This article was downloaded by: [Washington State University Libraries ]On: 27 November 2014, At: 20:56Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Asian Philosophy: An InternationalJournal of the Philosophical Traditionsof the EastPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/casp20

Naturalism and Intentionality: ABuddhist Epistemological ApproachChristian CoseruPublished online: 27 Aug 2010.

To cite this article: Christian Coseru (2009) Naturalism and Intentionality: A BuddhistEpistemological Approach, Asian Philosophy: An International Journal of the Philosophical Traditionsof the East, 19:3, 239-264, DOI: 10.1080/09552360903231018

To link to this article: http://dx.doi.org/10.1080/09552360903231018

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoever orhowsoever caused arising directly or indirectly in connection with, in relation to or arisingout of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Asian PhilosophyVol. 19, No. 3, November 2009, pp. 239–264

Naturalism and Intentionality:A Buddhist Epistemological Approach

Christian Coseru

In this paper I propose a naturalist account of the Buddhist epistemological discussionof svasam

_vitti (‘self-awareness’, ‘self-cognition’) following similar attempts in the

domains of phenomenology and analytic epistemology. First, I examine the extent towhich work in naturalized epistemology and phenomenology, particularly in the areasof perception and intentionality, could be profitably used in unpacking the implications

of the Buddhist epistemological project. Second, I argue against a foundationalist readingof the causal account of perception offered by Dignaga and Dharmakırti. Finally, I argue

that it is possible to read Dignaga’s (and following him Dharmakırti’s) treatment ofsvasamvitti as offering something like a phenomenological account of embodied self-

awareness.

Introduction

In this paper I propose a naturalist account of the Buddhist epistemological1

discussion of svasam_

vitti (‘self-awareness’, ‘self-cognition’) following similar attempts

in the domains of phenomenology and analytic epistemology. I want to examine the

extent to which work in naturalized epistemology and phenomenology, particularly

in the areas of perception and intentionality could be profitably used in unpacking

the implications of the Buddhist epistemological project. But I am also concerned

with naturalism more generally, and the ways in which specific models such as that of

embodied cognition, can benefit from some of the valuable insights of Buddhist

epistemology.

Let me begin, then, with what is fairly uncontroversial: the fact that Indian

epistemologies have never displayed the sort of non-naturalism characteristic of the

Cartesian tradition. As Mark Siderits reminds us,

the very idea of epistemology as praman_avada—that is, determining the number

and nature of reliable means of belief formation—suggests a non-foundationalistand externalist project. Non-foundationalist because the project presupposesthe existence of knowledge, instead of seeking to prove its very possibility.

Correspondence to: Christian Coseru, Department of Philosophy, College of Charleston, 66 George Street,

Charleston, SC 29424, USA. Email: coseruc@cofc.edu

ISSN 0955-2367 print/ISSN 1469-2961 online/09/030239-26 � 2009 Taylor & Francis

DOI: 10.1080/09552360903231018

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

And externalist because it seeks to distinguish between veridical and non-veridicalstates of the subject in terms of causal factors, and not in terms of states that arenecessarily accessible to the subject. (Siderits, 1995)

This focus on causal accounts of cognition is shared by all Indian epistemological

theories, Buddhist or otherwise. It is with Dignaga, however, that an examination

of the underlying processes of cognition becomes instrumental in determining which

inferential procedures are conducive to effective action (captured by the well known

theory of arthakriyasakti). By contrast, Western philosophers, beginning withDescartes, argue that justification, specifically the justification of reasons for why

certain beliefs ought to be classified as knowledge, is the main focus of epistemic

inquiry. What sets the two traditions apart seems to be the fact that, as has been

recently pointed out, ‘the distinction, common in Western thought, between the

causal question and the question of justification was not made by the Indian theories’

(Mohanty, 2000, p. 149).While Mohanty is right about the absence of this distinction in Indian

epistemology, this should not necessarily be seen as an unfortunate oversight.

Rather, I want to suggest that with the return to naturalism in epistemology (hence

to understanding cognition in causal terms) we are now in a better position to

appreciate the standpoint of Indian epistemologies. The fact that Indian

epistemologies did not make a distinction between the causal question and the

question of justification is indicative of the fact that epistemic inquiry in India is

primarily driven by pragmatic rather than normative concerns (that is, by concerns

about how we come to believe something rather than why might we be justifiedin believing it). Indeed, if Indian epistemologies treat as warranted only that

cognition which corresponds to its object and is produced in the right way, it seems

to me, they have a way of explaining epistemic dispositions as resulting from our

embodied condition (rather than as attitudes of a disembodied cogito serving

the justification of its beliefs). I will argue that naturalized epistemology and the

empiricist psychology that informs it share with the Praman_a Tradition of Dignaga

and Dharmakırti a common concern: developing a theory of reasoning that does not

divorce logical arguments from descriptive accounts of the psychology of cognition.

The ultimate source of these descriptive accounts is the Abhidharma, in itsspecifically Sautrantika-Yogacara synthesis of Vasubandhu.

One has to be careful not to read too much into a philosophical program whose

roots are historically and culturally different that those of the modern West. At the

same time one can, and perhaps should, profitably engage the thought of historical

(Buddhist) thinkers where such thought addresses perennial philosophical problems.But do so, mindful of Gadamer’s caveat against the objectivist claim that it is possible

to interpret the thought of a historical author in such a way as to suggest that the

interpreter does not enter in the event. Methodologically, then, we may identify

two predominant approaches, well captured by Richard Hayes in his study of the

overall scope of Dignaga’s philosophy: first, isolate ‘certain issues dealt with by the

thinker and to treat those issues in approximately the way that such issues are

discussed by modern philosophers, paying only minimal attention to the importance

240 C. Coseru

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

of those issues for the original philosopher’ (Hayes, 1988, p. 1). Second, determine

‘what significance a given range of topics had for him in his own historical context,paying relatively little attention to how or whether his treatment of those ideas

appeals to us in our own place in history’ (Hayes, 1988, p. 2).Neither approach, however, falls entirely outside the scope of comparative

philosophy, but the practice of comparative philosophy, where typically the methods,theories, and conceptual resources of one tradition are used for solving problems

in the other, all too often ends up with asymmetric and incongruous conceptualschemes. The typical response to the conundrums of comparative philosophy hasbeen to eschew it altogether in favor of new forms of scholasticism almost exclusively

concerned with text-critical issues and intradoctrinal disputes specific to any of themajor Asian philosophical traditions. A more interesting response, one that I pursue

in this essay, comes from Mark Siderits’ recent work in what he calls ‘fusionphilosophy’ (Siderits, 2003, p. 1). Simply put, fusion philosophy is the counter-

poising of distinct philosophical traditions for the purpose of solving problems thatare central to philosophy.

Whether or nor fusion philosophy holds the promise of making philosophy a trulyglobal and cross-cultural enterprise,2 the possibilities that it opens up, not just forsolving problems but also methodological dilemmas that are central to philosophy,

are numerous. From the perspective of fusion philosophy, then, it is legitimate toask whether the Buddhist epistemological program is to be treated as a system of

‘formal logic’, as a type of ‘cognitivism’, or as system of ‘natural reasoning’. Hence,the importance of correctly identifying the primary concerns of Indian and Buddhist

epistemology.An important first step in identifying the scope of epistemological inquiry in India

is the recognition that the systematic study of informal inference patterns isinextricably linked to the development of debate stategies, some of the earliest

examples of which can be found in Kathavatthu and the Nyayasutras of Aks_apada

Gautama. This is significant for at least two reasons. First, it showcases the primacy ofepistemology over ontology and metaphysics as the hallmark of Indian philosophical

inquiry. At least for the Buddhists, this means that ontological commitmentsare regarded as provisory and subservient to epistemological ends. The emphasis,

thus, is less on defending a specific ontology, and more on what we can knowgiven our reliance on an accredited source of knowledge (i.e. a praman

_a).3 Second,

it focuses primarily on the contents of experience, which take precedence overpurportedly objective accounts of reality, so that questions about external objects

are framed in terms of actions (including intentional acts) and their results.Indian philosophers have long recognized the importance of understanding

the psychology of individual cognizing agents when analyzing the form of inferential

patterns. The centrality accorded to perception as a reliable source of knowledge andthe practice of translating logical arguments back to their perceptual source are

the main reasons why Indian theories of inference have long been accused ofpsychologism, a derogatory label for the seeming conflation of logical reasoning

with the psychology of cognition mainly associated with Frege and the tradition

Asian Philosophy 241

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

of logical positivism. ‘Psychologism’ was first used to describe the claims of

19th century German philosophers such as Friedrich Eduard Beneke that empirical

psychology ought to serve as the basis of all philosophy given that philosophers arrive

at their truths by means of introspection. As Pelletier, Elio and Hanson (2008) argue

in their examination of what Quine (1969) called ‘the old anti-psychologistic days’

the newly expanded understanding of our cognitive architecture provided by

cognitive science makes it possible to be psychologistic about logic, albeit in a novel

way. I will briefly explore below some of the implications of this resurgent

psychologism for our reevaluation of the Indian theory of inference.

For now, let me draw attention to the consequence of using ‘logic’ (and, indeed,

‘epistemology’) in a transcultural sense: it can easily obscure the fact that the Indian

theory of inference (anumana), although neither a type of Aristotelian syllogistic

logic nor something similar to modern predicate calculus, nevertheless has its

internal order and coherence. As Mohanty has pointed out sometime ago, what we

have here is a ‘striking combination of cognitive psychology’ with the ‘needs of

a dialogical-disputational context and the strictly logical demand of validating a

belief’ (Mohanty, 1992, p. 105). Moreover, the recognition that nowadays syllogistic

logic covers just first-order logic and that there is a wide variety of ‘logics’ (modal,

temporal, counterfactual, many-valued, probabilistic, non-monotonic, fuzzy, etc.),

makes it possible to assume that the Indian theory of inference can either be

assimilated to one (or several) of these new logics or accommodated as a new type

of logic. Since we can think of intuitionistic logic (with its truth implying probability)

as psychologistic, it is more profitable to think of the nature of Indian theories

of inference in these terms.

That knowledge events could occur only if there is a proper causal connection

between the phenomenal content of experience and what we come to know, forms an

integral part of the praman_a theory. It is for this reason, I argue, that causal theories

of knowledge such as have been developed by Alvin Goldman and Fred Dretske

are relevant to understanding the role that Abhidharma accounts of cognition play

in Buddhist praman_a theory. Similarly, modifications to classical psychologism

proposed by Pelletier et al., which demonstrate that it is possible to explain logical

principles in terms of shared features of our cognitive architecture, could further

serve as a basis for reevaluating the scope of the Buddhist epistemological project.

Self-Awareness and Intentionality

Dignaga, and all philosophical thinkers following the tradition of epistemic inquiry

he helped initiate, makes an important and somewhat radical claim: a praman_a,

that is, a reliable cognition, is to be taken, against the received tradition of Indian

epistemological thinking, not as an instrument of knowledge such as perception

and inference exemplify, but rather as the reliable cognition ensuing from the exercise

of these cognitive modalities. For Dignaga, therefore, ‘a praman_a is effective only as

a result, because of being comprehended along with its action’.4 Thus the cognition

242 C. Coseru

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

that arises containing the representation of its object, while appearing to comprise

the action of cognizing, in fact is simply the result of cognitive activity.For instance, in apprehending an object, say a lotus flower, all that a cognizer

can be certain of is the representational aspect (akara) of that cognitive event, or,in Husserl’s terms ‘the intended object just as it is intended’ (vermeinten

Gegenstandlichen, so wie es vermeint ist) (Husserl, 1983, p. 312, Cx130). ForDignaga, thus, ‘that which appears in cognition, is the known object (prameya), the

reliable cognition (praman_a) and its result, are [respectively,] the subjective aspects

of [the cognition] (grahakakara) and the cognition [itself]’.5 Dignaga’s understand-ing of praman

_a in this context comes, I think, significantly close to something

like Husserl’s notion of noematic content (the ‘perceived as such’), which themethod of phenomenological reduction (epoche) is supposed to render access to.

For Dignaga, just as for Husserl, perception is ultimately constituted by intentionalcontent: perception, as self-awareness (svasam

_vitti), is a self-revealing (svaprakasa)

cognition.More specifically, according to Dignaga, every cognition has a double aspect, that

of a subjective appearance (svabhasa) and that of an objective appearance(vis

_ayabhasa). The subjective aspect (grahakakara) is just the individual’s self-

awareness as cognizing agent, while the objective aspect (grahyakara) represents the

intentional character of cognition or its object-directedness. I think Ganeri isgenerally right to claim that the problem of self-awareness (he calls it ‘knowledge

intimation’) cannot be addressed without taking into account the intentionality ofperception: ‘[for Dignaga] the double aspect theory of mental states is . . . motivated

as being the only way by which one can distinguish between thoughts and thoughtsabout thoughts, the intentional object of the latter being the subjective aspect of the

former’ (Ganeri, 1999, p. 474). I want to claim that Dignaga (and followinghim Dharmakırti and his commentators), while laying claim that self-awareness may

be said to be the only truly occurrent praman_a, does nothing more than advance

a peculiar representationalist epistemology closely aligned to a phenomenologicalaccount of cognition.

In a recent study that considers the essentially internalist account of cognition wefind at work in the Alambanaparıks

_a and Praman

_asamuccaya, Dan Arnold makes a

strong case for interpreting Dignaga as advancing ‘more like a full blownmetaphysical idealism than simply a representationalist epistemology’.6 given the

essentinally internalist account of cognition we find at work in the Alambanaparıks_a

and Praman_asamuccaya. His interpretation rests on the premise that svasam

_vitti,

which Arnold renders (following Kant) as ‘apperception’ is to be understood asendorsing the view that mental events are all that we can directly know. Such a viewwould, indeed, be deeply problematic: if apperception is the only accredited source of

cognition it becomes difficult to account for the difference between the particularaspects of the phenomenal content of experience. So Arnold asks: ‘How, that is, does

‘‘whatever appears as the content’’ of the cognition (yad abhasam_

prameyam) relateat once to the ‘‘cognition itself’’ (sam

_vitti) and its ‘‘subjective aspect’’ (grahakakara)?’

(Arnold, 2005a, p. 89). The problem with Dignaga’s account of self-awareness, claims

Asian Philosophy 243

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

Arnold, is that by assuming all cognitions (at least, all veridical ones) are perspectival,

one dispenses with the need to explain the subject–object relation. Presumably, even

though the Buddhist (unlike his Naiyayika opponent) is concerned with cognitive

aspects (akara) and not with ontologically distinct substances, the question of how a

subjectively phenomenal aspect can meaningfully tell us something about what

objectively appears remains unanswered. Does, then, Dignaga operate with some-

thing like the cognitive model of the sense–datum theorist, and is he, therefore, guilty

of creating an unbridgeable gap between the purely perceptual (hence, non-

conceptual) aspects of cognition and the purely discursive modes which seek to

explain it?I want to claim that it is possible to read Dignaga, and to a certain extent

Dharmakırti, as advancing something like a phenomenological account of cognition

such as we find particularly in Merleau-Ponty’s The Phenomenology of Perception.

Read through the perspective of a specifically embodied account of cognition such

as Merleau-Ponty offers, Dignaga’s understanding of the intentional character of

perceptual cognitions would then be less problematic (and also less amenable

to the sort of transcendental interpretation (of svasam_

vitti) that Arnold proposes).

I will argue, first, that Dignaga understanding of svasam_

vitti as a dual-aspect

cognition rests on a careful analysis of the phenomenology of first-person experience;

second, that it is possible to read Dignaga’s use of akara as something analogous

to Merleau-Ponty’s notion of body-schema; and finally, that Buddhist praman_a

theorists in general take the position that aspectual cognitions are inherently

intentional.Let me first clarify my suggestion that Dignaga understanding of svasam

_vitti is best

captured by the notion of ‘body-schema’. Against reductive models of embodied

cognition which assimilate the body to some kind of instrumental object, the

phenomenological method opens access to an investigation of the intentional relation

itself. From a phenomenological perspective, then, ‘body-schema’ and ‘body-image’,

although often used interchangeably in the literature, effectively intend different sorts

of objects. For instance, Gallagher identifies at least three distinctive features that set

apart the notions of ‘body-schema’ and ‘body-image’. First, the body image ‘has an

intentional status’ in that it is either ‘a conscious representation of the body or a set

of beliefs about the body’ (Gallagher, 1995, p. 228). By contrast, the body schema,

while having an effect on consciousness itself, it operates at the preconscious level.

Second, whereas the body image captures the sense of a body that is owned (or at

least one whose experiences are owned), the body schema functions at the

subpersonal level. Finally, the body image ‘involves an abstract . . . or articulated

representation of the body insofar as conscious awareness typically attends to only

one part . . . of the body at a time’ whereas the body schema functions in a ‘holistic

way’ (Gallagher, 1995, p. 229). Dignaga’s use of svasam_

vitti, as a cognition that

comes bound together with each instance of object-apprehension, and without which

no object (regardless of its ontological status) is apprehended as such, suggest a

pre-conscious, embodied and holistic function.

244 C. Coseru

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

An axiomatic principle of existentialist phenomenology in general (and of

Merleau-Ponty’s phenomenology in particular) is that we simply cannot conceptuallyarticulate the particular ways of our being-in-the-world. This is one of the reasons

why phenomenological accounts of cognition seek to go beyond the traditionalopposition between the ‘conceptual’ intentionality of belief and the ‘non-conceptual’

intentionality of sense-experience to the habitual matrix of experience which neither‘posits’ nor ‘represents’ its objects.

As Siewert notes in a reworked version of Merleau-Ponty’s account ofsensorimotor intentionality, however odd that might sound, it is possible to thinkof non-positional consciousness as being intentional without at the same time think-

ing of it as representational (Siewert, 2005, p. 274). Siewert claims, ‘the phenomenalcharacter of some of our experiences differs in ways we have reason to think cannot

be specified by attributing distinct representational content to the experiences’(Siewert, 2005, p. 276).

It was Husserl who first noted, while exploring the experiential background ofperception, that things in our perceptual field appear even before they are ‘seized

upon’. For instance, as I sit at my desk leafing through an old edition of Dante’sDivina Comedia I am surrounded by a whole range of other objects (books, papers,letters, pens, frames, memorabilia, etc.) which are perceptually present only as a ‘field

of intuition’. They are appearing even before they are properly perceived or seizedupon in attentive perceiving. Thus Husserl: ‘Every perception of a physical thing has,

in this manner, a halo of background-intuitions (or background seeings . . .), andthat is also a ‘‘mental process of consciousness’’ or, more briefly, ‘‘consciousness . . . of’’

all that which lies in the objective ‘‘background’’ seen along with it’ (Husserl, 1983,p. 70). When Husserl points us to the role that background intuitions play in

informing the phenomenal character of experience, he is also making an importantclaim about differences in the horizon of perceptual experience: less attended to

surroundings phenomenally look rather different than whatever we direct ourattention to. This difference is not easily verbalized, if at all, for in trying to ‘find out’what the surrounding areas look like one would have inevitably shifted the attention.

The resulting verbal report, therefore, is no longer about the initial surroundingsbut about whatever has now come into focus. On Siewert’s reading of this shift

in attention, ‘the effort to specify the character of the less attentive experience byattributing content to it invites a redirection of attention that risks assimilating the

prior experience to—and not distinguishing it from—a more attentive experience thatdiffers from it in phenomenal character’ (Siewert, 2005, p. 277). What does resisting

this assimilation (of the prior, essentially undefined and unrepresented experience)to the new experience that is fully attended to reveal that the resulting verbal reportcannot? As Siewert puts it, it reveals ‘a difference in character between the inattentive

experience of the visually apparent surroundings of what you are looking at andthe attentive experience of what you are looking at’ (Siewert, 2005, p. 278).

How might we profitably employ this insight about the intentional characterof perceptual experience for the purpose of explicating Dignaga’s account of self-

cognition (svasam_

vitti)? First, let me briefly review Dignaga’s position. As noted

Asian Philosophy 245

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

above, Dignaga claims that the difference between a cognition and its result is in fact

merely apparent: ‘we do not admit, as the realists do, that the resulting cognition(praman

_aphala) differs from the accredited sources of cognition (praman

_a)’.7 It is

only because ‘the resulting cognition arises bearing in itself the form of the cognizedobject’ that it is regarded as comprising the cognitive act itself, and is thus called a

reliable source of cognition. For Dignaga, this resulting cognition, whose contentbears formal resemblance to the apprehended object, is effectively the self-cognition

(svasam_

vitti) itself in its two-fold appearance, as subject and object. Furthermore,Dignaga tells us that what we ordinarily take to be the object of cognition is but theform or the phenomenal character of experience itself. This account of cognition

raises one important question: how do we know that cognition has this two-aspectualcharacter? Dignaga’s answer is quite categorical: because otherwise object-cognition

without self-cognition and self-cognition without object-cognition would beindistinguishable.8

I take Dignaga to be offering here something like a phenomenological accountof embodied self-awareness: we sense not only the properties of the objects with

which we come into contact, but also perceive ourselves as the locus of aqualitative experience (which explains why sensing is both transitive andintransitive). The intentionality of svasam

_vitti, then, may be understood as something

like the immediate sense of embodied agency that is characteristic of internal statessuch desire, pain, anger, ignorance, etc.

Recall that Arnold interprets Dignaga’s use of svasam_

vitti as advancing somethinglike a full blown metaphysical idealism. His interpretation is based on an overtly

Kantian understanding of the unity of apperception, which takes the faculty ofapperception to be the highest form of understanding comprising logic and

transcendental philosophy, if not ‘understanding itself’ (Kant, 1998, p. B134). Citingfavorably Pippin’s reading of Kant’s use of apperception ‘as a logical condition’,

in the sense that ‘all consciousness, including what Kant is calling experience, is aspecies of self-consciousness’ (Pippin, 1989, p. 20), Arnold seems to imply thatDignaga holds precisely such a view. At stake is Dignaga’s notion of svasam

_vitti

as a dual-aspect cognition, which Arnold (following Pippin) regards as vulnerable to‘the iteration problem’:

That is, if it is thought that any act of consciousness must, in order to countas such, be accompanied by a further act of consciousness, then the latter—again,if it is to count as such—must in turn be accompanied by a yet further act . . . .[T]his could serve just as well as a concise statement of thecharacteristically Madhyamika argument against Dignaga’s idea of svasam

_vitti.

(Arnold, 2005a, p. 82)

As we have seen above, the problem with this mentalistic conception of

consciousness is that it takes for granted the apparent structures of perceptualexperience. Rationalist traditions in epistemology (of the sort that Arnold draws

upon) still carry through, it seems, the old legacy of distinguishing betweentransparent consciousness and opaque sensation. Even Husserl, whose phenomen-

ological reduction was primarily conceived as a method for suspending belief in a

246 C. Coseru

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

natural world and examining not the objects that we ordinarily intend, but the pure

intentions themselves, remained committed, as Carman has recently argued, to‘the sort of intellectualism that in effect renders embodied consciousness metaphys-

ically unintelligible’ (Carman, 1999, p. 215).Echoing Merleau-Ponty’s view that the intentional gap between mind and world

can only be bridged in our understanding of bodily skills (and not through cognitiveattitudes, as Husserl thought), Carman captures rather well the non-positional

consciousness characteristic of our embodied condition. Thus, having a body doesnot consists in ‘having abstract thoughts about a body or concrete sensationslocalizable in a body, since embodiment is what makes possible the very ascription

of thoughts and sensations to subjects’ (Carman, 1999, p. 206). Now, Husserl is notan empiricist in the narrow sense of the term, nor does he subscribe to anything like

a sense-datum theory. For him sensations, while not the immediate and originalobjects of awareness, are however the basic constitutive material of consciousness,

but Husserl does not see bare sensations as having any intentional character.For that we must turn to Merleau-Ponty, who finds the key to our understanding

of sensorimotor intentionality in the notion of body schema, which he describes as an‘inter-sensory or sensorimotor unity of the body . . . that is not confined to contentsactually and fortuitously associated in the course of our experience . . . [but rather] is

anterior to them and makes their association possible’ (Merleau-Ponty, 1962, p. 114).Note that something like an analogous notion of schematic bodily representation

can be traced at least as far back as Kant’s Critique of Pure Reason (originallypublished 1781). Thus Kant’s ‘transcendental schema’, which is supposed to provide

a solution to the problem of bridging the gap between concepts and pure intuitions,is precisely the ground in which sensible concepts are anchored not as internal

representations but as the specific rules and procedures that govern the faculty ofimagination. For Kant, ‘it is not images of objects but schemata that ground our pure

sensible concepts’ even as he admits that how sensible objects appear to us by meansof these schemata, as well as their operations ‘is a hidden art in the depths of thehuman soul’ (Kant, 1998, p. A141). This much Kant does concede: that images are

the product of the empirical faculty of productive imagination, whereas schema area product of pure imagination itself. Unlike Kant, however, Merleau-Ponty does

not understand the conception of schema as explicit formal rules for negotiatingthe application of categories to experience. Rather, the schema for Merleau-Ponty

operates as ‘habit’ or ‘coping skills’—it constitutes the perceptual background thatconditions how we actually experience and respond to the environment we inhabit.

This embodied account of perceptual agency goes well beyond Husserl’s attempt toexplain intentionality by appeal to various hypothetical inferences and associationsbetween kinesthetic and outward-directed sensations. Rather, as Carman clearly

illustrates, our sense of embodiment ‘is nothing like an object-directed awarenessfocused on any of its distinct parts, as for example when we locate tactile sensations

on our skin or in our joints . . . [instead it] is bound up . . . with a primitiveunderstanding of the body as a global and abiding horizon of perceptual experience’

(Carman, 1999, p. 221).

Asian Philosophy 247

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

This double apprehension of the body as the locus of particular sensations and

at the same time as the horizon of perceptual experience comes very close, I think,

to Dignaga’s understanding of svasam_

vitti. Take the example of my two hands

touching each other: whereas Husserl would see it as an interaction between a

kinesthetic sense of voluntary movement and a passive sensation, Merleau-Ponty,

much like Dignaga, interprets the phenomenon as a dual-aspect perception that shifts

from an embodied sense of agency (effectively the result of past habituation) to being

the locus or ‘bearer of sensations’. It is only at the level of pure sensation, however,

that one can directly experience the inadequacy of the overtly mentalistic notion that

the body is a quasi-objective thing one identifies with by means of localized subjective

sensations. Thus, Merleau-Ponty: ‘If I wanted to render precisely the perceptual

experience, I ought to say that one perceives in me, and not that I perceive. Every

sensation carries within it the germ of a dream of depersonalization such as we

experience . . . when we really try to live at the level of sensation’ (Merleau-Ponty,

1962, p. 250).

Naturalism and Its Discontents

It is well known that the scope of epistemology itself has undergone radical change

with the advent of cognitive science in the 1960s. Significant advances in the

empirical study of consciousness and cognition have called into question both

the premise and the viability of a priori modes of inquiry. It was Quine who

first articulated this challenge in his celebrated 1969 essay ‘Epistemology naturalized’.

In a bold move, Quine proposed that epistemology be henceforth treated as a chapter

of psychology (that is, of what has come to be known as ‘cognitive science’ or ‘the

sciences of cognition’). The naturalism turn in contemporary epistemology,

therefore, is almost entirely informed by a growing body of scientific data pointing

to the fact that ordinary everyday experience does not constitute a reliable source of

knowledge. In its more radical form, naturalism represents an invitation to consider

the empirical evidence from psychology in settling questions about how we actually

form beliefs.

My proposal, then, is a relatively modest one: to examine whether the Buddhist

praman_a theorist’s acount of cognition shares any basic premises with naturalized

epistemology. As is well known, Quine’s influential analysis of the failure of

traditional epistemology to answer the challenge of skepticism ended with a proposal

that we abandon a priori reasoning and instead devote ourselves to simply studying

the psychological processes by which we form beliefs. Not surprisingly, his radical

proposal and the research program he eventually helped launch were met with strong

criticism. Jaegwon Kim, for instance, has argued on more than one occasion

that Quine’s proposal effectively amounts to abandoning justification and, with it,

‘the entire framework of normative epistemology’ (Kim, 1988, p. 48). Kim’s

trenchant criticism of the naturalism thesis follows four simple steps (adopted, with

some modifications, by most critics of naturalistic epistemology): first, set out the

248 C. Coseru

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

agenda of Western epistemology (as inherited from Descartes); second, explain the

strategy of this epistemological agenda as the (a priori) exercise of rationality itself;

third, show that naturalism strips epistemology of its knowledge claims; and finally,

introduce under the notion of supervenience the claim that normative epistemology

is not susceptible to ‘psychologistic’ reduction.Following Kim’s example, Laurence BonJour (1994, p. 62) argues that ‘skepticism

about the very possibility of a priori justification appears to undermine the rational

cogency of reasoning and argumentation’. Indeed, for BonJour claiming that there is

no a priori justification, as Quine does, obscures the fact that this claim itself needs to

be justified, something that cannot be plausibly accomplished a posteriori. Besides,

BonJour regards this claim as slightly odd if not downright suspicious coming from

someone so dedicated to rebuilding epistemology on a strong empiricist basis.

The main point of contention in BonJour’s rejection of the naturalistic paradigm

is the skeptical conclusion it inevitably upholds: that all our beliefs are based on

observational ‘sentences’ and that we have no way of determining whether or not

they are, indeed, true. What BonJour takes issue with is not so much Quine’s view

that the Humean condition is the human condition, but rather his rejection of

epistemic priority.

For Quine skeptical doubts are quintessentially scientific doubts, for it is science

and its empiricist program that reveal illusions (e.g. rainbows, afterimages, double

images) for what they are: phenomena that do not fall into the strict category of

genuine physical bodies and, as such, cast doubt on the very possibility of epistemic

certainty. BonJour sees skepticism about normative epistemology as a purely

philosophical move, one that challenges ‘the adequacy of our reasons for accepting

our beliefs’ (BonJour, 1994, p. 87), not the evidential basis of this acceptance (as may

be the case with an appeal to illusions).The main reason for rejecting the naturalist paradigm, claims BonJour, is that

it makes differentiating between the degrees of veracity of different classes of belief

irrelevant. Just as naturalized epistemology has nothing to say about the justification

of our scientific or even common sense beliefs about the world, so too it has nothing

to say about why we are justified in holding a particular set of religious beliefs,

something that any Buddhist philosopher presumably would take issue with. While

it may give us a descriptive account of how we come to hold such beliefs (allegedly

its main concern) it cannot tell us whether we have any reasons to prefer one set of

beliefs over another.Despite these criticisms, Quine has helped inaugurate a movement in epistemology

that strong advocates of naturalism such as Kornblith call ‘replacement naturalism’,

which now exists side by side with other, more moderate, forms of naturalism.9

The majority of philosophers, however, favor a weaker form of naturalism that

allows for evaluative questions about rationality, justification and knowledge to be

pursued in a traditional manner. Advocates of this later approach agree that the

sciences of cognition have much to contribute in resolving epistemological

problems.10

Asian Philosophy 249

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

Traditional analytic epistemology is foundationalist and normative.11 It has

inherited from Descartes its two criteria for establishing certainty: the conditions foraccepting or rejecting beliefs, and the conditions under which acceptable beliefs

count as knowledge. Note that the criteria for determining which beliefs are justifiedwere not initially established on naturalistic grounds. Having a ‘good reason’ or being

‘beyond reasonable doubt’ was sufficient for accepting the veracity of a given belief.Justification, thus understood, also explains why traditional (Western) epistemology

is a normative discipline: it does not simply tell us why we are justified in believingsomething, but also that it is reasonable and permissible to do so. If epistemologyis essentially a normative discipline, the story goes, it is because its foundation lies

on a bedrock of indubitable beliefs, usually about common sense experience and theimmediate accessibility of our mental states.

Now, unlike the traditional Cartesian epistemologist, the Buddhist praman_a

theorist (just like his Naiyayika counterpart) shares with Quine and other

proponents of naturalism the view that causality, not justification, is the determiningfactor in deciding which cognitive events count as knowledge. For Quine, causality is

understood in terms of the stimulatory conditions and situations that generatepropositional content. The first problem, in that case, is explaining how the cognitiveequivalence of sentences is possible, given the difference between universal sentences

and sentences expressing particular states of affairs. The second problem is the factthat even equivalent universal sentences of the type ‘Socrates is a man’ (or ‘Devadatta

is a human being’) are not always prompted by the same stimulations. For Quine,it is an individual’s disposition to treat sentences as cognitively equivalent even

if they are occasioned by different circumstances that informs his causal accountof cognition. On the other hand, for someone like Dignaga, who follows the

Abhidharmikas’ account of perception, there is an understanding that the perceptualaspect (akara), while causally related to the stimulus, represents more than is

available at the sensory level: it apprehends particulars (svalaks_an

_as) as uniquely

qualified phenomena, such that words and sentences can never truly designate, otherthan by a process of exclusion of all other possibilities (accounted for by the theory of

apoha). What Quine calls ‘occasion sentences’ (that is, sentences whose truth valuechanges as the occasion changes), might correspond to what Dignaga calls ‘verbal

utterances’ (vacana) of the sort that connect a word with a thing.12 Keep in mind,however, that the Buddhist, unlike Quine, generally warns against relying on our

ordinary language intuitions.Moreover, for the Buddhist the question of reliability is set out in pragmatic terms:

all actions, even impulsive and habitual ones (including utterances), have their sourcein preceding cognitions, so that the successful accomplishment of any human goal iswholly dependent on having correct knowledge. Operating with the premise that all

cognitive events are intentional (i.e. they are about some object of their own(savis

_ayaka)), Indian epistemological theories seek to identify the specific causal

order in which intentions translate into actual cognitive events. For the Buddhistepistemologist who follows Dignaga’s account of intentionality, the intentional act

250 C. Coseru

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

and the resulting apprehension of an object are not differentiated, such that a reliable

source of cognition (i.e. a praman_a) and its ensuing apprehension (praman

_aphala)

are in fact identical.

Praman_avada and the Naturalism Project

Leaving aside exegetical concerns about working out the best possible reconstruction

of the Buddhist epistemological project (one that is both historically and textually

accurate) critics might still point out that Buddhist epistemology operates on

soteriological rather than naturalist grounds: its ultimate aim is demonstrating the

validity of the Buddhist path and its corresponding metaphysical doctrines. Precisely

such a case is made by Steinkellner when he says that Dharmakırti’s motivation is

soteriological, in that ‘he wants to investigate whether a kind of ‘‘a progressive,

proleptic causality necessarily to be acknowledged as a real soteriological fact in the

conception of progress toward Buddhahood is supportable on a rational level’’

(Steinkellner, 1999, p. 350). Of course, on a purely doctrinal reading of the Buddhist

epistemological enterprise the soteriological concerns of the praman_ika might not

resonate well within the generally secular framework of modern scientific and

philosophical inquiry, despite the fact that reliance on perception and inferential

reasoning as the only accredited sources of knowledge seems to disqualify any appeal

to religious authority.Naturalizing Buddhist epistemology, then, amounts to more than simply placing it

in a modern psychological setting. Rather, it also means examining the presupposi-

tions of Abhidharma psychology, which inform it in a modern setting, while

at the same time recognizing its contribution to expanding our knowledge of the

phenomenology of first-person experience. This expanded knowledge can contribute

alongside the Husserlian phenomenological tradition to a widening of our

conception of nature currently at work in the sciences of cognition to include

consciousness and intentionality. For, as David Woodruff Smith (following Husserl)

has recently noted our encounter with, and categorization of, our cognitive activities

is prior to our theorizing in domains such as physics and neuroscience (Smith, 1999,

p. 83). An expanded conception of nature would allow for material and formal

categories to interweave, thus acknowledging the ontological complexity of our world

of experience. On such an expanded understanding of nature, a cognitive activity

such as perception is an intentional event which can be categorized materially under

‘event’ and formally under ‘intentionality’. So, following the practical efficacy model

of cognition (arthakriya), a perception has a subjective aspect (that is, is self-

revealing), carries intentional content, such as the perceptual image as intended,

and, if veridical, is directed to an actual object of experience.How are we, then, to understand Dignaga’s and Dharmakırti’s innovative synthesis

of logic and Abhidharma psychology? Why is it that, despite Nagarjuna’s and

Candrakırti’s critiques of epistemology and Wittgenstein’s invitation to abandon

Asian Philosophy 251

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

the illusions of the knowledge project, epistemology goes on? For the Buddhist

tradition the answer lies not simply in the need for establishing certainty—specifically, certainty about Buddhist metaphysical truths and the effectiveness of the

Buddhist path—but also in the pragmatic concerns that are at the heart of anyepistemology: what must I know in order to achieve a given goal?

A brief summary of Dignaga’s and Dharmakırti’s epistemological project shouldsuffice to make its pragmatic goals explicit. First, Dignaga gives us a most basic

reason why we ought to accept only two sources of knowledge as warranted: that is,because there are only two types of objects, those bearing their own characteristicmark (i.e. particulars (svalaks

_an

_a)) and those whose characteristic is that of

universality (i.e., universals (samanyalaks_an

_a)), each corresponding to a specific

mode of apprehension. What we have here is an attempt to translate the two

fundamental categories of particular and universal back to their specific knowledgesource: particulars to perception and universals to inferential reasoning. Dignaga,

thus, appears to be claiming that all objects of cognition, whether external or internal,cannot be properly understood if separated from the types of cognitive events in

which they are instantiated. This claim is, of course, consistent with Vasubandhu’sAbhidharma account, which states that perception, for instance, opens us to a worldof characteristically unique phenomenal properties: shapes, shades, textures—a

vast array of sensory experiences each having its distinctive qualitative nature(its svalaks

_an

_a). That we apparently come to apprehend this fragmentary and

discontinuous sense data either as enduring external objects or as universals is due tothe constructive function of conception.

However, Dignaga is not entirely faithful to Vasubandhu’s account of particulars.13

In Vasubandhu’s case, ‘attending’ (manaskara) to a particular results is understand-

ing it to have a specific characteristic, for instance, in the case of ‘matter’ (rupa),that of being disrupted by contact (rupan

_alaks

_an

_a) on the one hand. On the other

hand, attending to a universal is akin to reflecting on such basic characteristics ofentities as impermanence (at least, according to Buddhist metaphysical principles).Thus, whereas Vasubandhu understands particulars as types, Dignaga, it seems,

understands them as tokens. For instance, visual perception apprehends entitieswhose specific characteristics are visible types: shape, color, size, etc. For

Vasubandhu, these are all particular characteristics in the sense that they all havethe quality of being visible (rupayatana).14 With Dignaga, svalaks

_an

_a comes to

designate an utterly unique and ultimately real entity. The property of being visible,round, or having a color in the blue spectrum, are no longer regarded as types (of

which, presumably, there can be numerous tokens). The specific shades of blue andbloom-like shape by which something is apprehended as a blue lotus become forDignaga instances of uniquely characterized phenomena.

Reflecting on the type-token distinction, while tracing the semantic shifts of thenotion of particular from Vasubandhu to Dignaga and Dharmakırti, is helpful for

another reason: it shows that Dignaga and Dharmakırti are much more carefulto work out the implications for Buddhist epistemology of the debate among the

Sanskrit grammarians regarding the generation of meaning from verbal content.15

252 C. Coseru

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

For Dignaga the debate about the nature of word-meaning relations is important

for at least two reasons. First, it explains why he excludes from his definition

of perception the possibility that word-meaning relations could be perceptually

apprehended. Instead, Dignaga claims that verbal expressions are primarily

associated with conceptual cognitions, because they require the deployment of

categories such as genus and species, as well as reasoning and deliberation, all

of which are by definition outside the domain of sensory experience. Second,

it explains why Dignaga rejects the notion that universals can be perceptually

apprehended as relations of inherence: in perception we do not apprehend whole

objects, but only their ‘perceptible’ parts, and perception does not involves any

inferential component.16 For Dignaga, the debate over the nature of perceptual

objects is primarily centered on his analysis of the manner in which various classes

of names are attributed to phenomena as directly experienced. It is in direct

experience that he sees a total separation of perception from conception, which is

why he restricts the domain of the former to inexpressible particularities (viz.

phenomenal qualia) and the latter to universal properties (viz. genera). Of course,

Dignaga’s use of particulars and universals in this manner is not new. Vasubandhu

writes:

One examines the body in its proper and general characteristics, as well assensation, mind and constitutive elements of existence (dharmas). [Their] ownbeing is their proper characteristic. But the general characteristic is the non-eternity of produced [things], the fact that [everything that is] connected withthe [four] afflictions is suffering, and the fact that all things are empty and notthe self.17

Note that while a specific characteristic may be expressed in terms of a

phenomenal feature as apprehended, such as a unique shade or shape, it is also

structureless and momentary in nature. Perception, thus, enables us to apprehend the

specific particularities of objects as aggregated entities, but at the same time reveals

their momentariness. Inference, which relies on the testimony of direct experience,

extents this awareness to entities as classes of objects or as universals. Dignaga,

as it is well know, follows Vasubandhu in his explicit association of perception with

particulars. For Vasubandhu, direct perception in its dual aspect, as apprehension

of external objects and as awareness of internal states such as feelings and emotions,

also lacks any association with conception.

Now, by limiting the sphere of perception to indeterminate content only, Dignaga

(following Vasubandhu) parts with the received view (chiefly that of Nyaya) that

indirect perception is the sort of cognition naturally suited to connect objects and

universals. Dignaga sees this connection as being the function of inferential rather

than perceptual cognitions, a view that accords with the Abhidharma account of

cognitive dynamics. On this account, cognition results from a maturation of the

flow of discrete elements of consciousness. This flow of consciousness—called

‘mind-continuum’ (cittasantana) by the Sautrantikas and ‘storehouse consciousness’

(alaya-vijnana) by the Yogacaras—is ultimately seen as reflecting the dynamics of

Asian Philosophy 253

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

three types of action: volitional (cetayitva karman), vocal (vakkarman) and mental

(manaskarman). Vasubandhu again:

[I]t is through the force of its seeds supported on the material organs that the mindposterior to the absorption comes into existence. In fact, the seeds which give riseto the mind and to mental states rest according to circumstance on [one] of thetwo following series: the mental series or the series of the material organs.(Lamotte, 1936/1988, p. 58)

It is this mind continuum that effectively supports all cognitive activity and

generates the sense of continuity that comes with self-awareness. In its indeterminate

aspect, this mind continuum is nothing but the repository of a direct and non-

constructed form of cognitive awareness. For Vasubandhu, as for all Abhidharmikas,

apprehending subject and apprehended phenomena are emergent properties of the

threefold transformation of consciousness. The main point of this theory of cognitive

emergence, then, is to take into account the role that residual forces of past cognitive

events (vasana) have in ‘seeding’ this receptacle consciousness. On this model, it is

the dynamics of the residual forces of past cognitions, then, which generates the

intentional forms of apprehension that subsequent cognitions depend upon.

Exploring some possible parallels between this Buddhist account of cognitive

dynamics and cognitive science models of cognition, Waldron (2002) argues that the

notion of a ‘cognitive unconscious’ as the structuring principle of experience provides

a compelling account of the sort of coupling of world and its conscious apprehension

that the phenomenology of perception exhibits.18 It is possible, I think, to take the

notion of a ‘cognitive unconscious’ as indicating one way (the other being Huseerl’s

eidetic phenomenology) in which the Cartesian dilemmas of a thinking subject

seeking the justification of her beliefs can be overcome. Indeed, the psychological

account of cognitive maturation put forth in works such as the Sam_

dhinirmocana

Sutra suggests an intimate link between ‘the substratum of the material sense-faculties

along with their supports’ and ‘that which consists of the predispositions toward

conceptual proliferation in terms of conventional usage of images, names, and

conceptualizations’.19 As Waldron rightly points out, this model of the underlying

structures of cognitive awareness, is explained (in an encyclopedic work such as the

Yogacarabhumi) as the result of a threefold process of causal coupling in which a

residual cognitive awareness arises together with the sensory systems, and the habitual

tendencies generated by past linguistic experience such as conceptualization and

naming. Waldron (2002, p. 40) nicely sums up the explanatory gap that any theory of

consciousness must confront: ‘we live . . . in a world whose predominant structuring

influences—linguistic and physiological structures built up over time through

extended organism-environment interaction—we cannot fully discern’.Whether the Buddhist Yogacara model of cognitive dynamics can bridge this gap is

less significant than its implicit epistemic claim: namely, that cognition in the mode

of linguistic and symbolic representation is habitual and driven by the power of the

predispositions inherent in ordinary language expressions.20 In other words, ordinary

language expressions, which purport to describe the world of common sense

254 C. Coseru

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

experience, subtly condition the very way we commonly conceive the world.

Consequently, instead of an awareness of the mutual co-determination of self-cognition and cognition of object, which Dignaga’s dual-aspect model of cognition

exemplifies, we ordinarily experience a world of external objects standing over againstan apparent self.

Recall that Dignaga’s concern, unlike Vasubandhu’s, extends beyond causalaccounts of the cognitive process to questions of reliability and pragmatic efficacy.

His ‘spartan’ epistemology, which reduces the number of praman_as to perception

and inferential reasoning alone cannot be understood, therefore, apart from the(Abhidharma) cognitive psychological account that informs it. It is for this reason

that Arnold’s decidedly Kantian understanding of svasam_

vitti leaves our broaderdiscussion of the scope of the praman

_avada enterprise vulnerable to precisely the sort

of criticism he thus advances: that the tradition of epistemic inquiry initiated byDignaga and Dharmakırti with its ‘foundationalist emphasis on constitutively

caused perceptions . . . adduce[s] not reasons, but only causes’ (Arnold, 2005b, p. 51).I have addressed the question of whether the Buddhist praman

_avada enterprise is

properly characterized as a type of foundationalism elsewhere (Coseru, 2009). Whilearguing that the justification of knowledge—which is the primarily concern offoundationalism—falls outside the scope of praman

_avada enterprise, I claim that

it should be possible to offer an anti-foundationalist reading of the Buddhistepistemological project if we abandon the requirement that perceptual awareness

provide a justification for basic empirical beliefs.Granted Dignaga’s discrimination between sensory experience and self-awareness

(sva-sam_

vedana), we might conclude that self-awareness is simply a special mode ofnon-conceptual awareness.21 It is this non-conceptual awareness that provides direct

access to the data of experience, which, although non-conceptual, is meaningfullygiven. Without self-awareness providing an intentional anchorage for all other

cognitive events, cognition would happen in the dark. Cognition requires subjectiveapprehension and that, in turn, necessitates consciousness.

Thus, Dignaga, does not merely systematize into a system of epistemic warrants

what has traditionally been known as empirical awareness and self-awareness.He goes a step further and assigns to each cognitive modality a specific epistemic

role: perception does the job of apprehending particulars as uniquely characterizedphenomena, but only if operating in a non-constructive mode. If accounting for

the sense of intimation that accompanies each cognitive event is the goal, it is to thephenomenology of perception that one must turn for answers. Essentially, perception

appears in its dual aspect as awareness of something coupled with self-awareness.Thus Dignaga: ‘Every cognition comes about with a double appearance, namely thatof itself [i.e. as awareness of itself] and the appearance of the object. The awareness of

itself (viz. of cognition) as [possessing] this double appearance or self-awareness isthe effect [of the cognitive act].’22 On this definition, self-awareness is precisely the

type of cognition in which there is awareness that perceiving (or thinking) is takingplace without there being any engagement in it. Such is the case, also, with feelings

and memories.

Asian Philosophy 255

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

Recall that for the Buddhist epistemologists perception is not simply a

psychological process to be understood within the framework of classical

Abhidharma psychology, but an epistemic modality for establishing certainty.

Appeal to psychological observation is also at work in Dignaga’s formulation his

method of reasoning known as the triple inferential mark (trairupya). Central to this

method are the principles of ‘co-occurence’ (of the reason and the thesis) and

‘exclusion’ (of all other possibilities). So, in order to yield knowledge the logical proof

(hetu) must be present in the thesis (paks_a), i.e. in the position that is stated, be also

present in similar positions (sapaks_a), and be absent from dissimilar positions

(vipaks_a).23 In practical terms, this means that to prove a thesis, for example,

that sound is impermanent because it is a product of various causes and conditions,

what is needed is a thesis (A: sound is impermanent, because of being produced and

the contrary of the thesis (non-A: no instance of eternal non-produced sounds can be

found). A detailed discussion of the limits of Dignaga’s inductive method, its

criticism by the Brahmanical philosophers, and Dharmakırti’s attempt to reform it,24

is beyond the scope of this essay. For now, suffice it to say that appealing to empirical

observation as a premise ties logical reasoning to the ability to establish causal

connections between the things we directly experience.Exploring the limits of our ability to establish various causal connections

between the elements of experience is more readily associated with the psycho-

logical study of cognition, which for Dignaga is a legitimate move in the

overarching context of the Abhidharma psychology in which he operates. As a

result of this peculiar manner of addressing epistemic questions by appealing to the

mechanism of knowledge acquisition (and not simply by stating the formal aspects

of inference), Buddhist logicians appeared, at least to early generations of Western

scholars who came under the influence of logical positivism, suspect of

psychologizing tendencies.Following the return to naturalism in epistemology and significant recent

developments in cognitive science, it is debatable whether anti-psychologism is still

tenable as a philosophical attitude. Even the reservations that defining philosophy

too narrowly as ‘naturalism’ would make the broader questions of Buddhist (and

Indian) philosophy (viz. the goal of human life, the idea of emancipation from

suffering, etc.) irrelevant may be abandoned. In his now classical work, Perception.

An Essay on Classical Indian Theories of Knowledge, Matilal saw the empirical study

of cognition as having only marginal philosophical import: ‘‘‘What do we perceive

directly?’’ is, to be sure, not a scientific question. For it is not a question that is

generally answered by observation and experiment. It is a philosophical question.

Its answer requires conceptual analysis and philosophical argument’ (Matilal, 1986,

p. 224). Although he conceded that philosophers could not ignore the

experimental data, Matilal resisted the naturalist move in epistemology. Such

resistance may be legitimate when a historical reconstruction of classical

philosophical arguments is primary concern, but it is less profitable from the

perspective of fusion philosophy.

256 C. Coseru

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

Conclusion

Quine’s ‘naturalistic’ conception of knowledge began with an analogy meant to

illustrate the scope of his proposal: just as mathematics may reduce to logic and set

theory, so also our knowledge of natural kinds may reduce to sense experience.

This analogy assumes an implicit bifurcation of knowledge into conceptual

and doctrinal. The problem, as we now know, is that mathematics mostly reduces

to set theory, whose axioms, as Godel demonstrated, have less certainly than

logical principles. Similarly, our sense experience lacks the certainty that a

foundationalist epistemology would claim for it: basic, naıve beliefs about sense

experience are less certain than doctrinal positions about what is it that we actually

perceive. We have also learned that experience cannot be fully translated into

phenomenologically neutral terms. Should we, then, abandon epistemology

altogether, as the late Wittgenstein suggested? Is the view that there are real

epistemological problems simply a delusion that philosophers must be curred of?

Wittgenstein surely thought so, and so have Madhyamika philosophers from

Nagarjuna to Candrakırti and beyond, but Quine (much like Dignaga and

Dharmakırti) is quite categorical on this point: epistemology is an effective discipline

and brings about real results. So epistemology continues, thought in a new setting:

that of psychology. This is a new epistemology, one that is contained in sciences of

cognition rather than containing them.

Quine’s explanation of this mutual containment strategy hinges on the fact that

what should count as observation can now be settled in terms of our understanding

of the structure and dynamics of our perceptual systems. As we no longer need to

justify our knowledge of the external world through rational reconstruction,

awareness can finally be understood in causal terms. The empirical evidence from

vision science is, indeed, quite compelling. Take a few examples: first, color sensitivity

is only available in a small central region of the visual field, yet it does not seem as

though we perceive a difference in our experience of color saturation between the

central and peripheral regions of the visual field;25 second, there is a discrepancy

between the seeming richness and presence of the visual world and the rather

poor and fragmentary data processed by the visual system (see e.g. O’Regan, 1992,

pp. 484–485); finally, we ordinarily see the world as complete, dynamic, and

uniformly detailed, when, in fact, it is constructed out of momentary retinal images

within a certain threshold of awareness (see e.g. Blackmore, Brelstaff, Nelson &

Troscianko, 1995, pp. 1075–1076). Examples such as these are now routinely used to

support an epistemology that is conversant with the findings of cognitive science.It is out of the co-occurence of pre-reflective perceptual experience and an

awareness of perceiver and object perceived that questions about the cognitive status

of perception are born. The phenomenological tradition warns that moving all to

quickly over this pre-reflective awareness to an analysis of content of experience

results in object properties being determined in principle as being observer-

independent. Indeed, recent advances in perception research seem to suggest that

object properties are in fact the result of information processing of sensory input in

Asian Philosophy 257

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

the visual and somato-sensory cortex. The information at the level of retinal

receptors is quite different from what ‘we actually see’. In vision science, it is nowacknowledged that most of the actual processing is in fact unconscious, with the

visual awareness arising farther away on the pathways of sensory informationprocessing.26

Let me clarify the standpoint of my proposal. By setting the epistemology ofDignaga and Dharmakırti in a modern psychological setting, I do not mean to suggest

that we ought to prove or disprove their position in light of current scientific research.Nor is my intention to neutralize their arguments by showing that after all these arepre-modern philosophers whose ideas reflect the cultural concerns of their historical

time and place and thus have little to offer contemporary philosophers.Rather, granted that in the tradition of reflection initiated by Dignaga and

Dharmakırti arguments are already called to bear on the insights of Abhidharmapsychology, the Buddhist praman

_avada enterprise ought to be, at least in principle,

receptive to new findings in cognitive science. Conversely, cognitive science stands tobenefit from a tradition of philosophical inquiry with a rich history of first-person

phenomenological reflection, if the interdisciplinary study of the practice ofmindfulness meditation is any indication to this effect.27

While Buddhist meditative traditions may be regarded as supplying effective

methods for a phenomenological exploration of the constituent elements ofexperience, they cannot be divorced from the practice of epistemological inquiry

which has evolved to bear on the results of such experience. Take, for instance,Dharmakırti’s resemblance theory, which admits that our categorizing tendencies

follow primarily our aims and expectations but only within the constraints imposedby the essential qualities of things.28 If historically, Dharmakırti’s resemblance theory

may be seen as providing a middle ground between the extremes of nominalismand realism,29 within the broader research program of naturalist epistemology, it

may be seen as doing work that is similar to that of cognitive psychological theorieson the role of prototypes in concept formation (see e.g. Rosch, 1981, pp. 177ff).It also suggests some interesting correspondences to recent neuroscientific research

on the role of imagery in behavioral and cognitive simulation (for which, seee.g. Thomas, 1999).

Central to Buddhist epistemology from the time of Dignaga had been the view thatperception as self-cognition can be cognitive in a non-discursive sense. Consequently,

even natural occurrences of misperception have been attributed to the inferentialaspects of cognition. Whether this insight, which the Buddhist tradition had sought

to defend mainly on phenomenological and logical grounds, does find empiricalsupport, can now be settled beyond speculative arguments. So, let me finallyconclude with an example that looks at the ways in which imagery, for instance,

appears to interfere with and determine the contents of perceptual experience.Reviewing some of the neuroscientific data relating to the cognitive functions of

imagery, Hesslow (2002, p. 242) has recently observed that ‘imagining perceivingsomething is essentially the same as actually perceiving it, only the perceptual activity

is generated by the brain itself rather than by external stimuli’. The most impressive

258 C. Coseru

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

fact is that, the same is true also for simulations of action, which activate areas in the

brain in a manner similar to actual performance of physical action. Simulation,

using imagery appears thus be able to ‘elicit perceptual activity that resembles theactivity that would have occurred [original emphasis] if the action had actually been

performed’ (Hesslow, 2002, p. 242).To the extent that these cognitive scientific uses of imagery capture the sense

of Dignaga’s use of kalpana (usually translated as ‘conception’ or ‘conceptualconstruction’), it is possible to argue that indirect perception (savikalpapratyaks

_a) is

such that its contents are functionally indistinguishable from those of imagery

(kalpana). In other words, indirect perception is not free from the activity of imagery:it both apprehends and construes the object of perception. Direct perception alone,

as a cognition that is free of imagery (kalpanapod_ha) meets Dignaga demand for

providing access to the perceived as such. It is ordinary untutored perceptions (of the

indirect sort) that imagery interferes with, thus generating the illusion of aggregated

entities having intrinsic properties. The problem lies not in perception, whosemechanisms are generally not alterable at will, but in the awareness that attends to it.

Indeed, it is the latter that gives perceptual experience, when devoid of anyconceptual content, the directness that the Buddhist claims for it.

I have argued that in view of its analysis of cognition in causal terms, Buddhistepistemology shares with naturalism a common orientation. I have also argued that

the Buddhist tradition of epistemic inquiry does not divorce reasoning from an

understanding of the phenomenology of first-person experience. My discussion ofthe Buddhist notion of svasam

_vitti within a naturalized epistemological and

phenomenological setting may be taken, then, simply as suggesting a promisingdirection for a fusion philosophical approach to the problem of intentionality.

Notes

[1] For the purpose of this paper I use ‘Buddhist epistemology’ in a narrow sense as primarilyreferring to that preoccupation with establishing the sources of valid cognition historicallyassociated with Dignaga, Dharmakırti, and their Indian and Tibetan interpreters. Theoccasional use of ‘praman

_avada’ in the sense of praman

_a-doctrine or praman

_a-view, as well

as defining the chief scope of this classical Indian Buddhist philosophical enterprise,also attests to the difficulty of finding an adequate English equivalent.

[2] Various obstacles still stand in the way of such an enterprise given that, as recentlynoted, ‘we operate in the shadow of colonialism and its intellectual wing, orientalism’ asGarfield notes (2002, p. 230). The problem, as Garfield sees it, is not the legitimacy, let alonethe absence of works dedicated to translating and commenting Asian texts forEuro-American audiences, or vice versa, but the deeply problematic nature of theseintercultural philosophical ventures. This problematic concerns the types of inquiries thatmust accompany such enterprises, without which we cannot begin to talk about a singletruly intercultural tradition. Garfield’s list includes inquiries into ‘the historical conditionsunder which it takes place; into the respective characters of each of the textual traditionsbrought thereby into contact; into the relations they bear to one another; into the natureand the very possibility of the linguistic and cultural translation such interchange involves;and finally into the very possibility of reading a text into a tradition that is not one’s own’(Garfield, 2002, p. 231). While I agree that such inquiries are a valuable component of any

Asian Philosophy 259

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

cross-cultural philosophical enterprise, they strike me as overly restrictive when it comesto pursuing questions that are central to philosophy. Siderits’ fusion philosophy model, itseems, operates with the assumption that addressing central problems in philosophy can,and ought to, exceed hermeneutical concerns.

[3] As Matilal observes ‘praman_a (along with its possible interpretation as prama) is a very

important term in Indian philosophy. The sense of the word seems to be so comprehensivein some contexts and so limited in other contexts that it defies all our attempts at findinga happy English translation’ (Matilal, 1985, p. 203). Matilal renders prama as ‘true cognition’and pramatva as ‘truth’, which is one of the meanings of praman

_ya (i.e. the theory of the

means for the apprehension of truth or true knowledge). Ruegg following his brief surveyof various translations in French and German for praman

_a, notes that ‘en epistemologie

et en gnoseologie, la traduction la plus adequate (sinon parfaitement satisfante) de praman_a

demeure ‘‘(moyen/instrument de) connaisance/cognition correcte/exacte’’, ou encore‘‘connaisance droite’’’ (Ruegg, 1994, pp. 403–405). In this paper, I adopt the term‘knowledge’ for prama and ‘accredited’ or ‘reliable source of cognition’ for praman

_a.

[4] ‘savyaparapratıtatvat praman_am

_phalam eva sat’ (Hattori, 1968, p. 97n1.55).

[5] ‘yad abhasam_

prameyam_

tat praman_aphalate punah

_. grahakakarasam

_vittı trayam

_natah

_pr

_thak kr

_tam’ (Pramam

_asamuccaya I.10, translation per Hattori (1968, p. 29)).

[6] See Arnold (2005a, p. 88) and also Arnold (2005b, p. 35). In his very extensive andphilosophically interesting discussion of the concept of svasam

_vitti, Arnold (2005a) takes

into account Dignaga’s position, its refutation by Candrakırti and later reworkings of it byDharmottara, Santaraks

_ita and Moks

_akaragupta. I am not concerned with the historical

underpinnings of this debate, which Arnold explores in great detail and with subtlephilosophical acumen. Instead, I am primarily concerned with Arnold’s reading ofsvasam

_vitti in the Buddhist praman

_a literature as doing the same kind of work that the

notion of ‘apperception’ does for Kant.[7] ‘di la phyi rol pa rnams kyi bzhin du tshad ma’ (Hattori, 1968, p. 182). Translation per

Hattori (1968, p. 28).[8] See Praman

_asamuccayavr

_tti I k.11ab, text and translation in Hattori (1968, pp. 30, 185).

[9] See Kornblith’s (1997) edited collection Epistemology Naturalized, and his contributiontherein, and more recently his essay ‘In defense of naturalized epistemology’ (Kornblith,1999). For arguments against rationality as a foundational principle for traditionalepistemology see Stich (1984, 1990).

[10] See, for instance, Goldman (1992) for a critique of the notion of justified belief inepistemology, and his examination of the mutual relevance of epistemology and cognitivescience (Goldman, 1993). Goldman’s (2002, pp. 3–23) review of the internalist/externalistdebate in naturalized epistemology also includes an extensive critique of recent attemptsto play the a priori against the radical empiricism that naturalism appears to endorse.

[11] I follow Kim (1988), whose detailed account of the Cartesian legacy in epistemology clearlyspells out the divisions between the causal and normative questions.

[12] As Richard Hayes notes in his examination of Dignaga’s semantic theory, ‘the conclusiontowards which Dignaga argues is the fact that underlying the truth of the sentence‘‘Devadatta is a human being’’ is no more than the fact that, by whatever conventions theremay be that govern the use of symbols in a given language, the equivalent in that language tothe word ‘‘human being’’ is applicable to the individual Devadatta’ (Hayes, 1988, p. 205).

[13] For more on this see Dunne (2004, pp. 82f.).[14] See Abhidharmakosabhas

_ya ad Abhidharmakosa 10cd.

[15] There is no agreement among the early Abhidharmikas regarding the word-meaningrelation. The Sautrantikas, for instance, regard word types (naman), phrases (pada) andphonemes (vyanjana) as forms of speech (vac) capable of generating meaning. They alsoinsist that ‘speech’ should actually refer to any verbal utterance that calls a certain object toattention, that is, as something agreed on by a community of speakers. The Sarvastivadins,

260 C. Coseru

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

for their part, contend that word types are not speech, because speech is merely articulatedsounds and, by themselves, articulated sounds (e.g. a yell) do not make their objectcomprehensible (Cf. Abhidharmakosabhas

_ya ad Abhidharmakosa II, 46). Finally, the

Vaibhas_ikas regard word types as ‘constructs dissociated from thought’ (cittaviprayukta-

sam_

skaras). For the Vaibhas_ikas, thus, speech is nothing but vocal articulation (ghos

_a),

which, alone, cannot generate meaning (Cf. Abhidharmakosabhas_ya ad Abhidharmakosa II

47ab: ghos_asvabhavatvad vacah

_sarvam

_ghosamaram

_nama prakasayis

_yati adr

_so va ghos

_avis

_esa

is_yate namnah

_prakasakah

_sa eva arthasya dyotako bhavis

_yati /. text per Pradhan (1975)). Of

these three positions, that of the Vaibhas_ikas seems to come closest to the grammarians’

understanding of the word-meaning relation, particularly as developed in Bhartr_hari’s

theory of meaning disclosure (sphot_a). Cf. Pind (1991, pp. 278f).

[16] The question of whether or not a perceptual appearance (pratyaks_abhasa) is inferential

in nature is treated at length by the Naiyayikas. See e.g., Nyaya-sutra 2.1.30–2 andNyayasutrabhasya, Nyayavarttika, and Tatparyat

_ıka ad cit. For a detailed discussion of this

issue, see Matilal (1986, pp. 255–291).[17] Abhidharmakosabhas

_ya ad Abhidharmakosa VI, 14ab: kayam

_svasamanyalaks

_an

_abhyam

_parıks

_ate / vedanam

_cittam

_dharms ca / svabhava evais

_am

_svalaks

_an

_am/samanyalaks

_an

_am

_tu anityata sam

_skr

_tanam

_duh

_khata sasravan

_am

_sunyata’anatmate sarvadharman

_am

_/.

[18] Waldron borrows the notion of the ‘cognitive unconscious’ from Lakoff and Johnson(1999), which Waldron claims, I think convincingly, is structurally similar to the Yogacaranotion of alayavijnna.

[19] Waldron (2002, p. 39) quotes this passage following its reconstruction and translationin Schmithausen (1987, p. 357, n. 511).

[20] See Mahayana-sam_

graha, I. 58 as discussed in Waldron (2002, p. 40).[21] The first attempt to interpret svasam

_vedana as an aspect of mental perception is found

in Hattori (1968, p. 27). Similar interpretations have been proposed by Nagatomi (1979)and Franco (1993). Franco’s account is also a critique of Wayman’s (1991) view thatDignaga did indeed treat self-awareness as a different type of perception. Recently, Yao(2004) has reopened this debate by adducing some further evidence to support Dignaga’sendorsement of svasam

_vedana as a type of perception. Yao bases his claim primarily on the

Chinese translations of Nyayamukha by Xuanzang (600–664) and Yijing (635–713) as well ason Kuiji’s (632–682) commentaries on Dignaga’s principal works, all of which seem toindicate, in no ambiguous terminology, that Dignaga treats svasam

_vedana as a distinct form

of perception.[22] Praman

_asamuccayavr

_tti ad Praman

_asamuccaya I.1: shes pa ni snang ba gnyis las skyes te rang

gi snang ba dang yul gyi snang ba’o, snang ba de gnyis las gang dang rig pa de ni ‘bras bur‘gyurro, ci’i phyir zhe na. Text and translation (slightly altered for consistency) in Hattori (1968,pp. 28, 182).

[23] For a detailed treatment of the relationships that obtain between these three conditions seeHayes (1988, pp. 154ff), and Matilal (1998, pp. 90ff). A comprehensive analysis of thetrairupya, which traces its evolution across several sources and authors, includingVasubandhu’s Vadavidhi, Sa _nkarasvamin’s Nyayapravesa, Uddyotakara’s Nyayavarttika,Dignaga’s Nyayamukha, and Dharmakırti’s Praman

_avarttika and Nyayabindu, is found in

Oetke (1994).[24] For Dignaga’s formulation of the trairupya, see PS II, 5cd: anumeye‘tha tattulye sadbhavo

nasti tasti and translation in Hayes (1980, pp. 253f). See also Katsura (1983) for an extensivediscussion of the triple inferential mark.

[25] For a discussion of the philosophical implications of current research in color constancysee Thompson (1995, pp. 97–98).

[26] For a discussion of the dynamics of visual awareness, see e.g. Palmer (1999, pp. 616ff).[27] For a detailed survey of the recent literature on the topic of neuroscientific research on

various types of mindfulness meditation, see Lutz, Dunne and Davidson (2007).

Asian Philosophy 261

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

[28] See e.g. Hetubindu I.6.[29] As noted, in detail, by Dreyfus (1996, pp. 132ff).

References

Arnold, D. (2005a). Is Svasam_

vitti transcendental? A tentative reconstruction following

Santaraks_ita. Asian Philosophy, 15(1), 77–111.

Arnold, D. (2005b). Buddhists, brahmins, and belief. New York: Columbia University Press.Blackmore, S. J., Brelstaff, S. J., Nelson, G., & Troscianko, T. (1995). Is the richness of our

visual world an illusion? Transsaccadic memory for complex scenes. Perception, 24(9),

1075–1081.BonJour, L. (1994). Against naturalized epistemology. Midwest Studies in Philosophy, 19, 283–300.Carman, T. (1999). The body in Husserl and Merleau-Ponty. Philosophical Topics, 27(2), 205–226.Coseru, C. (2009). Buddhist ‘foundationalism’ and the phenomenology of perception. Philosophy

East and West, 59(4), 409–439.Dretske, F. (1995). Naturalizing the mind. Cambridge, MA: The MIT Press.Dreyfus, G. (1996). Can the fool lead the blind? Perception and the given in Dharmakırti’s thought.

Journal of Indian Philosophy, 24, 209–229.Dunne, J. (2004). Foundations of Dharmakırti’s philosophy. Sommerville, MA: Wisdom.Franco, E. (1993). Did Dignaga accept four types of perception? Journal of Indian Philosophy, 21,

295–299.Gallagher, S. (1995). Bodhi schema and intentionality. In J. L. Bermudez (Ed.), The body and the self

(pp. 225–244). Cambridge, MA: MIT Press.Ganeri, J. (1999). Self-intimation, memory and personal identity. Journal of Indian Philosophy, 27,

469–483.Garfield, J. (2002). Empty words: Buddhist philosophy and cross-cultural interpretation. Oxford:

Oxford University Press.Goldman, A. I. (1992). Liaisons: Philosophy meets the cognitive and social sciences. Cambridge, MA:

MIT Press.Goldman, A. I. (1993). Philosophical applications of cognitive science. Boulder, CO: Westview Press.Goldman, A. I. (2002). Pathways to knowledge: Private and public. Oxford: Oxford University Press.Hattori, M. (1968). Dignaga, on perception. Cambridge, MA: Harvard University Press.Hayes, R. (1988). Dignaga on the interpretation of signs. Dordrecht: Kluwer Academic.Hesslow, G. (2002). Conscious thought as simulations of behavior and perception. Trends in

Cognitive Science, 6(6), 242–247.Husserl, E. (1952/1983). Ideen zu einer reinen Phamenologie und phanomenologischen Philosophie:

Erstes Buch (Karl Schuhmann, Ed.). The Hague: Martinus Nijhoff. [Revised edition pubished

as: Ideas Pertaining to a Pure Phenomenology and to a Phenomenological Philosophy I (Frank

Kersten, Trans.). Dordrecht: Kluwer Academic, 1976].Kant, I. (1998). Critique of pure reason (P. Guyer & A. W. Wood, Trans.). Cambridge: Cambridge

University Press (Originally published1781).Katsura, S. (1983). Di _nnaga on Trairupya. Journal of Indian and Buddhist Studies, 32(1), 15–21.Kim, J. (1988). What is ‘Naturalized Epistemology’? In J. E. Tomberlin (Ed.), Philosophical

perspectives, 2, Epistemology. Atascadero, CA: Ridgeview.Kornblith, H. (1997). Epistemology naturalized (2nd edn). Cambridge, MA: MIT Press.Kornblith, H. (1999). In defense of naturalized epistemology. In J. Greco & E. Sosa (Eds.),

The Blackwell guide to epistemology (pp. 158–169). Malden, MA: Blackwell.Lakoff, G., & Johnson, M. (1999). Philosophy in the flesh. The embodied mind and its challenge to

Western thought. New York: Basic Books.

262 C. Coseru

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

Lamotte, E. (1936). Karmasiddhiprakara _na: le traite de l’acte de Vasubandhu, Brugers: Imprimerie

Sainte Catherine. (Translated as: Karmasiddhiprakara _na. The Treatise on Action by Vasubandhu

(Leo Pruden, Trans.). Fremont: Asian Humanities Press, 1988.).Lutz, A., Dunne, J., & Davidson, R. (2007). Meditation and the neuroscience of cognition.

In P. Zelazo, M. Moscovitch, & E. Thompson (Eds.), Cambridge handbook of consciousness

(pp. 499–554). Cambridge: Cambridge University Press.Matilal, B. K. (1985). Language, logic and reality: Indian philosophy and contemporary issues. Delhi:

Motilal Banarsidass.Matilal, B. K. (1986). Perception: An essay on classical Indian theories of knowledge. Oxford: Oxford

University Press.Matilal, B. K. (2000). The character of logic in India (J. Ganeri & H. Tiwari, Eds.). Oxford: Oxford

University Press.Matilal, B. K. & Shaw, J. L. (1985). Analytical philosophy in comparative perspective. Dordrecht:

D. Reidel Publishing Company.Merleau-Ponty, M. (1962). Phenomenology of perception (Colin Smith, Trans.). London: Routledge

[originally published in 1945 as: Phenomenologie de la Perception. Paris: Gallimard].Mohanty, J. N. (1985). Psychologism in Indian logical theory. In Matilal and Shaw (1985) 203–12.Mohanty, J. N. (1992). Reason and tradition in Indian thought. New York: Oxford University Press.Mohanty, J. N. (2000). Classical Indian philosophy. Lanham, MD: Rowman and Littlefield.Nagatomi, M., Matilal, B. K., Masson, J. M., & Dimock, E. C. (1980). Sanskrit and Indian studies.

Festschrift in Honour of Daniel H. H. Ingalls, Dordrecht and Boston: D. Reidel Publishing

Company.Oetke, C. (1994). Studies on the doctrine of Trairupya. Wien: Wiener Studien zur Tibetologie und

Buddhismuskunde.O’Regan, J. K. (1992). Solving the ‘real’ mysteries of visual perception: The world as an outside

memory. Canadian Journal of Psychology, 46(3), 484–485.Palmer, S. (1999). Vision: From photons to phenomenology. Cambridge, MA: MIT Press.Pippin, R. B. (1989). Hegel’s idealism: The satisfactions of self-consciousness. Cambridge: Cambridge

University Press.Pelletier, F. J., Elio, R., & Hanson, P. (2008). Is logic all in our heads? From naturalism to

psychologism. Studia Logica, 86, 1–65.Pind, O. (1991). Dignaga on Sabdasamanya and Sabdavises

_a. In E. Steinkellner (Ed.), Studies in

the Buddhist epistemological tradition: Proceedings of the second international Dharmakırti

conference, Vienna, 11–16 June 1989 (pp. 269–280). Wien: Verlag der Osterreichischen

Akademie der Wissenschaften.Pradhan, P. (Ed.). (1975). Abhidharmakosabhas

_yam of Vasubandhu (Revised by A. Haldar). Patna:

K.P. Jayaswal Research Institute.Quine, W. W. O. (1969). Epistemology naturalized. In Ontological relativity and other essays

(pp. 69–90). New York: Columbia University Press. (Reprinted in Kornblith, H. (Ed.). (1994).

Naturalizing epistemology (2nd edn). Cambridge, MA: MIT Press.).Rosch, E. (1981). Principles of categorization. In J. Haugeland. (Ed.), Mind design. Cambridge, CA:

MIT Press.Ruegg, D. S. (1994). La notion du voyant et du ‘connaisseur supreme’ et la question de l’autorite

epistemique. Wiener Zeitschrift fur die Kunde Sudasiens, 38, 403–419.Schmithausen, L. (1987). Alavijnana: On the origin and the early development of a central concept of

Yogacara philosophy. Tokyo: International Institute for Buddhist Studies.Siderits, M. (1995). Madhyamaka on naturalized epistemology. Paper presented at International

Seminar on ‘Concepts of knowledge East and West’, Calcutta, 9 January 1995.Siderits, M. (2003). Personal identity and Buddhist philosophy, Burlington, VT: Ashgate.Siewert, C. (2005). Attention and sensorimotor intentionality. In D. Woodruff Smith &

A. L. Thomasson (Eds.), Phenomenology and philosophy of mind. Oxford: Clarendon Press.

Asian Philosophy 263

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14

Smith, D. W. (1999). Intentionality naturalized? In J. Petitot, F. J. Varela, B. Pachoud, & J. Roy(Eds.), Naturalizing phenomenology: Issues in contemporary phenomenology and cognitive science.Stanford, CA: Stanford University Press.

Stich, S. (1984). Relativism, rationality, and the limits of intentional description. PacificPhilosophical Quarterly, 65, 211–235.

Stich, S. (1990). The fragmentation of reason. Cambridge, MA: MIT Press.Steinkellner, E. (1999). Yogic cognition, tantric goal, and other methodological applications of

Dharmakırti’s karyanumana theorem. In S. Katsura (Ed.), Dharmakırti’s thought and its impacton Indian and Tibetan philosophy: Proceedings of the third international Dharmakırti conference,Hiroshima, 4–6 November 1997 (pp. 349–362). Wien: Verlag der Osterreichischen Akademieder Wissenschaften.

Thomas, N. J. T. (1999). Are theories of imagery theories of imagination? An active perceptionapproach to conscious mental content. Cognitive Science, 23, 207–245.

Thompson, E. (1995). Color vision: Study in cognitive science and the philosophy of perception.(pp. 97–98). London: Routledge.

Waldron, W. S. (2002). Buddhist steps to an ecology of mind: Thinking about ‘thoughts withouta thinker’. The Eastern Buddhist, 34(1), 1–52.

Wayman, A. (1991). Dharmakırti and the Yogacara Theory of bıja. In E. Steinkellner (Ed.), Studiesin the Buddhist epistemological tradition: Proceedings of the second international Dharmakırticonference, Vienna, 11–16 June 1989 (pp. 419–430). Wien: Verlag der OsterreichischenAkademie der Wissenschaften.

Yao, Z. (2004). Dignaga and the four types of perception. Journal of Indian Philosophy, 32, 57–74.

264 C. Coseru

Dow

nloa

ded

by [

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es ]

at 2

0:56

27

Nov

embe

r 20

14