02/28/16 – 01:34:57 PM

Phenomenology, idealism, both or neither?

Making sense of Yogācāra-Vijñānavāda arguments against
external objects
Birgit Kellner ([email protected])
Institute for the Cultural and Intellectual History of Asia / Austrian Academy of Sciences

Workshop
“Buddhist Philosophy of Consciousness: Tradition and Dialogue”
National Chengchi University, Taipei
March 11-12, 2016

1. Introduction

The modern study of Buddhist philosophy has always included characterizations of

various doctrines, theories and frameworks in terms of identifiers derived from west-
ern philosophy, be it as empiricist or rationalist, realist or idealist, phenomenological
or otherwise. Such identifiers were at times applied without much argument, simply
out of the impulse to characterize the unknown in terms of the known (or, to put it
differently, the foreign in terms of the domestic). But especially in the late 20 th century
such characterizations have become the subject of more explicit discussion and, occa-
sionally, controversy. Yogācāra, or more specifically what Buescher 2008 calls
Yogācāra-Vijñānavāda, is a case in point. Practically by default, Yogācāra philosophy
that advocates “mere-cognition” (vijñaptimātratā) was presented and approached as a
form of idealism, until this position began to be denied in several critical studies. Vari-
ous non-idealist or even anti-idealist readings of Yogācāra were articulated that in-
stead promoted affinities of the tradtion with phenomenalism, representationalism
and even phenomenology.1

1 In Kellner and Taber 2014, we have listed the following studies as in one way or another denying
that Yogācāra-Vijñānavāda is (subjective) idealism, although they differ otherwise considerably in
style, rigour, and focus: Wayman 1979, Kochumuttom 1982, Hall 1986, Hayes 1988, Oetke 1992, and
Lusthaus 2002. We also note that Rahula 1978: 79-85 anticipates this position. Garfield 2002 also
refers to contributions on the e-mail-list BUDDHA-L by John Dunne, Dan Lusthaus and John
Powers in 1996 (but the archives of that list are no longer accessible; they used to be at
https://listserv.louisville.edu/archives/buddha-l.html).

1
 02/28/16 – 01:34:57 PM

One may suspect that part of the reason why idealist readings of Yogācāra were
called into question was that idealism in general has gone out of fashion as a philo -
sophical position. Philosophers have moved on to debating other positions, and to act-
ively pursuing positions that are better compatible with natural sciences, positions
that are, in other words, naturalistic. Christian Coseru more specifically articulates a
“phenomenological naturalism” in which a Husserl-inspired phenomenology, under-
stood as a disciplined and methodical investigation of consciousness that takes the
first-person perspective seriously, provides the main conceptual resources for a philo-
sophical account of the mind.2 Coseru’s argument, if I understand it correctly, is that
phenomenological naturalism is a better framework for accommodating Buddhist epi-
stemology than the lenses of representationalism and idealism through which that tra-
dition’s perspectives on consciousness and cognition are typically viewed.

Naturalistic accounts of consciousness and cognition may however turn out to be

misplaced if we are to make sense of anti-realist tendencies in Buddhist philosophical
traditions, and of the specific features of their metaphysics. Claus Oetke once ob-
served that most Buddhist philosophers pursued what Peter Strawson called ‘revision-
ary metaphysics’: a metaphysics is driven by the intent to produce a better structure
for our thinking about the world, while (Strawson’s own programme of) ‘descriptive
metaphysics’ would confine itself to simply describe its actual structure. To be sure,
revisionary elements in Buddhist philosophy might not render attempts at naturaliza-
tion completely futile. For, as Oetke also observed, the revisionary Buddhist systems
did not detach themselves completely from the conceptual frameworks of common
sense, and in fact included elements of everyday thinking in their ontologies that were
not ultimately consistent with their revisionary intentions. 3 (And such frameworks
might be easier naturalised than ones which regard our common intuitions and theor-
ies about the world to be fundamentally flawed.) Philosophical systems are complex

2 Coseru 2012.
3 Oetke 1988: 37. Oetke also warned against what he thought to be a simplistic dichotomy in
Strawson’s model, that any given philosopher could only be either descriptive or revisionary in their
metaphysics. Recently, Alex Watson also picked up Strawsonian revisionary metaphysics in his dis-
cussion of the light analogy for consciousness (Watson 2014).

2
 02/28/16 – 01:34:57 PM

and may contain elements that are in tension with one another, hence it is possible for
naturalistic accounts to have merits while at the same time encountering limitations.

Even if we one for this reason might not wish to discard the naturalization of
Buddhist philosophy completely, naturalistic accounts will tend to downplay revision-
ary elements in Buddhist thinking. On this background, it is perhaps no coincidence
that Coseru on the one hand takes Śāntarakṣita’s and Kamalaśīla’s “Examination of
the External Object” (Bahirarthaparīkṣā) chapter of the Tattvasaṅgraha and -pañjikā as
his point of departure, but on the other hand devotes precious little attention to the
fact that the overarching goal pursued in that chapter is the proof of vijñaptimātratā
by refuting external objects. As for Dharmakīrti's attitude toward the ontological
status of external objects, Coseru wryly remarks that it is “at best ambiguous and at
worst contradictory”.4

It would certainly be problematic if the doctrine of vijñaptimātratā that appears to

flatly deny the existence of external objects were simply to be brushed aside because it
does not fit into a particular framework in the philosophy of mind that can currently
claim some prominence. On the other hand, one would be ill-advised to stubbornly
insist on Yogācāra-Vijñānavāda’s idealist character and denounce all attempts to make
use of phenomenological ideas as nothing more than kowtows to current intellectual
fashion, and let that be the end of discussion. For, looking back in history, it may well
be that idealist readings of Yogācāra-Vijñānavāda similarly had more to do with the
dominance of a particular philosophical paradigm, and with its tacit recognition as a
default framework. Some might well have promoted idealist readings of Yogācāra
simply because it seemed to be the most “natural” way to read what appear to be deni-
als of an external world. Idealist construals of Yogācāra readings might upon closer ex-
amination also turn out to be misguided and flawed. Our understanding of vijñap-
timātratā discourse (and here and in the following I also subsume Buddhist pramāṇa
discourse that is geared towards proving vijñaptimātratā even if other terms are being
used) can only advance if we reexamine Yogācāra-Vijñānavāda literature and try to

4 Coseru 2012: 39 is the only place in the monograph where the problem of external objects is men-
tioned, though not elaborated.

3
 02/28/16 – 01:34:57 PM

make the best possible sense of its positions – as well as of its arguments.

Advocates and defenders of vijñaptimātratā have, after all, produced a rich reper-
toire of arguments to support their claims, and to defend their proofs against objec-
tions voiced both by fellow Buddhists and by contenders in their respective intellec-
tual environments (in India, contenders were especially Mīmāṃsakas and
Naiyāyikas).5 I will limit myself to Indian Yogācāra-Vijñānavāda here, as a matter of
course, as East Asian traditions are beyond my area of expertise. The terrain of debates
surrounding vijñaptimātratā in Indian philosophy between, roughly, Vasubandhu (ca.
between 350 and 420) and Ratnakīrti (ca. 990-1050) – one of the last Indian Buddhist
philosophers to explicitly defend vijñānavāda – is far from being mapped with any
reasonable degree of clarity, although the number of signposts in that terrain keeps
steadily increasing, in recent years focussed on debates surrounding the status of
forms/aspects (ākāra) and the concept of self-awareness (svasaṃvedana).6 Buddhist
thinkers in India felt compelled to defend vijñaptimātratā well into the final period of
Buddhist intellectual flourishing in the monastic centers of Northeast India. Jñā-
naśrīmitra (ca. 980-1040) devoted his monograph Advaitabinduprakaraṇa (not yet
translated into any modern language)7 to the subject. His disciple Ratnakīrti likewise
authored an independent treatise on the topic, the Citrādvaitaprakāśavāda (most of
Ratnakīrti’s works are commentaries on works by Jñānaśrīmitra).8 Ratnākaraśānti, the
third great philosopher from Vikramaśīla monastery, set out to defend vijñaptimātratā
against Śāntarakṣita's critique from a Madhyamaka vantage point 9 – and all three au-
thors touched upon the problem of the external world throughout several of their
works. Only recently two Sanskrit manuscripts of a hitherto unknown “Proof of Mere-
Cognition” (Vijñaptimātratāsiddhi) by Jitāri (ca. 940-1000) have come to light in the
5 Cf. Watson and Kataoka 2010 for an English translation and analysis of Bhaṭṭa Jayanta's (ca. 850–
910) refutation of Yogācāra-Vijñānavāda. This is based on Kataoka's critical edition of the pertinent
section of the Nyāyamañjarī, accessible at http://www2.lit.kyushu-
u.ac.jp/~kkataoka/Kataoka/NMvijR.pdf (last accessed 15 Feb 2016).
6 Cf. the two special issues on self-awareness and ākāra, Journal of Indian Philosophy 38 (2010) (ed.
Kellner) and 42 (2014) (ed. Kellner and McClintock).
7 A study group including Masahiro Inami (Tokyo Gakugei University) and Hisayasu Kobayashi
(Chikushi University, Kyushu) is currently working on this text.
8 Kitahara 1995
9 Moriyama 2014.

4
 02/28/16 – 01:34:57 PM

collection of manuscript copies kept by the China Tibetology Research Centre in Bei-
jing (CTRC). This work has so far not even been known to exist, and it has to current
knowledge not been translated into Tibetan.10 In sum, much still remains to be ex-
plored.

Research into Buddhist epistemology over the past decades has made it blatantly
obvious that the history of Buddhist philosophy after Dharmakīrti is far more than
just a series of footnotes to his work, and it would be odd if proofs of vijñaptimātratā
would not similarly reveal much original argument in the works of, say, Śāntarakṣita,
Dharmottara, Prajñākaragupta or either of the three late Vikramaśīla masters – sev-
eral studies already produced evidence for this. 11 Still, it is also evident that several of
Dharmakīrti's often cryptically expressed inferences to establish core philosophical
principles have exerted lasting influence on later debates. To arrive at a better under-
standing of debates about vijñaptimātratā after Dharmakīrti, we first need to get a
better sense of Dharmakīrti's pertinent proofs, at least in their general outline and
structure.

In examining these proofs, it is also important to attend to Dharmakīrti’s logical

theory, of explicit views on the validity of inference and on the possibilities and limi-
tations of reasoning. Indian logic underwent substantial changes between Vasubandhu
and Ratnakīrti, and not a few changes or innovations were introduced by Dharmakīrti.
The history of logical theory in India was especially after Dharmakīrti in no small part
driven by the discussion of particular philosophical proofs (e.g. the proof of the uni-
versal momentariness of all that exists led to the problematization of the role of the
example, the dṛṣṭānta, as something outside the scope of the subject, the pakṣa). It is in
turn conceivable that philosophical proof strategies on certain topics changed due to
changes in thinking about logical method – or that proofs were not voiced in stronger
terms because it was recognised that one’s own logical method would not allow for
this. What I have in mind here is that logical theory, once entertained with a certain
degree of sophistication, may act as a methodological constraint on just how far one

10 Chu and Franco forthcoming.

11 Iwata 1991, Kobayashi 2001, Kobayashi 2005, Matsuoka 2013, Matsuoka forthcoming, Saccone 2014.

5
 02/28/16 – 01:34:57 PM

can take one’s philosophical position.

In connection with proofs of vijñaptimātratā, it is then important to bear in mind

that these amount to establishing a negative thesis: the view that all objects are “mere-
cognition” is tantamount to claiming that there are no external objects, or that cogni-
tion does not apprehend external objects. For this reason we should in the end con-
sider how proofs of vijñaptimātratā might relate to approaches to negative proofs
more generally, which in the Indian philosophical context leads us to debates sur-
rounding “non-apprehension” (anupalabdhi) and the problem of how to know and
prove the non-existence of things.

Concerning the particular negative thesis that the concept of vijñaptimātratā repre-
sents, a distinction between epistemological and ontological positions has been sug-
gested. Consider, for instance the first words in Schmithausen's “On the Problem of
the External World” in the Cheng weishi lun:

“Yogācāra thought has traditionally been understood as advocating the

epistemological position that mind, or consciousness, does not – at least
not directly – perceive or cognize anything outside itself, but rather cog-
nizes only its own image of an object, and as propounding the ontologi-
cal position that there are no entities, especially no material entities,
apart from consciousness, or, more precisely, apart from the various
kinds of mind (citta) and mental factors or mind-associates (caitta).”12

Regardless their otherwise divergent positions, non-idealist interpreters of vijñapti-

mātratā generally agree that Vijñānavāda does not amount to a full-fledged ontologi-
cal denial. A Husserlian phenomenologist, intent on “bracketing” ontological dis-
course in favour of a methodical investigation of experience alone, would find such
denial rather difficult to assimilate, as Dan Lusthaus' “Buddhist Phenomenology”,
which occasioned Schmithausen’s rebuttal, demonstrates extensively. 13 (And phe-
nomenologists emphasizing the embedding of subjectivity in the world might find this
even harder to swallow.)

12 Schmithausen 2005: 9.
13 Lusthaus 2002.

6
 02/28/16 – 01:34:57 PM

The epistemological and the ontological denial, as Schmithausen and others have
considered them, differ in their logical force. One can maintain without contradiction
that cognition is not of external objects while still maintaining that physical objects
exist. But taking a step backwards, one may question whether the distinction between
these two forms of denial is chiefly significant in terms of logical force – or, for that
matter, whether the difference is all that great. What would be the point in maintain-
ing that there are physical objects, but that these remain inaccessible to experience?
And if one believes to have proven that external objects cannot be perceived (and also
not cognitively accessed in any other way), what evidence would then remain for their
existence? Dharmakīrti, in fact, seems to make pretty much this point: “If cognition
has the form of the object, what evidence is there for the external object?” 14 If the an-
swer is none, then there is no reason to think that it exists. 15 And if there is no reason
to think that something exists, is that not enough reason to think it to be non-exist-
ent? These questions are especially pertinent for approaching Dharmakīrti's argu-
ments against external objects. For, prima facie at least, his arguments are all epistem-
ological in kind, establishing most explicitly what appears to be just an epistemologic-
al denial. The question, then, is what to make of this.

2. Dharmakīrti's arguments against external objects

For a start, Dharmakīrti’s philosophical framework exhibits complexities that com-

plicate our endeavour to properly gauge the significance of his arguments against ex-
ternal objects in his larger philosophical theory. It would be misleading to consider
Dharmakīrti exclusively an advocate of vijñaptimātratā, a vijñānavādin. Several con-
flicting theories of the perceptual process, and the nature and status of perceived ob-
jects, are present in his main epistemological works, the Pramāṇavārttika and the
Pramāṇaviniścaya. Some have attempted to make this intelligible through the heuristic
metaphor of a “sliding” or “ascending” scale of analysis that arranges seemingly in-

14 PV 3.432ab: dhiyo nīlādirūpatve bāhyo ’rthaḥ kiṃpramāṇakaḥ / More literally, the question translates
as “What means of valid cognition does the external object have?”
15 In Kellner and Taber 2014: 717, we make this argument with reference to Vasubandhu's Viṃśikā.

7
 02/28/16 – 01:34:57 PM

consistent theories hierarchically.16 Not all versions of the “scales approach” are
equally plausible, but the ones that are (and here I side with Dreyfus and McClintock
more than with Dunne) allow to explain away seeming contradictions because con-
flicting positions are arranged hierarchically. 17 A low-level theory is first advanced
(and defended), then problems in that low-level theory are pointed out which cannot
be resolved without abandoning its core principles, and a more accurate and truthful
higher level theory is introduced to replace it.

Several levels on the scale have been distinguished. The distinction that matters
most to us here is that between a realist or externalist (Sautrāntika) model of percep-
tion and an internalist Yogācāra model (whose characterization as idealist is open to
debate).18 The latter is more accurate in Dharmakīrti's view; the two positions are hier-
archically arranged – on this point, Dharmakīrti is explicit enough: to account for a
philosophical principle as basic as causality solely on the basis of cognitions, of mental
entities, so to speak, is the doctrine of the wise (viduṣāṃ vādaḥ).19 Dharmakīrti accord-
ingly designates the low-level, externalist theory explicitly as being maintained due to
a fundamental “disturbance” or “confusion” (viplava) – a concept that no doubt points
to ignorance (avidyā), a Buddhist root-evil. As a result of this deep-seated mental dis-
turbance, an “object-form” or mental appearance that is in fact contained within cog-
nition is superimposed (āropa), projected outward.20 Elsewhere Dharmakīrti clarifies
that scriptural teachings of the five psycho-physical aggregates of living beings

16 Dreyfus 1997, McClintock 2003, Dunne 2004. (Dunne addresses specifically Dharmakīrti, Dreyfus
later Tibetan interpreters, and McClintock Śāntarakṣita and Kamalaśīla.) See Kellner 2011 for a cri-
tique of Dunne's ontologial version of Dharmakīrti's sliding scale.
17 Responding to critical remarks, Coseru more recently clarified that he considers Buddhist epistemol-
ogists to be ‘process externalists,’ subscribing to “the view that perceptual and inferential cognitions
depend on, or are continuous with, bodily processes that extend into the environment by virtue of
the tight relations between perception and action.” (Coseru 2015) This, in his view, allows to avoid
the question how someone like Dharmakīrti can seemingly argue for both external realism and
‘epistemic idealism’.
18 I am using “externalism” and “internalism” as labels to speak of theories about intentional objects,
i.e. whether or not these are external or internal to the mind, well aware that there are many con -
texts in contemporary analytical philosophy where these labels are applied in a different sense.
19 PV 3.397: asty eṣa viduṣāṃ vādo bāhyaṃ tv āśritya varṇyate / dvairūpyaṃ sahasaṃvittiniyamāt tac
ca sidhyati //
20 PV 3.431: svabhāvabhūtatadrūpasaṃvidāropaviplavāt / nīlāder anubhūtākhyā nānubhūteḥ parāt-
manaḥ //

8
 02/28/16 – 01:34:57 PM

(skandha) or of the twelve sense-spheres (āyatana) imply the existence of external

objects, but such teachings are spread by buddhas merely in accordance with worldly
understanding. In doing so they set aside non-duality (advaya) as that which is
ultimately true (tattvārtha). In this a buddha resembles an elephant who shuts one of
its laterally set eyes and then only sees to one side. 21 As a concession to common
assumptions in the world (loke, PV 3.430), one may go along with externalism up to a
certain point, but when the nature of cognition is correctly examined, externalism
turns out to be false. To presume external objects of perception is thus false, but
knowledge of its falsehood may be suspended in contexts where that presumption
serves pragmatic, dialectical or even didactic purposes. Coseru considers the “method-
ological approach” of both Sautrāntika and Yogācāra to be “much like Husserl’s phe-
nomenological method,” and to “consist in bracketing common assumptions about the
ontological status of external objects in favor of an analysis of the contents of experi-
ence itself.”22 Dharmakīrti as a Vijñānavādin, however, rather holds that once experi-
ence is properly analyzed, the assumption that the objects of experience exist extern-
ally, independently of their cognition, simply turns out to be false. He might not be a
full-fledged Vijñānavādin, but the way in which he conceives of the relationship
between Sautrāntika and Vijñānavāda reveals limitations to analogies with phe-
nomenology on the level of philosophical outlook and method: suspension of truth is
not the same as phenomenological bracketing. Indeed, the move from the externalist
theory to the more accurate internalist one constitutes an exercise in revisionary
metaphysics if there ever was one, and this seems hard to explain away naturalistic-
ally.

The Sautrāntika, externalist model can be regarded as Dharmakīrti's default model

in his overall exposition of epistemology and logic. This is the account of perception
that is presupposed when no explicit statement is made that cancels a commitment to

21 PV 3.219: tad upekṣitatattvārthaiḥ kṛtvā gajanimīlanam / kevalaṃ lokabuddhyaiva bāhyacintā

pratanyate // The explanation of the example is baed on Manorathanandin's commentary at M1M1
184,10-14.
22 Coseru 2012: 291. For a more elaborate discussion of “bracketing” along these lines, though not with
focus on Dharmakīrti, Garfield 2015, section 3.1 (Kindle version position 4565).

9
 02/28/16 – 01:34:57 PM

external objects. In this model a perceptual awareness is generated by a multiplicity of

causes, among which features an external object like blue colour that exists indepen-
dently of the mind. There is no underlying enduring substrate or subject of cognitions,
and all factors involved exist only momentarily.

Just how that external object exists and generates perception – as individual indi-
visible material particles or as some conglomeration of these – is a matter of discus-
sion, and even of internal tension in Dharmakīrti's own account (advocates of the slid-
ing/ascending-scales approach accordingly distinguish several levels on an externalist
section of the scale).23 Regardless, perception is caused by that external object and has
that object's form or appearance (ākāra, rūpa, ābhāsa) – in other words, perception re-
sembles the object. Causation and resemblance (sārūpya, sādṛśya) account for the in-
tentionality of perceptual states – and this intentionality is, in fact, the problem that
prompts many of Dharmakīrti's arguments against external objects: how to account
for perception's link with its individual object, for what I have elsewhere termed its
object-specificity. Dharmakīrti first defends his causation-and-resemblance-account –
especially resemblance or “form-possession” (sākārajñānavāda) – against brahminical
opponents who deny that cognition possesses form (nirākārajñānavādin).24 But while
Dharmakīrti evidently regards the causation-cum-resemblance model as the most ra-
tional externalist theory of perception, he then subjects it to criticism in various
places, and replaces it by the more accurate view that perception is only of an internal
object-form, and that it is not causally dependent on external objects. What, then, are
the main arguments that he offers in this connection?

Before I present and discuss these, let me briefly sketch the contours of my ap-
proach. Dharmakīrti offers these arguments in different sections of the chapters on
perception (pratyakṣa) in his Pramāṇavārttika and Pramāṇaviniścaya, which we are fi-
nally in a good position to compare because a Sanskrit text for the Pramāṇaviniścaya
is now available. The sections in question are devoted, respectively, to the exposition

23 Dunne 2004.
24 PV 3.301-319. Parallel PVin 1 30,9-33,10 (including vv. 34-37). See Kellner 2011 for a summary of the
arguments.

10
 02/28/16 – 01:34:57 PM

of the means of valid cognition and its result (pramāṇa/pramāṇaphala), and of self-
awareness (svasaṃvedana). I am going to discuss a variety of arguments that Dhar-
makīrti presents here not in their function within these debates, but rather in their
philosophical function, in terms of what approach they reveal to the status of external
objects. The way in which these arguments, found in different places, are related to
one another, is not always clear from the textual surface. One may arrive at different
interpretations if one reads them individually, as stand-alone proofs, or in combina-
tion with other arguments.

What I am trying to do here is offer a way how certain arguments that I regard as
crucial could be made to hang together – and where I disagree with earlier interpreta-
tions advanced by others, this is for the most part because I make connections be-
tween passages where others have limited themselves to a narrower range. The long
section on self-awareness in the Pramāṇavārttika (PV 3.425-539) still contains much
material that I have not been able to give full consideration, although a first reading of
that section that I undertook together with Shinya Moriyama, now almost ten years
ago, convinced me that it holds great interest (and I believe Moriyama's contribution
to the workshop will confirm this). What I am offering here is therefore preliminary,
following one of the main methodological principles that I learned from my teacher
Ernst Steinkellner: “One has to start somewhere.”

2.1. A first type of arguments: the attack on the causation-cum-resem-

blance-theory of perception

Dharmakīrti’s arguments against externalism can be divided into two main groups.
The first group of arguments attack the causation-cum-resemblance theory of percep-
tion, while the arguments in the second group take features of awareness itself as the
basis for the conclusion that perception cannot be of external objects. The arguments
in first group therefore do not refute the view that the objects of perception are exter-
nal in general terms, but merely demonstrate that external objects cannot be perceived
if their perception is to be accounted for by causation and resemblance. As Dhar-

11
 02/28/16 – 01:34:57 PM

makīrti argues while refuting nirākārajñānavāda, this however is simply the most ra-
tional externalist theory. If one has to be a realist at all, then one has to accept that
perception arises from its object bearing that object's form. Otherwise there is no way
to account for how perception is connected to its object (causation alone is not suffi-
cient). One can even go so far as to see this defense of causation-cum-resemblance as
the first step in an extended argument against external objects, for after having argued
that cognition has to have its object's form, Dharmakīrti then proceeds to question
just why cognition then should be of an external object at all – and concludes that it
cannot be, and need not be (PV 3.320-337).25

I have presented two main arguments in the first group elsewhere in greater de -
tail,26 so a summary will be enough. The first is in fact a cluster of arguments that I
have dubbed “arguments from incongruence”27 because they essentially claim a funda-
mental incongruence between what might serve as external object – in physical terms
– and what appears as perceptual content: the many individual atoms that are real and
allegedly cause a perception do not appear as such, while the singular “coarse”
(sthūla), spatially extended28 form that appears has no correspondence among real ex-
istents in the external world. Although Dharmakīrti in one place attempts to remedy
the problem by arguing that the many atoms cause a perception when mutually sup-
porting each other (which creates a greater coherence among them as a special fea-
ture, a viśeṣa),29 but apparently does not consider this to be an ultimate solution of the
problem. The option that there might be one singular coarse object of perception, as
e.g. the Vaiśeṣika's whole (avayavin), is also refuted.30

The arguments from incongruence – or argument in the singular, for one can easily
take them together to form one extended argument – no doubt take their inspiration
from the “Examination of the Object-Support” (Ālambanaparīkṣā) by Dignāga, Dhar-
25 Kellner forthcoming b.
26 Kellner 2011.
27 Kellner 2011, Kellner forthcoming a, Kellner forthcoming b.
28 Visual perception is the default type under discussion for the most part. Perhaps one could extend
the distinction between “subtle” (sūkṣma) and “coarse” (sthūla) also to other types of perception, al-
though I am not aware of any explicit discussions to that effect.
29 PV 3.195-196.
30 See Kellner 2011 for details.

12
 02/28/16 – 01:34:57 PM

makīrti’s (indirect) predecessor. A second argument that Dharmakīrti brings up in the

same context questions that the object-form in perception must derive from an exter-
nal object, by drawing up a particular scenario that yields an absurd consequence. As-
sume that there are two successive perceptions showing the same object. Both have
the same object-form, and the preceding perception is the cause for the subsequent
one. The conditions of causation and resemblance are therefore fulfilled—but then, the
preceding cognition (in technical terms: the “immediately preceding homologous con-
dition”, the samanantarapratyaya) could just as well be the object of the following one
as an external object. of a cognition. 31 The externalist's attempt to define the external
object of perception through causation and resemblance thus turns out to be inconclu-
sive.

These two arguments are straightforwardly concerned with the refutation of the
causation-cum-resemblance theory as an account of how perception applies to exter-
nal objects. Elsewhere in the Pramāṇavārttika, in a section that the commentator De-
vendrabuddhi introduces as presenting Dharmakīrti’s own arguments for self-aware-
ness (independent from Dignāga’s),32 Dharmakīrti points out more fundamental flaws
of the concept of resemblance. If cognition resembled its object in all respects, it
would no longer be cognition; it would be that object. If it were only partially similar
to the object, everything would be aware of everything else. 33 A cognition of a pot
would be a cognition of potsherds because pot and potsherds share the property of
“being cognizable” and thus resemble each other in one respect. If one were to counter
that the cognition of a pot has the form “pot,” and not the form “potsherds,” then the

31 Arnold 2008 reads this not as an absurd consequence for the Sautrāntika, but as a positive argument
by the Yogācāra, for whom the ālambana of an awareness-event is indeed the samanantarapratyaya.
But there is doubt whether this view would be shared by all of Dharmakīrti's interpreters. For a
start, Prajñākaragupta identifies the “latent impression” (vāsanā) responsible for producing a subse-
quent cognition on a Vijñānavāda view with the samanantarapratyaya, while his commentator
Yamāri tells us that Dharmottara regards the impression to be different from the samanantara-
pratyaya (Kobayashi 2001: 202, cited after Matsuoka forthcoming, n. 25). Sucaritamiśra, a
Mīmāṃsaka commentator on the Ślokavārttika, distinguishes “old” Buddhists who regard the vāsanā
to be distinct from the cognition from others who identify the two (Kāśikā 163,27 on ŚV ŚNV 175cd-
176ab; reference from Matsuoka forthcoming, n. 25)
32 De-t D242b5-243a1= P287a5-287b2 (cf. also Tosaki 1985: 105, n. 3).
33 PV 3.434: sarvātmanā hi sārūpye jñānam ajñānatāṃ vrajet / sāmye kenacid aṃśena syāt sarvaṃ sar-
vavedanam // On the transmission of this verse (with substantive variants) see Kellner 2009-2010.

13
 02/28/16 – 01:34:57 PM

problem remains that the pot-cognition be a cognition of all pots, and not indexically
linked to the one pot in front of one's eyes.34

If perception's being aware of an object were to be based just on resemblance, then

it would follow absurdly that all things that resemble each other are aware of each
other. If one then were to limit this kind of resemblance to cognition – only cognitions
are aware of other things by resembling them --, then surely being aware of an object
is not just defined by resemblance.35 In other words, there must be some other feature,
peculiar to cognition, that makes cognition aware of an object when it resembles it,
while, for instance, a mirror does not become aware of an object by taking on that ob-
ject’s form (this is my own example). Dharmakīrti concludes that cognition is “experi-
ence” by itself (svayaṃ so 'nubhavaḥ); its being an experience of something is not
caused by resemblance. Directly after this refutation of resemblance, there follows an
act of concession to the (false) lower-level theory: in the world (loke), resemblance
might well serve as condition for distinguishing the act of cognizing and the object of
cognition,36 i.e. to account for perception’s object-specificity in ways that are more at-
tuned to common-sensical (externalist) assumptions about reality and cognition, as-
sumptions that are also shared by Dharmakīrti's brahminical opponents in debate.

Two points about these arguments against resemblance. First, they drive a wedge
between cognition's “form-possession” (sākāratā) and resemblance (sārūpya, sādṛśya).
For the Sautrāntika externalist, cognition's resemblance is its possession of the object's
form. Here, Dharmakīrti rejects resemblance, but retains that cognition has an object-
form. Second, these arguments are coupled with the claim that cognition’s quality of
“awareness” (saṃvedana) or “experience” (anubhava) comes from cognition itself –
this is just how cognition is. Now, for Dharmakīrti things are what they are – they
have their “own-being” or identity (svabhāva) – because of their causes and condi-
tions. We can thus surmise that cognition is awareness or experience simply because
34 This expansion of the argument with the help of examples is based on Devendrabuddhi’s commen -
tary (De-t D 244b5f.=P289b4-7).
35 PV 3.429: prāptaṃ saṃvedanaṃ sarvasadṛśānāṃ parasparam / buddhiḥ sarūpā tadvic cet nedānīṃ vit
sarūpikā //
36 PV 3.430: svayaṃ so ’nubhavas tasyā na sa sārūpyakāraṇaḥ / kriyākarmavyavasthāyās tal loke syān
nibandhanam //

14
 02/28/16 – 01:34:57 PM

it is produced as such. But there is more: at the time when cognition arises, there is an
awareness (saṃvedana) of the object-appearance that belongs to the cognition, just
like there is also an awareness of cognition itself (svavedana) at that time.37 What
Dharmakīrti seems to be driving at is that cognition is intentional awareness simply
because it is produced as such: it is the very nature of cognition to be aware of the ob -
ject-form that it contains. This step from (a) cognition is awareness to (b) cognition is
intentional awareness and further on to (c) cognition is not aware of external objects
(and hence the object is not different from its cognition) will be made more explicit in
the second group of arguments, to which we can consider the refutation of resem-
blance on the background of self-awareness here a kind of stepping stone.

With these arguments, in any case, Dharmakīrti apparently believes that he has re-
futed externalism:

PV 3.432-433:

dhiyo nīlādirūpatve bāhyo ’rthaḥ kiṃpramāṇakaḥ /

dhiyo ’nīlādirūpatve38 sa tasyānubhavaḥ katham //

yadā saṃvedanātmatvaṃ na sārūpyanibandhanam /

siddhaṃ tat svata evāsya kim arthenopanīyate //

If cognition has the form of the object, what evidence is there for the ex-
ternal object?39 If cognition is without the form of the object, how could
it be an experience of that object?

If cognition’s having the nature of awareness is not conditioned by re-

semblance, then that nature of awareness is established of the cognition
just from cognition itself (svata eva). What is then contributed by an ex-
ternal object?

37 PV 3.425, especially the second half-verse: dvairūpyasādhanenāpi prāyaḥ siddhaṃ svavedanam /

svarūpabhūtābhāsyasa tadā saṃvedanekṣaṇāt // By proving that cognition has two forms, self-
awareness is indirectly established because at the time when an object-cognition arises, one ob -
serves the awareness of an (object-)appearance that belongs to cognition’s nature.
38 Pr-A’ reads nīlādirūpatvaṃ for ’nīlādirūpatve in 432c.
39 Again, more literally: “what means of valid cognition does the external object have”?

15
 02/28/16 – 01:34:57 PM

There is no evidence for the external object, and the external object would more-
over be superfluous, as it would add nothing to cognition which is not already part of
its own nature.

2.2. The second group of arguments: arguments from the nature of

awareness

The two arguments against cognition's having an external object that belong to this
second group have been surveyed and analyzed in Takashi Iwata's masterful study of
the sahopalambhaniyama-inference – which is one of these arguments: the object is
not different from cognition because both are invariably apprehended together. The
second argument in this group can be labelled the “awareness-inference” (translating
Iwata’s “saṃvedana-inference”):40 cognition is not of another object because it is by
nature an “appearing-in-a-certain-way” (tathāprathana), just like cognition's aware-
ness of itself (ātmasaṃvedana). Iwata pointed out two passages in the Pramāṇavārt-
tika where (some version of) the sahopalambhaniyama-inference is also expressed,41
and hence we can conclude that this was an argument that Dharmakīrti had worked
on for some time. There are also passages in the PV that prefigure the saṃvedana-in-
ference, or at least articulate some of the notions that are peculiar to it, including the
refutation of resemblance at PV 3.425-434 that I just summarized. This suggests that
the saṃvedana-inference, too, is comprised of ideas that Dharmakīrti had worked out
for quite some time, across his two main epistemological treatises PV and PVin. Both
these inferences were historically influential, but the sahopalambhaniyama-inference
attracted much more attention and has been more frequently discussed, and drew
much fiercer critique from Naiyāyikas and Mīmāṃsakas.42

In Dharmakīrti’s own tradition, both inferences were construed as having two pur-
poses: first, to prove “self-awareness” in the sense of proving that cognition is only
aware of itself, not of an external object, and second, that cognition has the form of

40 Iwata 1991.
41 PV 3.333-335 and PV 3.387-390ab (the latter being closer to what is the most widely cited passage on
this inference, PVin 1.54cd with prose).
42 Iwata 1991: 13.

16
 02/28/16 – 01:34:57 PM

the object (sākāratā).43 Both claims can be made to be consistent with the presumption
of external reality if they are taken in the more limited sense that cognition is only di-
rectly aware of itself (and indirectly of an external object), and that cognition has a
form of an object that represents some externally existing thing. Indeed, for the sa-
hopalambhaniyama-inference Dharmakīrti himself remarks that the appearing object
is not different from its cognition even if there is an external object, 44 which would re-
quire us to take the “object” proven to be non-different from cognition in the more
limited sense of the “object-form.” In this more limited reading both inferences would
therefore also be acceptable for a Sautrāntika, and they could well be used by a
Sautrāntika to refute a Mīmāṃsaka nirākārajñānavādin (although one should note that
when Dharmakīrti actually refutes nirākārajñānavāda theories of perception, he avails
himself of other arguments).45 The brahminical authors whose works Iwata considered
in his study unanimously understood the two inferences as establishing vijñānavāda
and refuting external objects, and this may well be their dominant interpretation out-
side Buddhist circles on the whole. How one reads these inferences affects the actual
interpretation of the relevant passages right down to minute linguistic details (as,
again, becomes clear from Iwata's meticulous studies of some of the commentarial in-
terpretations). I am not going to enter such details here, but will rather attempt to dis-
cuss these two arguments in their function of refuting external objects. This might not
be their only historically legitimate interpretation, but it is an important one – and I
would go even further and suggest that for Dharmakīrti it is the main interpretation
(although this needs more argument than I can provide in this paper).

2.2.1. The awareness-inference

The awareness- or saṃvedana-inference reads like an extension of the arguments

against resemblance at PV 3.425-434 that I discussed above. As a matter of fact, these

43 Or, alternatively, dvirūpatā: the fact that cognition has two forms, its own as well as that of the ob-
ject. But since it is evident that cognition has its own form (that, after all, defines it as cognition), the
focus is here too on the form of the object. See e.g. TSP 695,14 and 705,6, quoted in Matsuoka
forthcoming n. 56.
44 PVin 1 58ab: bāhye 'py arthe tato 'bhedo bhāsamānārthatadvidoḥ /
45 By this I mean, again, the arguments in PV 3.301-319 with their parallels in PVin 1, cf. above n.24.

17
 02/28/16 – 01:34:57 PM

arguments might be considered as stepping stones that are leading up to it, or as an-
ticipating some of the notions that feature in it. As Iwata pointed out, the inference is
most clearly stated in Pramāṇaviniścaya 1 42,3-6:

saṃvedanam ity api tasya tādātmyāt tathāprathanam. na tad anyasya

kasyacid ātmasaṃvedanavat. tato 'pi na tad arthāntare yuktam.

Paraphrase: What is called ‘awareness’ (saṃvedana), now, is appearing-

in-a-certain-way (tathāprathana) because awareness has that nature
(tādātmyāt). This awareness is not of anything else, just like cognition’s
awareness of itself. For this reason, too,46 it is not possible that aware-
ness is of another thing (i.e., of an external object).47
My paraphase is closest to Jñānaśrībhadra’s commentary, and, I believe, the most
straightforward reading of the Sanskrit text. 48 We can identify one main premise:
when we speak of cognition’s being aware of an object – saṃvedanam in the sense of
arthasaṃvedanam – , what is actually the case is that cognition appears or registers in
a certain way (tathāprathana), just as it is the case when cognition is aware of itself.
That cognition is reflexively aware is apparently already taken as a given in this infer-
ence, for this notion is relied upon as the inference’s example (dṛṣṭānta). In this read-
ing here, moreover, the fact that cognition is not of an external object would represent
the sādhya, and the fact that it merely appears in a certain way would constitute the
hetu. The inference’s main premise is justified on the ground of the axiomatic state-
ment that this is simply the nature of awareness (tādātmya), which in technical terms
then further supports the hetu. Parallels that similarly state that experience is the na-
ture of cognition and not of anything else drive home the point.49

To bring out the force of the argument, it may be helpful to focus on the formal fea-

46 “For this reason, too”: as well as based on the sahopalambhaniyama-inference that was just stated in
the text.
47 Cf. Iwata 1991: 9 (based on the Tibetan translation; the Sanskrit was not available at the time).
48 Cf., again, Iwata 1991: 9ff.
49 PV 3.326: ātmā sa tasyānubhavaḥ sa ca nānyasya kasyacit // pratyakṣaprativedyatvam api tasya
tadātmatā // Cf. also the parallel in PVin 1 35,11-12: tasmād ātmaiva buddher anubhavaḥ / sa ca
nānyasya kasyacit / pratyakṣaprativedyatvam apy asyāṃ tadātmataiva / Here Dharmakīrti adds that
the fact that objects are perceived individually (i.e. the object-specificity of cognition, pratyakṣapra-
tivedyatva) also is just cognition’s own nature – the implication being that it is not contributed by
an external object.

18
 02/28/16 – 01:34:57 PM

ture of the verb prathate, “appears”,50 that it is intransitive. It makes no sense to ask
the “what” question for an object: “What did you appear?” is a nonsensical question,
as opposed to “what did you read?” or “what did you dream about?” 51 Now, if
“appearing-in-a-certain-way” is here presented as the very nature of awareness, we
may consider Dharmakīrti here to advance an intransitive notion of consciousness, as
is also true when the verb prakāśate and its nominal derivates are used
interchangeably with prathate:52 to shine forth, to become manifest. The awareness-
inference thus seems to derive the conclusion that cognition is not transitively aware
of an object that exists independently of it from the premise that cognition is
intransitively aware of its object-form (it just appears in a certain way). Cognition’s
mode of awareness does not differ fundamentally for its reflexive awareness
(cognition registers as cognition) and its intentional awareness (cognition apprehends
an object). Here, indeed, analogies may be drawn to Sartre’s one-level account of
consciousness according to which every positional consciousness of an object is
simultaneously a nonpositional consciousness of itself. 53 Yet, Dharmakīrti and his
interpreters obviously go further than phenomenologists might be inclined to do: in
the saṃvedana-inference intransitive awareness is said to be all there is to both self-
awareness and (seemingly) intentional awareness. As Śāntarakṣita understands the ar-
gument: whatever appears in cognition has to partake in cognition’s luminous na-
ture.54 Insentient (jaḍa) objects like rocks do not have that nature of luminosity – of
“shining forth” – and therefore cannot manifest themselves. Whatever appears to cog-
nition has to have the nature of cognition. As Alex Watson rightly noted, the analogy
of light does not serve the argument particularly well, if light is considered to manifest
objects that exist separate from it.55 But cognition is according to the saṃvedana-infer-
ence precisely not a manifestor of others. Cognition just shines forth (i.e., illuminates
50 The verb form prathate is used in PV 3.349, in a closely related context: yathā niviśate so ’rthaḥ
yataḥ sā prathate tathā / arthasthites tadātmatvāt svavid apy arthavin matā // Self-awareness is
thought to be object-awareness because the determination of the object has that (self-awareness) for
its nature, given that self-awareness appears in just the way in which the object has entered it.
51 Cf. Legrand 2009 for a clarification of the transitivity/intransitivity of consciousness.
52 PV 3.327=PVin 1.38, PV 3.446, 478, 480, 481. (List cited in Watson 2014.)
53 See Zahavi 2005: 20.
54 TS 2000: vijñānaṃ jaḍarūpebhyo vyāvṛttam upajāyate / iyam evātmasaṃvittir asya yājaḍarūpatā //
55 Watson 2014: 418.

19
 02/28/16 – 01:34:57 PM

itself), together with the object-appearance that it contains within itself.

One may note a tension here. On the one hand, light is used as an analogy for
something that reveals itself while revealing other objects (thus helping to prove the
innate reflexive awareness of cognition), on the other hand Dharmakīrti uses light to
exemplify something that simply “shines forth” and does not (indeed, cannot) reveal
other objects.56 Prajñākaragupta seems to recognize the problem that the light analogy
has implications that run against Vijñānavāda, and effectively cancels its intuitive ap-
peal to the notion that objects are illuminated by a lamp that exists separately from
them: what actually happens is that a seemingly independent object such as a pot
arises from the lamp as shining forth.57 The pot arises from the lamp as shining forth –
this seems like a classic case of a Buddhist philosopher who has to cope with the prob-
lem that his revisionary metaphysics are in tension with the reliance on common-sen-
sical examples. Prajñākaragupta, for his part, faces the challenge by simply providing
a ‘revisionary’, counter-intuitive account of the example.

2.2.2. The sahopalambhaniyama-inference

In its most widely cited form, the sahopalambhaniyama-inference is stated in the

Pramāṇaviniścaya:

PVin 1 54cd with prose: sahopalambhaniyamād abhedo

nīlataddhiyoḥ | na hi bhinnāvabhāsitve 'py arthāntaram eva rūpaṃ
nīlasyānubhāvāt tayoḥ sahopalambhaniyamād dvicandrādivat. na hy
anayor ekākārānupalambhe 'nyopalambho 'sti. na caitat svabhāvaviveke
yuktaṃ pratibandhakāraṇābhāvāt.

56 It seems to me that in Dharmakīrti’s own works there are precious few places where the light anal -
ogy is used to show that cognition it self- and other-illuminating, but this needs more comprehen-
sive study. One such passage is is PVin 1 42,7f.: sa (sc. anubhavaḥ) ca tādātmyāt tathāprakāśamāno
'pi svaparātmanoḥ prakāśakaḥ syāt, prakāśavat.
57 PVABh 353,20-22 as discussed in Kobayashi 2006: 4 (cited in Watson 2014): yadi ghaṭaḥ pradīpena
bāhyātmanā prakāśyate / pradīpo ’pi tathābhūtenāpareṇeti na paryanuyogaḥ / na ca ghaṭo ’pi
pradīpena prakāśyate / api tu tathābhūtasyaiva tata utpattiḥ // “One (should not) put forward the
criticism that if a pot were illuminated by a lamp that is external in nature, then the lamp, too,
(would be illuminated) by another (lamp) being like that (i.e. external). For, not even the pot is illu-
minated by the lamp (let alone a lamp by another). Rather, there is an arising of the pot being like
that (i.e., shining forth) from that (lamp).” The interpretation of tathābhūtasya as meaning
prakāśarūpasya is confirmed by Yamāri’s commentary, and similar notions are found in Deven-
drabuddhi’s and Manorathanandin’s commentaries ad PV 3.329 (Kobayashi 2006: 6).

20
 02/28/16 – 01:34:57 PM

“Blue and its cognition are not different because they are
invariably58 apprehended together. For, the nature (rūpaṃ) of blue is
not at all a different thing from the experience of blue, even though [the
two] appear as different, because the two are invariably apprehended
together, like the double-moon [seen by someone with cataracts]. That is
to say: there is no apprehension of one of both when the form of the
other is not apprehended. And this is not possible if they are different in
nature (svabhāvaviveke) because there is no cause [of their co-
apprehension] consisting in [some other] connection.”
When the inference is taken to refute external objects, then the status of blue (the
object) must be left undecided at the outset, for it is precisely its difference or non-
difference from cognition that is at issue: whatever counts as the object of perception
is not different from that perception; the object is fully dependent on, or enclosed
within, perception.

The premise from which this conclusion is derived has two components: a percep-
tion is invariably apprehended when an object is apprehended, and the object is in-
variably apprehended when its perception is apprehended. The second component is
less controversial: when I apprehend my perception of blue by thinking “I am aware
that I now perceive blue”, I also apprehend blue. If perception is intentional, being
aware of perception also means being aware of the perceived object, if we surmise
that any cognition of an awareness will be a cognition of an awareness of something.

The first component of the premise, that perception is necessarily cognized when
its object is perceived, has been far more controversial in Indian philosophy. Obvi-
ously, it makes an appeal to cognition's innate self-awareness, and, indeed, the elabo-
ration of the sahopalambhaniyama-inference is one of the main, if not the main, occa-
sions where Dharmakīrti spells out the premises of svasaṃvedana and supports its ne-
cessity with an argument. To recall, the argument is, first, that perception has to be es-
tablished—known—in order to perceive its object, and, second, that perception cannot
be known in this way only after it has occurred because that would lead to an infinite
regress. For, any higher-order cognition would itself have to be known in order to
know a lower-order cognition, hence an infinite (forward) 59 progression would be gen-
58 “invariably”, or “necessarily”. Cf. Taber forthcoming for some remarks on the subject.
59 Cf. Siderits, “Dignāga's Philosophy of Mind”.

21
 02/28/16 – 01:34:57 PM

erated.

But why does perception have to be known if it is to apprehend its object? One
might expect that Dharmakīrti appeals to the analogy with light at this point, develop-
ing an argument along the following lines:

“What would it be like to be conscious of something without being

aware of this consciousness? It would mean having an experience with
no awareness whatever of its occurrence. This would be, precisely, a case
of unconscious experience. It appears, then, that being conscious is iden-
tical with being self-conscious. Consciousness is self-consciousness. The
claim that waking consciousness is self-consciousness does not mean
that consciousness is invariably dual in the sense that every instance of
it involves both a primary awareness and another instance of conscious-
ness which is somehow distinct and separable from the first and which
has the first as its object. That would threaten an intolerably infinite pro-
liferation of instances of consciousness. Rather, the self-consciousness in
question is a sort of immanent reflexivity by virtue of which every in-
stance of being conscious grasps not only that of which it is an aware-
ness but also the awareness of it. It is like a source of light which, in ad-
dition to illuminating whatever other things fall within its scope, ren-
ders itself visible as well.”60
The thesis that there is some kind of reflexivity to consciousness has widespread
support today, and certainly has strong support within phenomenology. 61 It is worth
noting, however, that when prompted to justify cognition’s reflexive awareness, Dhar-
makīrti offers a quite different line of argument:

PVin 1.54cd: apratyakṣopalambhasya nārthadṛṣṭiḥ prasidhyati ||

na hi viṣayasattayā viṣayopalambhaḥ, kiṃ tarhi tadupalambhasattayā. sā

cāprāmāṇikā na sattānibandhanān vyavahārān anuruṇaddhi. tadaprasid-
dhau viṣayasyāpy aprasiddhir ity astaṅgataṃ viśvaṃ syāt, sato ’py asid-
dhau sattāvyavahārāyogyatvāt. tasmān nānupalabhamānaḥ kasyacit
saṃvedanaṃ vedayate nāma kiñcit.

PVin 1.54cd: The perception of the object is not established for someone
who does not perceive its perception.62
60 Frankfurt 1988: 162, quoted from Gallagher and Zahavi 2012: 57.
61 See Zahavi 2005: 12, noting that the view that “the experiential dimension is characterized by a tacit
self-consciousness” was defended by all major figures in phenomenology, regardless of the partly
substantial differences between their views in other respects.
62 Or, with Taber forthcoming: “The seeing of an object is not established for someone for whom the

22
 02/28/16 – 01:34:57 PM

To explain: a perception of an object is not due to the existence of the

object, but due to the existence of its perception. And if the existence of
the perception is not established by a means of valid cognition
(aprāmāṇika), then it does not attach itself ( anuruṇaddhi)63 to forms of
behaviour that presuppose existence. If the perception is then unestab-
lished, then the object is also unestablished, so that everything would go
asunder, for even if something exists, it cannot be treated as existent un-
less it is established. Therefore, someone who does not perceive the
awareness of something is not aware of anything at all.64

Some semantic uncertainties notwithstanding, the argument is evidently that a per-

ception must be known so that certain forms of concept-involving behaviour can arise
that presuppose the existence of that perception. Given that perception itself is ac-
cording to Dharmakīrti devoid of conceptualization (kalpanāpoḍha), any conceptual
determination based on perceptual experience cannot be a part of it. Conceptual cog-
nitions that are experientially grounded must temporally follow after perception and
be caused by it (the relationship between perceptual content and concepts formed on
its basis is then further complicated by the apoha theory, but this need not concern us
here). Dharmakīrti thus argues here that if a perception is to induce any conceptual
behaviour, then that perception must be known, for, as he puts the main principle:
“even if something exists, it cannot be treated as existent unless it is established.”
Someone who has seen blue will only be able to conceptually determine the seen ob-
ject as blue when the seeing of blue was known.

Of course, one could object that this does not suffice to make reflexive awareness a
universal feature of all perceptual (let alone all mental) states because there could be
perceptions that simply fail to induce subsequent conceptual behaviour. Why would
these have to be known? One avenue for strengthening Dharmakīrti’s argument could
lie in drawing on the principle that all perceptions are followed by some conceptual-
ization, which is either erroneous – misidentifying the seen object – or correct (what

apprehension is not evident.”

63 PVin-t translates: “then a behaviour caused by existence does not occur” (yod pa’i rgyu mtshan can
gyi tha sñad ’jug par mi ’gyur ro). Taber forthcoming translates anuruṇaddhi as “... is not conform-
able to ...”.
64 Text and translation quoted from Kellner 2011: 420.

23
 02/28/16 – 01:34:57 PM

has been called “perceptual judgement”). 65 There wouldn’t be any perceptual dead-
end-roads, so to speak. Offering what might be a different avenue for saving the argu-
ment, Kamalaśīla (in a different context) argues that perception is only assumed to be
a means of valid cognition for the aspect of an object in regards to which a subsequent
determination arises. If perception does not produce a determination with regards to a
certain aspect of the object that it grasps, then that aspect is called ‘unapprehended.’ 66
In other words, if no determination follows, the previously perceived object is as good
as unperceived.

Is the sahopalambhaniyama-inference a strong argument against external objects?

A lot depends on its further analysis and explication, which I shall not go into here –
especially now that John Taber in a forthcoming paper is offering a much more thor -
ough and sophisticated philosophical reflection on the inference than I could ever as-
pire to produce.67 Taber there also notes how the appeal to reflexive awareness makes
Dharmakīrti’s analysis of consciousness consistent with widespread assumptions
among phenomenologists. However, phenomenologists are rather unlikely to share
the conclusion that Dharmakīrti draws from this: that cognition is not of external ob-
jects. And on Taber’s reconstruction, the sahopalambhaniyama-inference can be made
to work as a deductive argument against external objects (which would logically force
even phenomenologists to deny their existence).

2.3. Discrediting further evidence for external objects

When read as refuting externalism, the awareness-inference and the sahopalamb-

haniyama-inference are epistemological in nature: they then establish, respectively,
that cognition is not of an external object, and that the object is not different from its
cognition. While the first group of arguments are more specifically targeted against a
causation-cum-resemblance theory of perception, the second group can be taken to-

65 This can be derived from the apoha-digression of the PVSV discussed (among others) in Kellner
2004.
66 TSP ad TS 1972ff.: pratyakṣam aviśeṣeṇotpannam api sad yatraivāṃśe yathā parigṛhitākāraparā-
marśaṃ janayati sa eva pratyakṣam iṣyate vyavahārayogyatayā, yatra tu na janayati tad gṛhītam apy
agṛhītaprakhyam. See McClintock 2003: 157f., n. 21 for discussion.
67 Taber forthcoming.

24
 02/28/16 – 01:34:57 PM

gether as arguments that deny perception offers evidence for independently existing
external objects: perception cannot establish objects as independent from awareness.

There might of course be other evidence for external objects, provided by other
pramāṇas such as inference or scriptural authority. Setting aside Dharmakīrti’s reser-
vations about the pramāṇa-hood of scriptural authority in general, when we read that
the Buddha taught the “spheres” as external objects of perception merely in conform-
ity with the common understanding among ordinary people (PV 3.219), we may in
any case count this also as an attempt to discredit scriptural evidence for external real-
ity. Even if the Buddha’s words were fully authoritative in general, his teachings about
external objects have to be understood as governed by specific, limited intentions.
What could be inferential evidence for external objects? Here we might think of facts
that could simply not be explained without the postulation of external objects. In Va-
subandhu’s Viṃśikā, for instance, the fact that objects appear to us limited in space
and time, and not limied to one particular person, is initially presented as evidence for
external reality (which Vasubandhu then discredits).68 Kumārila similarly argues (in
the beginning of the nirālambanavāda chapter of the Ślokavārttika) that certain im-
portant scholastic and philosophical distinctions, as well as distinctions in ritual the-
ory would crumble if there were no external objects: the distinction between prior and
established position, or the distinction between valid and invalid cognition. 69

In other words, the external world could be proven as being simply the best expla-
nation for well-established facts of our experience, and as a prerequisite for various
types of theories about the world. Again, several of Dharmakīrti's arguments can be
taken to dispute this: the arguments from incongruence and the samanantara-
pratyaya-argument, for example, can not only be taken to refute an externalist theory
of perception, but also as demonstrating that external objects are not the best explana-
tion for why cognition is object-specific. Cognition’s form-possession, when accompa-
nied by a causal account that relies only on latent mental traces and causal processes
internal to consciousness, simply offers a superior explanation. Elsewhere Dhar-

68 Vś 2.
69 ŚV nirālambanavāda 1-3.

25
 02/28/16 – 01:34:57 PM

makīrti also adds that external objects are not required to account for the difference
between valid and invalid cognitions. Invalid cognitions arise from imprints left be-
hind by disturbed cognitions, and they therefore do not enable one to attain a desired
goal, which is what valid cognitions are supposed to do. Valid cognitions on the other
hand arise from strong imprints with an uninterrupted connection with desired goals
and are therefore reliable.70 There is a causal differentiation between valid and invalid
cognitions, one does not need external objects to account for it. Determining cause
and effect, and drawing inferences on the basis of causal relations, is equally possible
without assuming external objects, and this is actually the method preferred by the
“wise” (viduṣām).71

One might formulate an inference to prove external objects along the following
lines: When all other causes for perception are assembled, and perception still does
not arise, this implies that an additional cause is needed—and that further cause might
well be the external object. Some traditional interpreters construe this inference as a
Sautrāntika’s response to Yogācāra criticism, while others understand it to express the
Sautrāntika’s view that the external object is only inferred, and not perceived. 72 Dhar-
makīrti, in any case, expresses this inference in the hypothetical, and qualifies the con-
clusion: the missing cause might be the external object unless the Vijñānavādin should
claim that that additional cause is a special material cause of the cognition, that is, a
preceding mental episode in the same mental series that acts as the immediately pre-
ceding homologous condition.73 In other words: the non-arising of perception when a
certain number of its causes are present does not conclusively establish that the miss-
ing additional cause has to be an external object: it only does so if the possibility of an
internal cause is willfully ignored, or set aside. Seen in its context this, too, can be re -
garded as a way to discredit evidence for external objects. All these bits and pieces can

70 PVin 1 43,4–44,6; Krasser 2004: 143-144.

71 PV 3.392–397.
72 PVin 1 58d, elaborated in PVin 1 43,10–12; see also PV 3.390d–391ab (Krasser 2004 142–143). On the
various interpretations of the inference see Kyūma 2011: 314, n. 28.
73 PVin 1 43,11f.: … sa bāhyo ’rthaḥ syāt / yady atra kaścid upādānaviśeṣābhāvakṛtaṃ kāryavyatirekaṃ
na brūyāt / PV 3.390d-391: hetubhedānumā bhāvet / abhāvād akṣabuddhīnāṃ satsv apy anyeṣu hetuṣu
/ niyamaṃ yadi na brūyāt pratyayāt samanantarāt //

26
 02/28/16 – 01:34:57 PM

be seen as indicators that Dharmakīrti would not be inclined to accept that any
pramāṇa offers evidence of external objects, even though he does not confront the
question head-on, in one particular place of his works.

But if there is no reason to believe in objects outside of cognition, then it is reason-

able to assume that there are none. Would Dharmakīrti want us to draw this conclu-
sion? Did he think to have established that external objects do not exist, by having di-
credited all pramāṇas for them? Dan Arnold has Dharmakīrti rather argue for a
weaker position of “epistemic idealism”, as distinct from Vasubandhu’s “metaphysical
idealism.” Epistemic idealism, on Arnold’s use of the term, would subsume the
Sautrāntika and the Yogācāra model of perception, for it simply claims that “what we
are immediately aware of must be understood in terms of the intrinsic properties of
cognition”.74 In this connection Arnold drew attention to an intriguing passage in
Manorathanandin’s commentary on the PV which shows that within tradition, too, a
difference in arguments against external object has been recognized.
Manorathanandin recognizes that Dharmakīrti's arguments only go so far as to
demonstrate that “cognition appears, whereas an external object does not appear at
all”.75 The external object “behaves like a[n imperceptible] demon [and] is without a
means of valid cognition that proves it.” There is no evidence for it. But,
Manorathanandin goes on to say, “if the opponent were to strongly insist on negating
the [external object], he should be made to examine the master [Vasubandhu’s] nega-
tion of atoms according to whether one supposes that [the external object] has parts
or is without parts.”76 One way to read this passage is that Dharmakīrti’s arguments
(and it is not entirely explicit which ones Manorathanandin has in mind here, though
the sahopalambhaniyama-inference is certainly included) are weaker and only estab-
lish that there is no evidence for external objects, while Vasubandhu’s mereological
arguments do the extra work of denying their existence.

74 Arnold 2008: 15.

75 M1M1 220,17–20. Arnold 2008: 16. Note that I construe this sentence differently to Arnold.
76 See also Ratié 2014: 358, in critical engagement with Arnold 2008: 16. I am following Ratié’s reading
of the passage, but I regard some of the possible interpretations that she discusses as less compatible
with Dharmakīrti’s logical theory (see Kellner forthcoming b, as well as below).

27
 02/28/16 – 01:34:57 PM

But one can also read this passage differently (and here I am taking clues from Is -
abelle Ratié). Considering that these arguments are only for the stubbornly insisting
opponent, too obtuse to realize that Dharmakīrti has proven all there really needs to
be established, logical force might not be the main issue. Vasubandhu’s mereological
arguments, while not necessarily formally flawed, simply prove something that is no
longer of value once Dharmakīrti’s demonstrations have been considered: they enter-
tain the moot question as to whether an object for whose existence cognition offers no
evidence actually exists. But who in their right mind would still dare raising this ques-
tion? From this vantage point Dharmakīrti would not necessarily have put forward his
argument because he intended to support a weaker position than Vasubandhu. In fact,
the arguments discussed above show, when taken together, that while the Sautrāntika
and Yogācāra models converge on cognition’s direct awareness of an object-form,
Dharmakīrti’s arguments against externalism go further than just pointing out that
the direct objects of our awareness are internal to consciousness: they push for the
stronger point that there is no reason to believe in external objects, even though some
of the arguments individually might well be accommodated with weaker positions.

But why is the seemingly obvious conclusion that then follows not stated more
openly: that there are no external objects?77 A possible answer could be provided by
constraints of Dharmakīrti’s logical method. In the first part of my paper I drew atten-
tion to the fact that proofs of vijñaptimātratā are negative proofs, proofs of a negative
thesis: cognition is not of an external object, the object is not different from cognition,
and so forth. Now, Dharmakīrti is among the first Indian logicians to devote sustained
attention to how one can know nonexistence, and to whether nonexistence can be
proven inferentially. In his logical theory, a separate type of evidence (hetu) is reserved
for this purpose: non-apprehension (anupalabdhi). But what does the non-apprehen-
sion of a given object prove? Is the fact that something is not apprehended by any
pramāṇa sufficient evidence that it does not exist? Dharmakīrti would say: no.

In working out the scope of anupalabdhi, Dharmakīrti severely limits its applicabil-

77 PV 3.335d nārtho bāhyo ’sti kevalaḥ (with M1 attesting to the reading kevalaṃ) is the closest Dhar-
makīrti comes to stating that there is no external object, I think.

28
 02/28/16 – 01:34:57 PM

ity and effectively eliminates it as a method to prove the nonexistence of an entire

class of entities. One might wish to prove the nonexistence of “remote” (viprakṛṣṭa)
objects—objects that are distant in time or place, or by their very nature—on the
ground that they are not apprehended by any of the three means of valid cognition,
perception, inference, and scripture. That there is no scriptural statement proving the
existence of something cannot prove its nonexistence because scripture only teaches
what is relevant for a particular purpose. If something is not mentioned in scripture,
this cannot conclusively establish that it does not exist. Moreover, remote objects by
definition lack the capacity to produce a cognition of themselves—this is what distin-
guishes them from objects that are perceptible. Remote objects are therefore not of
such a kind that cognitions can be observed as effects proving their existence. Since
they therefore might exist without giving rise to a cognition of themselves, their non-
apprehension through a pramāṇa cannot establish their nonexistence.78 Moreover, a
general non-apprehension by all persons cannot prove anything, since it is not known
to oneself, nor to anyone else. Only one’s own non-apprehension of an object is evi-
dent, and when applying to remote objects, it is subject to the limitations just out-
lined.79 In its value as evidence, non-apprehension is limited to proving that objects
that would necessarily produce a perceptual awareness of themselves in a given situa-
tion where all other causes for the arising of that awareness are present can be justifi-
ably determined as absent. It can only establish the situationally specific absence of
particular objects, not the nonexistence of an entire class of objects.

Dharmakīrti’s logical theory, then, does not permit a straightforward ontological

denial on the basis of an argument from ignorance (the argument that concludes to
evidence for absence from absence of evidence).80 A Dharmakīrtian might wish to
claim that objects that are completely imperceptible are as good as nonexistent, but
when pushed to prove their nonexistence (s)he would have to resort to other argu-

78 PVSV ad PV 1.199 = 201; PVin 2, p. 62.

79 PVin 2 64,12–14.
80 In Kellner and Taber 2014, we argue that the Viṃśikā when read as a whole pursues just such an ar-
gument from ignorance against external objects, and we also suggest that, pace what contemporary
textbooks on logic state, arguments from ignorance are not always fallacious.

29
 02/28/16 – 01:34:57 PM

ments than the ones I have examined here. Such arguments could, of course, be pre-
cisely Vasubandhu’s mereological arguments against atomism found in Viṃśikā vv.
11–15. Dharmakīrti nowhere seems to make use of these, and at this point we must
leave the question open what he may have thought of them (which may or may not
coincide with Manorathanandin’s perspective). It is conceivable, in any case, that
Dharmakīrti's theory of non-apprehension might have acted as a kind of methodologi-
cal constraint on his approach to the external world, perhaps even pushing him fur-
ther to explore different and new approaches to its refutation than his predecessors.

3. Concluding Remarks: Phenomenology, Idealism, Both or Neither?

There is, obviously, no easy answer to the suggestive question that I have chosen as
the title of my paper, and certainly no general one that would apply to all theories and
arguments subsumable under the heading of Yogācāra-Vijñānavāda. For Vasubandhu,
John Taber and myself have recently proposed a new rereading of his Viṃśikā Vijñap-
timātratāsiddhiḥ that rehabilitates him as an idealist, in that the position he is arguing
for corresponds closely to one of Berkeleyan subjective idealism. Dharmakīrti’s argu-
ment strategy is quite different, and I have suggested that this is not because he pur-
sues a weaker philosophical position, but rather that he amasses numerous arguments
discrediting evidence for external objects without ever telling us straightforward that
they do not exist. My contention is that this might have to do with constraints im-
posed by his logical method. The probability that Dharmakīrti’s more complicated
stance can be reasonably explained through a method analogous to phenomenological
“bracketing” however seems to me relatively low, for this phenomenological method is
difficult to reconcile with the (I think evident) hierarchy of externalist and internalist
theories in Dharmakīrti’s overall approach, to the effect that the latter are simply more
accurate and truthful. The most intriguing aspect of what I referred to as Dhar-
makīrti’s second group of arguments against external objects, the awareness-inference
and the sahopalambhaniyama-inference, is that we encounter in them principles that
may well be widely accepted among phenomenologists (notably, reflexive awareness),
yet the conclusion that is drawn from them is quite a different one: there is no evi-

30
 02/28/16 – 01:34:57 PM

dence for external objects.

3.1. Bibliography

3.1.1. Sanskrit Sources and Tibetan Translations

Det Devendrabuddhi’s Pramāṇavārttikapañjikā. Tibetan translation. Tshad ma rnam ’grel gyi
dka’ ’grel, translated by Subhūtiśrī(śānti) and (Rma) Dge ba’i blo gros. D 4217 Che 1-326b4,
P 5717b Che 1-390a8.
M1 Dharmakīrti’s Pramāṇavārttika with a Commentary by Manorathanandin. Ed. by Rāhula
Sāṅkṛityāyana. Appendix to Journal of the Bihar and Orissa Research Society 24-26 (1938-
1940).
PrA’ A modern transcript of an incomplete paper ms. of Prajñākaragupta’s
Pramāṇavārttikālaṅkārabhāṣya, written by Vibhūticandra, extending from the commentary
on PV 3.302 to the end of the work. Microfilm of the Nepal-German Manuscript
Preservation Project in Kathmandu (reel no. A1219/26) (cf. Kellner 2009-2010: 167 and 168).
PV 3 Dharmakīrti’s Pramāṇavārttika, chapter 3 (pratyakṣa). See Tosaki 1979 (stanzas 1-319) and
1985 (stanzas 320-539).
PVin 1 Dharmakīrti’s Pramāṇaviniścaya. Chapters 1 and 2. Ed. by Ernst Steinkellner. [Sanskrit
Texts from the Tibetan Autonomous Region 2]. Beijing – Vienna: China Tibetology
Publishing House – Austrian Academy of Sciences Press, 2007. For Corrigenda, cf. Wiener
Zeitschrift für die Kunde Südasiens 51 (2007-2008) 207-208, as well as
http://ikga.oeaw.ac.at/Mat/steinkellner07_ corrigenda.pdf (last visited 26 June 2015).
PVSV Dharmakīrti’s Pramāṇavārttikasvavṛtti. The Pramāṇavārttikam of Dharmakīrti: the First
Chapter with the Autocommentary. Edited by Raniero Gnoli. Serie Orientale Roma 23.
Rome: Istituto Italiano per il Medio ed Estremo Oriente, 1960.
TS Dvarikadas Shastri: Tattvasaṅgraha of Ācārya Shāntarakṣita with the Commentary
‘Pañjikā’ of Shri Kamalashīla. 2 Vols. Varanasi: Bauddha Bharati 1968; reprinted 1981.
TSP Kamalaśīla’s Tattvasaṅgrahapañjikā. See TS.
Vś Vasubandhu’s Viṃśikā. Vijñaptimātratāsiddhi: Deux Traités de Vasubandhu: Viṃśatikā et
Triṃśikā. Edited by Silvain Lévi. Paris: Libraire Ancienne Honoré Champion 1925.
ŚV Svāmī Dvārikadāsa Śāstrī, Ślokavārttika of Śrī Kumārila Bhaṭṭa, with the commentary
Nyāyaratnākara of Pārthasārathi Miśra. Varanasi: Tara Publications 1978.

3.2. Bibliography

Arnold, Dan: Buddhist Idealism, Epistemic and Otherwise: Thoughts on the Alter-
nating Perspectives of Dharmakīrti. In: Sophia 47 (2008) 3-28.
Buescher, Hartmut: The Inception of Yogācāra-Vijñānavāda. Wien: Österreichische
Akademie der Wissenschaften, 2008.
Chu, Junjie / Franco, Eli: Rare Manuscripts of Works by Jitāri. In: Lasic, Horst (ed.):
Proceedings of the Panel on Sanskrit Manuscripts at the Fifth International Beijing Inter-
national Seminar on Tibetan Studies (August 1st - 6th, 2012). Beijing: China Tibetology

31
 02/28/16 – 01:34:57 PM

Publishing House, forthcoming, .

Coseru, Christian: Perceiving reality : Consciousness, Intentionality, and Cognition in
Buddhist Philosophy. consciousness, intentionality, and cognition in Buddhist philosophy.
Oxford [u.a.]: Oxford Univ. Press, 2012.
Coseru, Christian: Perception, causally efficacious particulars, and the range of phe-
nomenal consciousness: reply to commentaries. In: Journal of Consciousness Studies 22
(2015) 55-82.
Dreyfus, Georges B. J.: Recognizing reality. Dharmakīrti's philosophy and its Tibetan
interpretations. Albany, NY: State Univ. of New York Press, 1997.
Dunne, John D.: Foundations of Dharmakīrti's Philosophy. 1. ed. ed. Boston: Wisdom
Publ., 2004.
Frankfurt, Harry: The Importance of What We Care About: Philosophical Essays. Cam-
bridge: Cambridge University Press, 1988.
Gallagher, Shaun / Zahavi, Dan: The Phenomenological Mind: an Introduction to Phi-
losophy of Mind and Cognitive Science. 2. ed. Oxon: Routledge, 2012.
Garfield, Jay L.: Western Idealism through Indian Eyes - A Cittamātra Reading of
Berkeley, Kant, and Schopenhauer. In: Empty Words - Buddhist Philosophy and Cross-
Cultural Interpretation. Oxford: Oxford University Press, 2002, 152-169.
Garfield, Jay L.: Engaging Buddhism. why it matters to philosophy. Oxford [u.a.]: Ox-
ford Univ. Press, 2015.
Hall, Bruce Cameron: The Meaning of Vijñapti in Vasubandhu's Concept of Mind.
In: Journal of the International Association of Buddhist Studies 9/1 (1986) 7-23.
Hayes, Richard P.: Dignāga on the Interpretation of Signs. (Studies of Classical India,
Vol. 9). Dordrecht et al.: Kluwer Academic Publishers, 1988.
Iwata, Takashi: Sahopalambhaniyama. Struktur und Entwicklung des Schlusses von
der Tatsache, daß Erkenntnis und Gegenstand ausschließlich zusammen wahrgenommen
werden, auf deren Nichtverschiedenheit. Teil 1. Stuttgart: Franz Steiner Verlag, 1991.
Kellner, Birgit: Why Infer and not just Look? Dharmakīrti on the Psychology of In-
ferential Processes. In: Katsura, Shoryu / Steinkellner, Ernst (ed.): The Role of the Ex-
ample (dṛṣṭānta) in Classical Indian Logic. Wien: Arbeitskreis für Tibetische und Bud-
dhistische Studien, Universität Wien, 2004, 1-51.
Kellner, Birgit: Dharmakīrti’s criticism of external realism and the sliding scale of
analysis. In: Krasser, Helmut; Lasic, Horst; Franco, Eli / Kellner, Birgit (ed.): Religion
and Logic in Buddhist Philosophical Analysis. Proceedings of the Fourth International
Dharmakīrti Conference, Vienna, August 23-27, 2005. Wien: Verlag der Österreichischen
Akademie der Wissenschaften, 2011, 291-298.
Kellner, Birgit: Towards a Critical Edition of Dharmakīrti's Pramāṇavārttika. In:
Wiener Zeitschrift für die Kunde Südasiens 52/53 (2009-2010) 161-211.

32
 02/28/16 – 01:34:57 PM

Kellner, Birgit: Proofs of Idealism in Buddhist Epistemology : Dharmakīrti’s Refuta -

tion of External Objects. In: Tuske, Joerg (ed.): Indian Epistemology and Metaphysics. :
Bloomsbury Academic, forthcoming a, .
Kellner, Birgit: Proving Idealism in Indian Buddhist Philosophy: Vasubandhu and
Dharmakīrti. In: Ganeri, Jonardon (ed.): The Oxford Handbook of Indian Philosophy. :
Oxford University Press, forthcoming b, .
Kellner, Birgit / Taber, John: Studies in Yogācāra-Vijñānavāda Idealism I: The Inter-
pretation of Vasubandhu’s Viṃśikā. In: Asiatische Studien/Études Asiatiques 68/3 (2014)
709-756.
Kitahara, Yūzen: Ratnakīrti no Citrādvaitaprakāśavāda kōgai (*A Synopsis of the
Citrādvaitaprakāśavāda of Ratnakīrti). In: Journal of Indian & Buddhist Studies (=Indo-
gagu bukkyogaku Kenkyu) (Tôkyô) 87 (1995) 134-137.
Kobayashi, Hisayasu: Prajñākaragupta no senzai inshō setsu Prajñākaragupta's The-
ory of vāsanā. In: Indogaku bukkyōgaku kenkyū 50-1 (2001) 325-327.
Kobayashi, Hisayasu: Prajñākaragupta on what brings about diversity in cognition.
In: (ed.): The Annals of the Research Project Center for the Comparative Study of Logic.
Vol. 3. Hiroshima : Research Project Center for the Comparative Study of Logic, Hi-
roshima University, 2005, .
Kochumuttom, Thomas: A Buddhist Doctrine of Experience: A New Translation and
Interpretation of the Works of Vasubandhu the Yogācārin. Reprint 1989. New Delhi:
Motilal Banarsidass, 1982.
Krasser, Helmut: Are Buddhist Pramāṇavādins non-Buddhistic? Dignāga and Dhar-
makīrti on the impact of logic and epistemology on emancipation. In: Hōrin 11 (2004)
129-146.
Kyūma, Taiken: On the (im)perceptibility of external objects in Dharmakīrti's episte-
mology. In: Helmut Krasser, Horst Lasic, Eli Franco Birgit Kellner (ed.): Religion and
Logic in Buddhist Philosophical Analysis. Proceedings of the Fourth International Dhar-
makīrti Conference, Vienna, August 23-27, 2005. Wien: Verlag der Österreichischen
Akademie der Wissenschaften, 2011, 309-318.
Legrand, Dorothée: Two senses for 'givenness of consciousness'. In: Phenomenology
and Cognitive Sciences 8 (2009) 89-94.
Lusthaus, Dan: Buddhist Phenomenology. A Philosophical Investigation of Yogācāra
Buddhism and the Ch'eng Wei-shih lun. Oxon: RoutledgeCurzon, 2002.
Matsuoka, Hiroko: Santaraksita in Defence of the Alambanapariksa v. 2ab. In: Jour-
nal of Indian and Buddhist studies 61 (2013) 1241-1247.
Matsuoka, Hiroko: Śāntarakṣita on the Two Kinds of Arguments for Self-Awareness:
sahopalambhaniyama and saṃvedana. In: McAllister, Patrick; Kellner, Birgit; Lasic,
Horst / McClintock, Sara (ed.): Proceedings of the Fifth International Dharmakīrti Con-
ference (Heidelberg). Vienna: Austrian Academy of Science Press, forthcoming, .

33
 02/28/16 – 01:34:57 PM

McClintock, Sara: The Role of the 'Given' in the Classification of Śāntarakṣita and
Kamalaśīla as Svātantrika-Mādhyamikas. In: Dreyfus, Georges B.J. / McClintock, Sara
(ed.): The Svātantrika-Prāsaṅgika distinction: what difference does a difference make?.
Boston: Wisdom, 2003, 125-171.
Moriyama, Shinya: Ratnākaraśānti’s Theory of Cognition with False Mental Images
(*alīkākāravāda) and the Neither-One-Nor-Many Argument. In: Journal of Indian Phi-
losophy 42 (2014) 339-351.
Oetke, Claus: "Ich" und das Ich. Analytische Untersuchungen zur buddhistisch-brah-
manischen Ātmankontroverse. Stuttgart: Franz Steiner Verlag Wiesbaden GmbH, 1988.
Oetke, Claus: Doctrine and argument in Vijñānavāda-Buddhism. In: Wiener Zeit-
schrift für die Kunde Südasiens 36 (1992) 217-225.
Rahula, Walpola: Zen and the Taming of the Bull. London: Gordon Fraser, 1978.
Saccone, Margherita Serena: Śubhagupta on the Cognitive Process. In: Journal of In-
dian Philosophy 42 (2014) 377-399.
Schmithausen, Lambert: On the Problem of the External World in the Ch'eng wei shih
lun. Tokyo: The International Institute for Buddhist Studies, 2005.
Taber, John: Philosophical Reflections on the sahopalambhaniyama Argument. In:
McAllister, Patrick; Kellner, Birgit; Lasic, Horst / McClintock, Sara (ed.): Proceedings of
the Fifth International Dharmakīrti Conference (Heidelberg). Vienna: Austrian Academy
of Science Press, forthcoming, .
Tosaki, Hiromasa: Bukkyōninshikiron no kenkyū, Jōkan. *Studies in Buddhist episte-
mology, part 1. Tokyo, 1979.
Tosaki, Hiromasa: Bukkyō-ninshikiron no kenkyū, Gekan. *Studies in Buddhist episte-
mology, part 2. Tōkyō: Daitōshuppansha, 1985.
Watson, Alex / Kataoka, Kei: Bhaṭṭa Jayanta's Refutation of the Yogācāra Buddhist
Doctrine of Vijñānavāda: Annotated Translation and Analysis. In: Minamiajiakoten-
gaku South Asian Classical Studies 5 (2010) 285-352.
Wayman, Alex: Yogācāra and the Buddhist Logicians. In: Journal of the International
Association of Buddhist Studies 2/1 (1979) 65-78.
Zahavi, Dan: Subjectivity and Selfhood. Investigating the First-Person Perspective.
Cambridge: MIT Press, 2005.

34