzhinskye
ba can dag ni // rtag las skye ba mi 'thadphyir / Itos pa mi rungpa yi phyir /).
61. GT, pp. 239-240; Jackson, Is EnlightenmentPossible? pp. 564-566.
62. Compare, for example, Shah, chap. 2; and D. N. Shastri, passim.
63. Compare the concluding section for remarks on this issue.
64. kathancinnopakdryatvddanitye'py apramdn.ata-// (rnam 'gasphan gdags bya min phyir / mi
rtag na yang tshad med nyid //).
65. Prajiakaragupta, Pramdnavdrttika-bha.syaor Vdrttikalahkdra of Prajntikaragupta, ed.
Rahula Sankrityayana (Patna: Kashi Prasad Jayaswal Research Institute, 1953), p. 34.
66. GT, p. 240; Jackson, Enlightenment,p. 566.
67. PVV, Nyingma sDe-dGe, vol. 94, fols. 733-734, p. 185.
68. sthitva pravrttisamsthdnavisesdrthakriyddisu / sdod 'jugdbyibs kyi khyadpar dang / don byed
pa la sogs pa dag /). Cf Tattvasam.graha(TS) 46 (for full references, compare note 117 following).
69. Preceding, p. 6. Prof. Karl Potter has disagreed with me that the syllogism being refuted is
Nyaya-Vaisesika, noting (a) that no such exact syllogism is found in Nyaya-Vaigesika works and (b)
that no later Nyaya-Vaisesika works specifically defended the tradition against Dharmakirti's
attacks. Prof. Potter has suggested that Dharmakirti's opponent may, in fact, be a lost Sam.khya
work. This may well be, but it must be argued from silence, and it seems to me that (a) while the
syllogism refuted by Dharmakirti is not precisely like those found in Nyaya-Vai§esika works, there is
a significant overlap and (b) later Nyaya-Vai§esika works may not have specifically addressed
Dharmakirti's objections because by the time they were written, Dharmakirti's arguments perhaps
had been overshadowed by those of Santaraksita, Kamala§ila and Jiama§ri. Prof. Potter also has
pointed out-and in this I quite agree with him-that Dharmaklrti's opponent may be unidentifiable
for the simple reason that DharmakTrtihas distorted the theistic position in recasting it for discussion.
Thus, the Nyaya-Vai§esikas may have been the intended target, but not recognized for their own
position as restated by Dharmakirti. Alternatively, Dharmakirti may be combining the ideas of more
than one theistic school into the syllogism.
70. Preceding, p. 10.
71. GT, p. 241; Jackson, Enlightenment, p. 567. This does not depart substantially from the
interpretation of Devendrabuddhi, who differs only in describing that which must be created as
bodies, environs, and products, and cites as an example of "particular shape" not a pot, but a
mansion, PVV, fol. 734, p. 185.
72. istasiddhirasiddhirvi drstdntesam. ayo 'thavi // ('dodpa grubpa 'amdpe ma grub/ yang na the
tshom za ba yin //).
73. GT, p. 241; Jackson, Enlightenment,pp. 567-568.
74. siddham.yddrg adhisthdtrbhdvdbhdvdnuvrttimat / sam.nivesaditadyuktam tasmddyad anumTy-
ate// (byin rlabsyod med rjes jug can / dbyibssogs ci 'drarabgrubpa / de las rjessu dpoggang yin / de ni
rigs pa nyid yin no //).
75. GT, pp. 241-242; Jackson, Enlightenment,p. 569; compare TS 63.
76. vastubhede prasiddhasya sabdasiamnyid abhedinah / na yuktdnumitih pdn.dudravyddivad
dhutWsane// (tha dadngos la rabgrubpa / sgra mtshungstha dadmedpa'iphyir / rjes dpog rigspa ma yin
te / skye bo'i rdzas las me bzhin no //).
77. GT, p. 242; Jackson, Enlightenment,p. 570.
346 Jackson

78. anyatha kumbhakarena mrdvikdrasya kasyacit / ghatadeh karanat sidhyed valmFkasyapi

tatkrtih // (de Ita min na rdza mkhangyis / bumpa la sogs jim pa yi / rnam 'gyur 'ga' zhig byed pa'i
phyir / grog mkharyang des byas grub 'gyur//).
79. GT, p. 242; Jackson, Enlightenment,p. 570; compare TS 65.
80. "When a general result is probative / Because it is concomitant with the predicate, / (Then,)
when one (over-)differentiates the relator, / That differentiation is asserted to be the flaw called
kdryasama."sddhyendnugamdtkdryesdmdnyendpisddhane/ sambandhibhedddbhedoktidosahkdrya-
samo matah // (bsgrubbya'i rjes 'grophyir 'brasbu / spyis kyang sgrubpar byedpa la / 'bralba can nyid
the dad phyir / tha dad skyon brjod 'brasmtshungs 'dod//) (verse 14).
81. GT, pp. 242-243; Jackson, Enlightenment,pp. 570-573.
82. "(Although) one proves (a thesis) in regard to a particularclass, / It is not reasonable to prove
(a similar thesis) just from seeing / That there is a general term (that is similar to the reason); as if /
Words could be horned because (there is a term,) gotva."jatyantare prasiddhasyasabdasdmanyadar-
sandt/ na yuktam sddhanamgotvac cchasdlzndm.visdnivat// (rigs kyi khyadpar la grubpa / sgra yi spyi
ni mthongpa las / sgrub byed mi rigs ngag la sogs / go nyidphyir na rva can bzhin //) (verse 15).
83. GT, p. 244; Jackson, Enlightenment,p. 573.
84. "Because (words) are controlled by a desire to express, / There is nothing for which there is
not a word; / If one attained (objects) through the existence (or words for them), / All people should
attain all objects." vivaksaparatantratvdn na sabddhsanti kutra vd/ tadbhdvadarthasiddhautu sarvam
sarvasyasidhyati // (brjodpar 'dodpa'igzhan dbangphyir/ sgra rnamsgang la'ang medma min/ de yod
pas ni don grub na / thams cad kyis ni thams cad grub //) (verse 16). See GT, p. 244; Jackson,
Enlightenment,pp. 573-574.
85. "Through this (approach) one can also investigate (and refute) such Samkhya (and Jaina
syllogisms as, respectively,) / 'Buddhi is non-sentient, because it is impermanent,' / And '(A tree) is
sentient, because it dies / When its bark is stripped."' etena kdpilddTndmacaitenyddi cintitam /
anityddes ca caitanyam maran. t tvagapohatah// ('dis ni ser skya la sogs kyi / mi rtag sogs phyir yang
sems med / sogs dang shungpa bshusna ni / 'chiphyir sems Idandpyadpa yin //) (verse 17). GT, p. 244;
Jackson, Enlightenment,p. 574.
86. "If a general entitative (reason) is unproven, (then the syllogism is invalid,) / Whereas if the
(general reason) is proven, then even if particular (details) / Are unproven, (the syllogism) is not
invalidated, / As (whether or not) sound "depends on space" (does not affect the permanence or
impermanence of sound). "vastusvarupe'siddhe'yamnydyahsiddhe visesan.am/ abadhakamasiddh~v
apy dkddsarayavaddhvaneh// dngospo'i ngo bo ma grubna / tshul 'digrubna ma grubkyang / khyadpar
gnod byed ma yin te / sgra yi nam kha' la brten bzhin //) (verse 18). GT, pp. 244-245; Jackson,
Enlightenment,pp. 574-575.
87. "Even if a word is unproven, if the entity / Is proven, then (the reason) will be proven, as / The
Buddhists explain to the Aulukyas / "(Atoms are impermanent,because) they are physical." asiddhdv
api sabdasyasiddhe vastunisidhyati / aulukyasyayathd bauddhenoktam. murtyddisddhanam// (sgra ma
grubkyang dngospo ni / grubna grub par 'gyur te dper/ 'ugpa da la sangs rgy as pas / lus sogs sgrubbyed
bshadpa bzhin //) (verse 19). GT, pp. 245-247; Jackson, Enlightenment,pp. 575-577.
88. "If the (entity) is mistaken, / Then even if the word is unmistaken, / The proof must be known
as flawed, / Because an entity is (only) proven from an entity." tasyaiva vyabhic&rddausabde'py
avyabhicdrini/ dosavatsddhanamjneyam.vastunovastusiddhitah// (de nyid 'krulla sogs yin na / sgra ni
'khrulpa med na yang / sgrub byed skyon Idanshes bya ste / dngos las dngospo grubphyir ro // (verse
20). GT, p. 247; Jackson, Enlightenment,p. 577. Note the strong element of "realism" here: though
the connection between words and entities may be tenuous, it still is assumed by Dharmakirti that
there is a definite and discernible nature to entities, which may serve as the foundation for valid
reasoning.
89. "'Because it is a "goer"' and 'because it is "hand-possessing"' / (As reasons) proving (a
colored cow) is a cow and (an elephant calf) is an elephant / Are not (validly) asserted, for these are
verbal expressions / That are merely common (sayings)." "'gro ba 'iphyir dang lag Idanphyir / rva can
glangpo zhes sgrubbyed / 'diyi sgra yi brjodbya ni / gragspa yin gyis brjod 'dodmin // (verse 20a). GT,
p. 247; Jackson, Enlightenment,pp. 577-578.
90. yathd tat karanan? vastu tathaiva tadakdranam / yadd tat kdranam kena matam nestam
347

akdranam// (ji Itardngos de rhyuyinpa / de lta de nyidgang gi tshe / rgyumingang gis de ni rgyur/ 'dod
la rgyu ma yin mi 'dod//).
91. GT, p. 247; Jackson, Enlightenment,pp. 578-579.
92. Compare preceding, p. 16, alternative (c).
93. sdstrausadhdbhisambandhdc caitrasya vranarohane/ asambaddhasyakim sthdnohkdranatvam
na kalpyate // (mtshon dang sman sogs 'brelba las / nag pa'i rma dang 'drubsyin na / 'brelmed sdong
dum ci yi phyir / rgyu nyid du ni rtog mi byed //).
94. Nagatomi, A Study of DharmakTrti'sPramdnavdrttika,p. 33.
95. GT, p. 247; Jackson, Enlightenment,p. 579. rGyal tshab probably was unfamiliar with the
instance of "homeopathic magic" cited by Dharmakirti, and glosses the verse as having the weapon
inflict the wound and the medicine heal it. Devendrabuddhi (fol. 742, lines 6-7) supports the reading
we have given. Incidentally, an instance of homeopathic magic is cited in Dante's Inferno (XXXI,
4-6), where the poet recalls the lance of Achilles and his father, which could both wound and heal.
96. svabhdvabhedanavind vydpdro 'pi na yujyate / nityasydvyatirekitvdtsdmarthyan ca duran-
vayam// (rang bzhinkhyadpar medpar ni / byedpar yang ni mi rungngo / rtag la Idogpa medpa'i phyir /
nus pa nyid kyang rtog par dka' //).
97. GT, p. 247; Jackson, Enlightenment,p. 579.
98. Ibid.
99. yesu satsu bhavaty eva yat tebhyo 'nyasyakalpane / taddhetutvenasarvatrahetunamanavasth-
itih // (gang dag yod na gang 'gyurnyid / de dag las gzhan de yi rgyu / rtog pa yin na thamscad la / rgyu
rnams thugpa med par 'gyur //).
100. GT, pp. 247-248; Jackson, Enlightenment,pp. 579-580.
101. GT, p. 248; Jackson, Enlightenment,p. 580.
102. svabhavaparindmenahetur ahkurajanmani/ bhuimyddistasya sam.skdretadvisesasya darsandt
// (myugu skyed la sa la sogs / rang bzhinyongs su gyur nas ni / rgyu yin de legs byas pa na / de yi khyad
par mthongphyir ro //).
103. GT, p. 248; Jackson, Enlightenment,p. 580.
104. yathd visesena vind visayendriyasam.hatih/ buddherhetus tathedamcen na tatrdpi visesatah //
(gal teji Itaryul dbangpo / tshogspa khyadmed blo rgyuyin / de Itar 'diyin zhe na min / de las khyadpar
yodphyir ro //).
105. NS,I, 1,4.
106. This is the import of Devendrabuddhi's reading at PVV, fol. 745, line 2, p. 188.
107. GT, p. 248; Jackson, Enlightenment,p. 580.
108. prtak prtag asaktdndam svabhdvdtisaye'sati / samhatdvapy asdmarthyam.sydt siddho 'tisayas
tatah // (so so so sor nus med rnams/ rang bzhinkhyadpar medpas na / tshogs kyang nuspa med 'gyur
bas / de phyir khyad par grubpa yin //).
109. GT, p. 248; Jackson, Enlightenment,pp. 580-581. Emphasis mine.
110. tasmdtprtag asaktesu yesu sam.bhdvyategunah / samhatauhetuta tesdm.nesvardderabhedatah
// (dephyir so sor gang nus med / tshogs na yon tan srid 'gyurba / de dag rgyuyin dbangphyugsogs / ma
yin khyadpar med phyir ro //).
111. Compare preceding, p. 9.
112. Verse 183; GT, p. 219; Jackson, Enlightenment, pp. 713-714. The verse is found in the
discussion of the aspect of origination (samudaya) of the truth of origination.
113. GT, p. 251; Jackson, Enlightenment,p. 586.
114. Compare, for example, the writings of Chattopadhyaya, N. Bhattacharyya, and Potter
mentioned in this article (preceding, notes 2, 4, and 29).
115. Bodhicaryavatara IX, 118-123. It ought to be noted that Santideva is misrepresenting the
Nyaya-Vaisesika view, whereby 7~varais not the creator of the paddrthas, but their arranger.
116. Ibid., IX, 124-125.
117. Tsvarais rejected at TS 46-93, purusa at TS 153-170. Compare Ganganatha Jha, trans., The
Tattvasamgraha of Sdntaraksita with the Commentary of Kamalahsla(Baroda: Oriental Institute,
1937), vol. 1, pp. 68-101, 132-139.
118. TS 47-48 and 56-60; Jha, Tattvasamgraha,pp. 69-71 and 75-79.
119. TS 85; Jha, Tattvasamgraha,p. 92.
348 Jackson

120. TS 155-161, Jha, Tattvasamgraha,pp. 133-135.

121. TS 162-167; Jha, Tattvasamgraha,pp. 135-137.
122. Compare Chattopadhyaya, Indian Atheism, chap. 15, for a good summary of MTmarmsaka
arguments.
123. For references, compare preceding, note 4.
124. NydyamanjarT125-133; summarized in Potter, pp. 371-373.
125. VyomavatT40;summarized in Potter, Encyclopedia,pp. 435-436.
126. NydyavdrttikatdtparyatrkdIV, 1, 21; summarized in Potter, Encyclopedia,pp. 481-482.
127. The works of both Jiianaarimitraand Ratnakirti have been edited by Anantalal Thakur in the
Tibetan Sanskrit Works Series (Patna: Kashi Prasad Jayaswal Research Institute, 1959 and 1957,
respectively).
128. Chemparathy,An IndianRational Theology,p. 28. 1 have not as yet studied Jfinansrmitra'sor
Ratnakirti's arguments.
129. Compare preceding, notes 8 and 9.
130. Compare preceding, note 4.
131. Indeed, there is not even agreement on whether the syllogism corresponds to the argument
from design: G. Bhattacharyya (p. 44) calls it the "cosmological argument," while Potter (p. 102)
considers it "cosmoteleological." In fact, the syllogism refuted by Dharmakirti-which seeks to
prove that entities are preceded by a conscious designer because of intermittence, particular shape,
and efficiency-seems most like the argument from design, while the later syllogism proposed by
Jayanta-in which the existence of Tsvarafollows from the world's being an effect-seems a bit more
"cosmological"-although the focus there still is on the analogical appeal to design.
132. The most concerted attack on Western theism by a Buddhist is that of Dharmasiri (preceding,
note 38), who does not, however, often directly relate Buddhist arguments to Western ones, but,
rather,criticizes modern Western arguments directly, interspersinghis discussion with passages from
and reflections upon the Theravadin tradition. Theravada does not develop a rational a-theology to
anywhere near the degree that the Sanskritic "Pramdna"tradition does.
133. Umberto Eco, The Name of the Rose, trans. William Weaver (New York: Harcourt Brace
Jovanovich, 1983), pp. 492-493.