Introduction to Sapans ‘Treasury of valid cognition


Introduction to Sapans ‘Treasury of valid cognition[1]’.

____ Sonam Jamtsho [Dzongsar][2]


During what is historically known as a golden period in Indian civilization, numerous religious-philosophical schools came into existence including, among many other, Buddhism. If we carefully read and examine extant writing and literature of those periods, we can find that it is permeated by intensive, critical and intellectual engagement with rival schools, religious sect, etc. Dialogue and debate, guided by Indian logic, forms the central themes of those ancient manuscripts. For example Samkya, a Hindu school, tries to establish self whose nature is unchanging conscious, Vaishekas proposed a theory of atomism in which, inherently existing permanent entity that is the fundamental building blocks of empirically given phenomena. Theistic school arguing for the existence of an intelligent creator; a transcendence principle that created everything, Carvaka, a sort of ancient nihilistic-materialistic school, argues against possibilities of life after death and karma; and even going to doubted the empirical causality. Meanwhile Buddhists developed arguments against the existence of ‘Self’ , independent of the psycho-physical constituents. They also developed theory of rebirth, psychology, and cosmology, with each sub-school providing proofs for the validity of its own practices and theories.[3]

It was within this intellectual milieu that, Buddhist saint-scholars like Dignaga and Dharmakirti lived and wrote.[4]While there is no doubt that they are historical person, their exact date is a matter of speculation. Although Dignaga is regarded as the founder of the Buddhist logico-epistemological[5] tradition, Dharmakirti surpassed him both in literary outputs and influence. These two scholars, especially the latter developed a systematic and coherent presentation of Buddhist principle in response to non-Buddhist attacks. His central project is formulation of an inter-traditional standard of validation free the from prejudices and metaphysical assumptions of particular schools, in order to prove the validity of Buddhist principle like the Four noble truths, impermanence, the omniscience of Buddha, the existence of rebirth and karma etc. To this end he employs logical principal that is accepted universally in the intellectual environment in which he lived[6]. He was accredited with authoring what are commonly known as “the seven treatise on valid cognitions’’, seven technical treatise that deals with Indian logic (gtan tshig rigs pa), epistemology (blo rigs khi rnam bzhag), theory of philosophy of language (brdyul), relations and contradictions (‘drel ba dhang ‘gal ba), concepts formations (rtog pairnam bzhag), and perception, to name but a few. All of his major’s works were translated into Tibetan by the end of 11th century by Tibetan scholars collaborating with their Indian counterparts. Ngok lotsawa blo lden shes rab(1059-1109), a famed Tibetan translator is believed to be the first native Tibetan to write on Indian logic. His works was propagated and continued by many of his students and other scholars. Among these Phya pa chokyi senge (1182-1251) was the most formidable and innovative. This great scholar created many seminaries, composed texts, and developed many original academic methodologies which are still practiced in Tibetan monasteries. Monastic debates being one of the epitomes of his legacy[7].

Sa kya pandita kun dgrgyel mtshen (1182-1251, henceforth refered to as Sapan), was one of the most important figures to appear in Tibetan history. Sapan, politically is the first to rule over unified Tibet after the collapse of Tibetan empire in 842, albeit under Mongolian supervision. In the religious aspect of Tibetan buddhism, his Differentiation of three vows (sdom gsum rab dbye)[8] played an important role in bringing together the monastic and tantric strands of Buddhism in Tibet. Sapan is probably the most important Tibetan thinker in the field of logic and epistemology, and also one of the greatest commentators on Dharmakitian tradition. One of his main contributions to this field was the promotion of Commentary on valid cognition (pramanavarttika karika, tshad ma rnam ‘grel) as the primary texts of logico-epistemological studies. The primary of the Commentary on valid cognition in the Tibetan study of logic is result of Sapan’s influence. He is also probably the main source of all the Pramana lineages of Commentary existing today within Tibetan tradition[9]. Sapan’s masterpiece on logic and epistemology is his , Tshad Ma Rigs gTer, evinces his mastery of Dharmakirti’s difficult works. Instead of relying on the commentarial material and tradition as most of the Tibetan’s do, Sapans directly quotes the Dharmakirti’s work.

Prior to Sapan, Dharmakirit’s original works were mainly understood through the interpretive lens of Ngok and Chaba among the Tibetan scholars. Through his great proficiency in Sanskrit language and with his contact with Indian scholars.[10] Sapan realized the degree to which Dharmakirtis works[11]and philosophical position had been ‘polluted by the innovation of earlier Tibetan,’ as he addressed his predecessors. He saw the early interpretation from Ngok’s tradition’s as reading into the master texts positions that Dharmakiti either rejected or never propounded. He came to believe doctrinal purity of Indian Buddhist pramana[12] system has become littered with the Hindu tenets[13] and new concept developed by Tibetan. He saw those mistakes among earlier Tibetan not as exception but as a norm. It was with this knowledge of unfaithfulness of earlier Tibetan towards Indian works, Sapan, composed his master piece, “The treasury of valid cognitions’’. The positive purpose of his work is to propagate correct understanding of the Dharmakirti’s thought and to serve as a corrective measure to the wrong done by early Tibetan, and negative purpose is to refute and reject flawed interpretation that he detected in earlier Tibetan[14]. His stringent criticism overall is mainly against the interpretation of what is essentially anti-realist philosophy as a realism and the epistemological and logical implication of such diversion. He works strenuously to correct these mistake. Realism here refers to reification of universal. Dharmakirti relegates these generals to quasi-phenomena, unlike his Nyaya and Samkhya counterpart, a sort of conceptual linguistic construct. It does not have any ontological bearing or status. His Hindu adversaires on the other hands reified the constructs and gave an ontological bearings , which was also accepted by earlier Tibetan albeit in a moderate form, who nevertheless simultaneously claimed to be the follower of Indian Buddhist pramana system.

The Treasury is arguably the most authoritative and coherent commentarial treatise exposing the pramana system of Dinaga and Dharmakirti, from native Tibetan scholar. In it major ideas covered up in Indian source are discussed. The text is divided into eleven chapters of unequal length, including preamble and colophon. The root texts were written down in verse form and were famous for its brevity, pithiness and mnemonic nature. This monumental work is also accompanied by an auto commentary (though the authenticity of it is challenge by other). The Treasury is divided into two main sections; 1) ascertaining the object and, 2) ascertaining the nature of cognition that comprehend the objects. The former is again subdivided into three parts; a) the object of comprehension, b) cognitions that comprehend it and c) the manner in which cognitions comprehends its objects. According to Buddhism, ignorance is the root of all suffering and dissatisfaction, which are the pervasive characteristics of our experience. To get convince of this hypothesis one must undertake to investigate several subjects. In the Buddhist Pramana tradition this fundamental Buddhist premise is subjected to logical analysis. Ignorance is mistaken perception. Thus it is essential to analyze the perceived objects as well as perceiving subject and the ways in which latter comprehend the former. Since ignorance which is constituted by mistaken perception and since it is the root of our suffering, in order to get rid of it, its roots must be extirpated. This process presupposes the knowledge about the object (Yul), subject (Yul Can) and relations between them (Yul Yul Can Gyi ‘Brel Ba).

Object of comprehension (prameya,bzhal bya) is taken up in the first chapter. Here objects are defined as ‘possible object of valid cognition.’[15] It is crucial here to understand what exactly is meant by  object. Within the overall system, there seems to be a degree of ambiguity with regard to Dharmakirti and Sapan ideas of what it means to be exists. It arises out of Dharmakirti’s division of his ontology (yul) into two types of object which does not exist at same level. And he provides two different criteria of what is mean by the term ‘’Yul[16]” or ‘existence’. The first one is something that can be apprehended by valid cognition, while the second one is object that has efficacy. In the first sense intelligibility is primary while in the second existence means having a casual capacities. Two types of objects are specifically characterized phenomena (svalaksana,rang tshen) and generally characterize phenomena (samanyalaksana,spyi tshan). First one are real things substantially existing (rdzas yod) and second one are conceptual entities like universals (samanya,spyi) that are mere construct. Moments of consciousness and infinitesimally small atoms are illustration of the first, while appearing object of conceptual thought and direct/actual signification of our language are illustration of the second. The first types also has a definite identities of three determinate,i.e, they have specific spatio-temporal dimension and nature(yul dus rang bzhin ma ‘dres pa), while the generally characterized lack this. Dharmakirti and Sapan mainly understand concepts of existence as equal to having a casual capacities[17]. This understanding in turn excludes second types of object, namely generally characterised, from being existence. Yet, Dharmakirti includes the second type in his typologies of object. And also there is a problem of whether our common sense objects does exist independent of our epistemic practices or not (Blo la ma ltos pa’i ngo bo). What does it means to be common locus of existence and that is merely construct? It strike at the heart of ontological status of universals, object of our general noun, taken up in the third chapter. This seems to be one of the most problematic issues as it is inextricably linked with the conceptual thought and language and with logic.

This is the general background of the first chapter of the texts. Immediately, in the same fashion with other writer of his time, he structured his work with the three heading; refutation of other’s system, presentation of one’s view and refuting possible criticisms (dgag bdzhag spong gsum). His main target is early Tibetan who divides ‘held object’(gzung yul)-an object which serves as one of the three necessary conditions[18] in production of sense perception-into three different types. In their presentation of these particular objects, conceptual thought and mistaken non-conceptual consciousness[19]are said to possess held object on top of usual non-mistaken sense perception. This he sees as wrong reading of Indian precursors. According to Sapan, only external material objects of our sense perception, is considered as held object by Indian masters. He quotes directly from Dharmakirti to substantiate his claims. He flung/hurl two unacceptable consequences to his opponents based on their own premises. He rhetorically questions them to accept that conceptual thought and mistaken sense perception to be non-mistaken because they have held object like sense perception. If they accept this, then they are in flagrant contradiction with the cardinals tenets of Dignaga and Dharmakirti, who are ultimate source of Buddhist Pramana system. His second attack is that if these mental episodes have held objects, those objects should be also apprehended by someone nearby under normal circumstances. If they accept this, an obviously ridiculous position, they are taunted as unworthy of meaningful discussion, because objects of conceptual cognitions and mistaken perception are limited to person who experiences those cognitions. They are encroaching rational boundary into wilderness. It is normal to find some scathing remarks of various lengths after his successful refutation. So what is the rationale behind early Tibetan to take this position? First of all Sapans and his orthodox followers thinks that the concepts of held objects is only accepted by and taught within the Sautrantika (mdo sde pa) system, within the overall works of Dharmakirti[20]. And within this system it is only first two perception[21] that has held objects which is essentially material objets that cast there aspects (rnam pa gtad pa) towards apprehending perception. This understanding of the causality between sense perceptions and their corresponding objects implicitly underscores the representationalism (shes pa rnam pa dang bcas pa). Since the objects and subjects are cause and effect, for Dharmakirti they are necessarily sequential. Early Tibetan commentators seems to be advocating a theory of sense perception that is at variance with the Dharmakiti’s thought. Chapa is famous for his critique of this theory. For him the object subject co-exists, which is denied by Dharmakirti. The radical time gap between the object and subject is unacceptable for Chapa and he advocates something that is charitable view of our common intuitions. We think that object we see are present before us ‘now’.

Though Sapan did not develop, his followers like Gorampa(1429-1489) and Sakya Chog lden developed a theory of four typology of objects. They are appearing objects (snang yul), held objects (gzung yul), conceived objects (zhen yul) and objects of engagements (‘jug yul). According to Gorampa all conceptual thoughts (rtog pa) has appearing objects; the universals that directly appears to it. But the case with the non conceptual objects are different, non mistaken non conceptual cognitions like four perceptions has appearing objects but mistaken perceptions like the sense perception of two moons does not. Only mental perception and sense perception has held objects. They are the material objects that give rise to or act as the objective condition for these two perceptions. Conceived objects are purview of conceptual thoughts. Last one is objects of valid cognitions and subsequent certainty (rjes kyi nges shes) that is induce by them.

He then proceeds to presents his own reading. Here he makes one of his most startling claims. Sapan write’s “Specifically characterized phenomena are the only objects of cognition.”[22] It is  very controversial statement and some of his critics like Bodong went far to dubbed him as nihilistic in his ontology, who denies the validity of inferential knowledge, one of the two types of valid knowledge supported by Dharmakirti tradition. Since inference depends upon reasoning which is general properties and since Sapan seems to denies their existence he also denies the validity of inferential knowledge! Sapan quotes abundantly from Indian source to corroborate his own theory, though what he quotes is not free from inherent ambiguities. Sapan in the auto-commentary to the verse says, “Only specifically characterised are the objects of consciousness,’’[23] and he seems to thinks that it is so because only this type of phenomena has the capability to fulfill the practical needs of beings. Gorampa while commenting on this passages makes a technical distinction between objects of modes of apprehension (‘jug tshul gyi gzhal bya) and objects of engagements (‘jug yul gyi gzhal bya). He also introduces the concepts of conventional valid cognitions (kun tu tha snyad pa’i tshad ma) and explains that the Sapan’s comments has to be read as explaining the objects of this type of cognitions, which takes specifically characterized phenomena as its object since this types of objects alone has the capabilities to fulfill practical needs.

As mentioned earlier Dharmakirti divides the objects into two categories, namely generally characterized and specifically characterized. Based on this binary division, valid knowledge is also divided into two as perception and inference.[24] While perception without any mediation of concepts and language relates with real entities, inference which is other valid knowledge relates to reality only through the mediation of universals. But if universals are only conceptual constructs that lacks the identity criteria of determinate spatio-temporal and nature, how can inference ever be valid given it only touches upon those constructed universals? This is probably why early Tibetan revised the more radical rejection of universals and turned it into a moderate theory of universals. Because denying universals seems to imprison inferential knowledge into conceptual realms without any relevance to real world. Though this revisionist theory of universal is helpful to ground inferences and other conventional practices which assume the existence of universals, it is very problematic as the orthodox interpretation of Dignagas and Dharmakirtis works. On top of which there seem to philosophical problems of how to account for relation between universals and individual which instantiate them.

Under the third heading some real and virtual polemics against his system is refuted. Here Sapan’s briefly touches upon the general framework he uses to explain whole of the Buddhist canon.[25] It is Sapan’s own allusion to his basic position on hermeneutics, which consist of three types of world view with respective ontological commitments or lack thereof and its implied epistemological theory. They are 1)accepting the validity of external world, 2)entering into conventional reality,i.e, mind only theory and 3)entering into ultimate.[26] Sapans is implicitly saying that the first is the position adopted by Sautrantika, second by Cittamatra and the third by Madhyamaka. Sapan again get crossed with the early Tibetan who based their theory of perception on common sense. He says that if we base our epistemological theories on common intuitions, then there is no need for the existence of philosophically reflected system develop in seven treatise. He does not name his opponents but the opponents seems to propose an epistemological theory very similar to Chandrakirti based not on theoretical reflection but on the common intuitions.

He also points the ultimate tenets expounded in the work of Indian Pramana teacher as Mind only school (sems tsam pa), idealistic school of Mahayana Buddhism.[27] Here he ultimately rejects the reality of external world and develops idealistic world view based on doxographical tenets of Mind only school.He briefly provides some rationality in establishing mind as the main factor in our experience and non-existence of the world independent of perceiving mind. Two reasons are ‘sign’s of clarity and sign’s of simultaneous apprehension,(gsal rig kyi rtags dang lhen cig dmigs nges kyi rtags). Both are technical reason that entails subjectivity of everything that comes in our experiences. Sapans also develop a theory explaining and differentiating valid experiences from other, given if all we experience are our subjective projection. This arises due to difference between firm and unstable mental imprints on our minds (Bag chag brtan pa dhang mi brtan pa). The first chapter is mainly based on the theory of object as explicated in the initial parts of chapter on perception in Compendium (pramanasammuccaya, tshad ma kun las btus pa) and Commentary.

In the second chapter of the text consciousness that comprehends the objects is discussed.

Consciousness is defined as clarity and knowing.[28] Next consciousness is classified in different ways, conceptual and non- conceptual, mistaken and non-mistaken etc. Non valid cognition were primarily analysed here in this chapter, while valid cognitions are explored later. Non valid cognitions are those mental factors which are factually disconcordant one. They are defined as cognition which is not non deceptive[29]. Casual relations and conditions that give rise to specific mental states are not investigated in this chapter.

Corresponding to the other parts of the work, major portion of the chapter is taken by refuting the flawed interpretation of the early Tibetan scholars. Sapan painstakingly criticized how far they are from the true intention of Dharmakirti and then goes on to refute them. Earlier Tibetan gives five typologies of non valid cognitions, namely, correctly assuming consciousness (yid dpyod),inattentive cognition (snang la ma nges pa), subsequent cognition (bcad shes), doubt (the tshom) and distorted cognition (log shes).According to Gorampa early Tibetan scholar arrived at the division of non valid cognition into five types as follows.[30] When consciousness engages its objects its apprehension might waver or not. The first one is a case of doubt. Those that apprehends its objects unwaveringly may do so not in accordance with the modes of existence of its objects or in accordance with it. The first one is distorted cognition. Those that apprehend its objects in accordance with its mode of existence might do so as not contradicting the mode of apprehension of its opposite or contradicting its opposites. The first is inattentive cognition. In the second case a cognition may be engaging with an object that is apprehended earlier or a new one. If it is the first then the cognition is subsequent cognition. In the later case it may engage with its objects based not on reason or not. The first is correctly assuming consciousness. In the second case a consciousness engage with its object by the power of experience or reasoning. First is perceptual valid cognition and second is inferential valid cognition.

Sapan does not have problem with the last two categories of non valid cognations. He focused his critique on the first three category. The first category, according to him does not even exists, it is pure fabrication, another diversion from Indian precursors. A correct assumption according to the early Tibetan scholars is mental events that holds its object without evidence or grounding in reality, which alone make a consciousness valid cognition. Minutely analysing, Sapan show the non existence of the consciousness which is endowed with the attributes and features described by his adversaries. For him it is a phoney classification and another diversion from master’s intention. Sapan ask does correct assumption depends on reasoning or not? If it does what is the status of reason, is it endow with three mode (tshul gsum pa can gyi rtags) or not? If reason on which correction assumption depends is valid reasoning (rtags yang dag), then the cognitions must be inferential valid cognitions. If it does not depend on reasoning then there is no need to devise a separate category because it can be subsume under either three of typologies of cognitions that Sapan thinks is implicit in master’s works. They are non realising, distorted and doubt (ma rtogs pa log shes dhang the sthom) . The first one arise dependent upon non established reason, second through contradictory and the third from uncertain reason.

Inattentive cognition as a distinct/separate category of non valid cognition is pseudo classification for Sapan. Inattentive cognitions are those that does not determine their object. Every instance of inattentive cognition for earlier Tibetan necessarily is invalid, while for Sapan it is not. Early Tibetan contend that valid cognition has inherently cognitive content in the sense that they determine their object[31], while Sapan and by extension Dharmakirti sees perception as mere meeting with the bare reality without any determination whatsoever, which nonetheless is valid cognition. There are times when the given consciousness is both inattentive and invalid like erroneous perception (rtog med log shes) resulting from faulty sense organ, but they are not universally true. For him all direct perception is example of mental factor that is common locus of both valid cognition which nevertheless is uncertain. Chaba’s theory of an inattentive cognition as a separate kind seems to be result of his understanding that valid perceptions determine its object (rang yul nges pa), which probably is also related with his admission of real universals and importance he credit to common sense in Pramana tradition. For Sapan, no perceptions determine their objects to begin with. It just holds its object in perceptual ken. Sapan thinks that postulation of epistemological theory that hold some case of perception determing its objects has ramifications unacceptable to basic Buddhist logico-epistemological traditions initiated by Dignaga and refined by Dharmakiti; namely ontology that supports real universals.

He also does not agree with the account of subsequent cognition given by early Tibetans, which for him is necessarily conceptual in nature while for early Tibetans there are instances of non conceptual subsequent cognition (bcad shes dhang mngon sum gdzhi mthun). It just belongs to category of non-realizing cognitions, not as a seperate class by itself. Subsequent cognition exists on the same plane as non apprehending cognition, difference being only of degree and not of kind. Sapan was especially critical of the idea that second moment of perception as subsequent cognition. For him subsequent cognitions is necessarily conceptual and always invalid, while perception is always non-conceptual.

In the last part of the chapter, a working definition for invalid cognition is given[32] and individual families within this general category is defined. For him there are only three different class of non-valid cognitions; they are non apprehending cognition being the chief, which is strikingly absent in the list of his opponents. Doubt and distorted are the next two. For Sapan doubt is also form of indecisive awareness on top of hesitancy. This chapter is mainly based on the chapter on inference for self (rang don le’u) of Compendium and Commentary, part of chapter on perception where non valid perception are dealt with and, on the part of Drops of reasoning (gtan tshig thigs pa).

Next few chapters deals with the ways in which consciousness apprehends its objects. General and particulars (samanya and visesa, spyi dhang bye brag), appearance and elimination (pratibha and apoha, snangba dhang sel ba), signified and signifier (vacya and vacaka, brjod bya dhang rjod byed), relations and exclusive (sambhandha and virodha, ‘brel ba dhang ‘gal ba) are discussed vis-à-vis Dharmakiriti, his non Buddhist opponents and early Tibetan scholars.

All of these dual topics are based on the binary system of specifically characterized phenomena and generally characterized phenomena on the ontological side and corresponding direct perception and inferential knowledge on the side of epistemology, and the positive engagement and eliminative engagement by subjects when apprehending their corresponding objects. These three sets of dyad system is the string that runs through all other strands of Dharmakirti’s system. Hence without the grasp of this background knowledge, rest of the chapters does not make much sense. In all those chapters, he critically examines the views of not only of his usual Tibetan adversaries but also of general Indian orthodox Vedic philosophical schools stances on the subject. His sharp critical investigation subjects them to relinquish their position. He then proceeds to give his own theory and treatment on the subjects, and laboriously quotes directly from master’s works to support his own stance. This is a very significant and powerful strategies adopted by Tibetan, since they practice philosophy as a way of writing down commentary on authoritative Indian texts, and not rely only on the philosophical merit of the topic discussed. He also refutes any possible doctrinal attack, both real and imagined that his tenets might stimulate in other.

The issue of universals and particulars occupies the central position in the epistemological system of Buddhist thinker like Dignaga and Dharmakirti. In order to understand these, the basic Buddhsit doctrine of impermanence has to be appreciated. Their radical rejection of real universals seems to be the results of Buddhist doctrine of universals flux which denies any stability over duration and Buddhist critique of wholes. For this thinker, universals are quasi phenomena that are conceptually construed based upon our use of language and superimposed on basic elements that Abhidharma lists. It is the creation of conventions like conceptual thought process and language. The Hindu realist, specially Nyayayika (rigs pa chen pa) on the other hand believe and posit that these phenomena have real existential status. They are part and parcel of the real world, fabric of our world and not just mere product of conventions. They not only really exist but exist independent of their particulars. They are what validates our cconventions like language and conceptual thought and provide norms; the structure of those conventions reflect truly and non-distortedly the structures of world. All of these categories are essential for us to explain the relationship between our thought and world outside.

In Dharmakitian’s system perception is hampered by lack of certainty, while conceptual is hampered by lack of specificity. Our perception being free from conceptual (rtog pa dhang bral ba) and abstraction process (mngon sum gyi dgag sgrub mi byed), relate directly with the ultimately established individual substance, that has unmixed spatial-temporal dimension and nature. Its objects are directionally part less or temporally part less object. In other words, our perception relate to real phenomena whose spatial-temporal dimensions and nature is not mixed with those of other, a condition for being objectively established phenomena. When this real object appears to perception, all the features and characteristics of that object which are one undifferentiated entity with it also appear. This is the reason why non-conceptual cognitions are referred to as ‘complete engager.’(sgrub ‘jug)

On the other hand object which appear to conceptual thought, that subject is prevented from clearly perceiving the unmixed features of its objects, because the nature of the object appears to thought as mixed with its universals. For example, when I think of my village, the intentional object or conceived object (zhen yul) is small country side hamlet occupying a real geographical position, while the appearing object (snang yul) is its universals. This universals is constructed by conceptual thought based on its objective aspects (grahyakara,bzung rnam), which then it take to be real village, inherently conditioned as it is by primordial ignorance. All cases of conceptual thought involves this mistaking its objective aspects for real world. And this mistaken tendency is inherent and internal one.

The problems of universals and particulars are taken up in the third chapter of the texts. The problem related with the ontological status of universal is one very tenacious and for which Dharmakirti paid very large attention in his writings. Universal is concept which assist us to recognize and classified diverse particulars. Linguistically it is the reference of our common noun. For example, when we see piles of particulars book, we immediately and intuitively thinks that all particulars books are subsume under universals book-ness. That is all particular book instantiate universal book-ness. Or we think that all particulars books are book. The concepts of universals is also invariably invoke with our ability to build systematic and meaningful universe out of disoriented perceptions. Another important issue related with the problem of universals is question of the validity of our conceptual reasoning/inferential knowledge and naming process/linguistic conventions. Both of this conventions involve the use of universals; both are directly link with it. Universals are the appearing object of the inferential knowledge while it is the directly signified object of the names.

Since for Dharmakirtian tradition[33] in which Sapans works is to be contextualise, universals does not satisfied the three criteria of real thing. They belong to the quasi entity of generally characterized phenomena. But this ontology is intimately related with the epistemology of inferential knowledge and languages as shown earlier, therefore how can we be able access to real particulars through language and conceptual cognitions if what they really relate does not really exist? Dharmakirti’s ontology of anti realism[34] in realm of universals and by extension that of Sapans seems to keep conceptual thought (kalpana,rtog pa) and language imprisoned within their own generated fiction. This is also the greatest predicament within the overall structure of Dharmakirtian tradition; his system revolves around the ontological theory of two object and corresponding two valid knowledge. He strives to give philosophically informed explanation about the epistemology based upon his ontology. But his rejection of universals make it difficult to explain reasoning process and naming.

Following Dharmakiti lead, for Sapan too, there exist two classical orthodox Hindu school who advocates realist theory of universals. They are associated with Nyaya (rigs pa chen pa) and Samkhya (grang chen pa). First school advocates the reality of universals which is eternal, part less, ubiquitous and independent of its particulars (rtag pa cha med khyab pa dhang rang dbang ba). It is an example of extreme form of realism. The second school postulates that universals are substantially identical with its particulars. Sapan’s refutes both theories by saying that they can be simply dismiss since we does not observe any real universals either as independent of its particulars or as substantially identical with its particulars.

Sapans own theory of universals seems to be pragmatically oriented one. His division of two contexts, that of theoretical considerations (‘chad pa’i tshe) and of practical standpoint (‘jug pa’i tshe) seems to apply to this problem also.[35] Folowing Sapan Gorampa defined universals as “comonly appearing which is superimposed”(thun mong bar snang ba’i sgro btags). And he enumerates it into three division of kind universals (rigs spyi), meaning universals (don spyi) and collection universals (tshogs spyi). Gorampa in his commentary says that not only universals is non-entity (dngos med), a whole which is composed of many parts, duration which is collection of many moments too are non-thing. Gyaltshap on the other hands in his commentary warns that if we denies any reality to duration, that will hamper our understanding of ethical law (las dhang ’bras bu).

He ends the chapter by answering objections made against his theory of universals. The opponents accuses him of not grounding our common practices and reasoning in reality which will surely undermines validity of those practices. Sapans answer seems to be merely describing the practical procedure of how we are successful without explaining the reason for the success. According to Gorampa third chapter is mainly based on the perception chapter of Compendium and Commentary and first chapter of Commentary.

The mode of engagement of non conceptual cognition like sense perceptions is by mere appearance of its objects, while conceptual thought and languages engages with its objects by elimination of others, to express technically, it engages its object by other eliminations(anyapoha,gzhen sel). With this observation Sapans begins the fourth chapter. It deals with one of the central theme of Buddhist logico-epistemological system; that of (it is very difficult concept) apoha theory. Essentially it is the anti realist answer to explain the validity of reasoning and language in the absence of real universals. The theory seems to state that our language and conceptual thought engages with its objects negatively through excluding it from class of phenomena that it does not belongs. The chapter is revealing in more than one ways about the deviations between Sapan and early Tibetan. He vehemently attacks early Tibetan tenets that our perceptual faculties determines its object (rang yul nges pa), on the ground that it goes against the intention of Master and also philosophical ground as it implies acceptance of real universal, which he invariably equate with the Hindu school.

Sapan begins the chapter by discussing about the objects of perception and how it engages the objects. The objects of perceptions are parts and fabric of the real world, which exist with their uncommon identity. These objects if they are one cannot be many nor if they are many cannot be one. Like-wise on the epistemological side of perception it engages its objects as it is reflecting the structure of its object un-distortedly. Then Sapan proposes his own theory of how perception determines its objects. Since perception itself does not affirmed or negates, it does not determine its object; it just hold it in its perceptual ken. What it does is it induces conceptual thought (nges shes ‘drang pa) which determine the objects. For example eye consciousness which hold blue colour does not determine its object. What it does is it induces subsequent cognition (bchad shes) which will determine it as ‘blue.’

Sapan proceeds to discuss the apoha theory first by defining the subjects which engages it, namely thought and language as, a mode of engagement by excluding its opposites. Gorampa closely following Sapan defines exclusion as ‘superimposition which appears as this objects through eliminations of its direct opposites’. For example when a conceptual thought which is conceiving a cup engages its objects, it does so by the process of eliminating the cup from class to which it does not belong. Or when the conceptual thought that conceives cup it does so by mistaking its objective aspects with the universals cup-ness that real cup instantiate. This direct object of conceptual cognition is exclusion.[36] Gorampa group indian account of apoha into two groups, the first one were those who postulates three exclusion and second were those who postulates two types of exclusion. Abbot Santarakshita and Shakya Buddhi were proponents of first. Three types of exclusion that they accepts are non-implicative negative exclusion (med dgag gi gzhan sel), cognitive exclusion (blo’i zhan sel) and exclusion of real specifically characterised phenomena (don rang than gyi zhan sel). The last two are nominal exclusion since they are specifically characterised phenomena which exist independent of the activities of thought and language. Exclusion being universals necessarily exists in dependence upon the constructing activities of thoughts and language. The second theory is propose by Dharmadhotara. The two exclusion are non-implicative negation exclusions and superimposed exclusion (sgro btags kyi zhan sel). Gorampa also summarised Tibetan understanding on the subject into five unifying categories. First those who content that exclusion is invariably exclusion of non-implicative negations. Second who takes that exclusion is necessarily either of non-implicative negatives or of cognition. Third were those scholars who takes that exclusion is necessarily exclusion of specifically characterized phenomena. Fourth were those who think that the exclusion of this particular phenomena must be either of the three exclusion though this does not apply generally. Last were those who accept equivalence between existence and exclusion.

In this chapter Sapan offers a penetrating analysis into the structure of abstraction and conceptual coalescence (gcig tha dhadh dhu rtog pa dhang tha dhadh gcig tu rtog pa). Generally the mode of apprehension of conceptual thought can be divided into four exclusive group which is also exhaustive. Apprehension of many as one, one as different, one as one and many as many. He discuss first and second. Both are treated by analysing each into three parts, viz., 1)reasons for reification, 2)aspects of the conceptual mind that which reifies and 3)purpose. The conceptual coalescence coalesce different objects into a single unifying category which subsume them. This process of mistaking happens because all of the particulars share either similar attributes or the performs similar functions. For example Maple and Oak though not substantially identical share a similar features of being endowed with branches and leaves. Like wise bottle water and tape water both quenches thirst, which is a case of performing similar functions. These similarities are the reason on which thought construct a unifying concepts of tree-ness in the first case and water-ness in the second case though nothing in the reality correspond to those concepts. The mode of apprehension of these coalescing subjects are taking different particulars as one as a results of beginning-less ignorance. The purpose or the function of these coalescing thoughts is that they make it possible for us to engage in affirmation and negation in the absence of real universals.

The conceptual abstraction results because though the objects of abstracting subjects are substantially identical yet because they are distinguished (log pa) by different distinguisher (ldog pa tha dhadh) are abstracted differently. For example the impermanence and compounded aspects of sound are substantially identical yet they have different distinguisher, which is taken by conceptual thoughts and are reified as different entities. The apprehending mode of those abstracting thoughts is it engages its objects as differently. This abstracting concept assists us in realizing the identical nature of different aspects of a given entity.

A systematic presentation of other exclusion, through enumeration of its various typologies is given. Exclusion (sel ba) or other exclusion (gzhan sel) can be classified into two groups as exclusion of thing and of non thing base on the types of conceived objects of conceptual thought that construct the exclusions. Gorampa further sub divides them into different groups based on the types of subjects and negation(that is what is excluded or negated).

In the last section of the chapter, Sapan succinctly summarizes classical criticism of apoha theory and its rebuttal founded in the first chapter of Commentary. The following five are the critiques marshalled by Hindu philosopher against Dignaga, who is the first to promulgate the theory. They are;1)If there are no real universal, conventions of naming process can not be successful, 2)other exclusion leads to circular arguments, 3)the theory does not have universal applicability, as it can’t be use when there is no negandum, 4)the theory cannot explain the analysis of the ontological status of universals and 5)lack of universals as implied by theory will lead to denial of lack of real common locus. Sapan closely follows the arguments developed by Dharmakirti to rebut the rebuttal. Gorampa in his commentary writes on the topic of eliminative engagers and complete engagers. The first are thoughts and languages which proceeds its objects negatively by excluding it from the class in which it does not belong. Second are non conceptual perception which does not depends on exclusion to engage with its objects.

At the outset of the chapter five Sapan refutes the linguistic theory of classical Indian grammarians (brdha sprod pa) and Hindu’s who argues that our language signify (brda’i dngos yul) and relate directly with the reality out there, implying there is a natural relationship between our terminological convention on the one hand and reality on the other. Some Hindu schools argue that what is directly signified by our language is the individuals, while some thinks it is the universals that is the directly signified by language. According to Sapan our language and naming process does not really signify fabric of reality. If we apply name to a real entity x at time t, then at time t+1 we can’t apply same conventions since then the basis of the name first applied cease to exist as it is impermanent. This then leads to destruction of primary purpose of naming. He also criticises the position of Tibetan and Abhidharma schools which held that the universals and non-composite phenomena (ldan min ‘du byed) as what is being directly signify by the names, on the ground that that won’t lead us to the successful naming process. This is so because his opponents thinks that name can be applied successfully since the two type of phenomena directly signified is “like” the particulars on which name is applied. This Sapan argues is a fallacy since we can’t be always successful if the names are applied to particulars on the basis of theirs being similar to the directly signified.

For Sapan naming and conceptual identification is mere conventions albeit useful one. Therefore it is essentially a mistaken case in which all unenlightened sentient beings participates under the influence of avidya (ma rig pa).Central to Sapans project of naming process is his division of two levels of discourses;critical or theoretical standpoint (chad paitshe) and practical standpoint (‘jug paitshe).Were we to appraise and assess naming from former level, it won’t be able to withstand examinations. It will only result in revealing great hollowness of our common practices. It is only from non critical standpoint that it makes sense. Since universals is initially involved in the naming process as the directly signified (dngos khi brjodh bya), i.e, they are direct object of names, and since it does not really exist outside the framework of conceptuality and terminology which construct it, it only makes sense within the second framework.

Sapan’s account of naming is closely following that of Dharmakirti’s. Central to the Sapan’s account is that naming is a conventions at the centre of which is mistake of taking the objective aspect of thoughts as the objects outside by thought itself. This process of inherent mistake takes place as a result of nature of thought itself. Under the influence of inherent mistakes our conceptual thoughts fails to make the distinction between universals and individuals.

He ends the chapter by refuting some qualms about the naming process developed by Dharmakirti and that he continues. The main being that in the absence of real phenomena being directly signified by the name, naming process will only lead to arbitrariness. Gorampas points to the third chapter of Commentary as the source of this ideas. Gorampa on top of commenting on the chapter complements it by extensively developing the theories of subject and predicate expression (chos dhang chos can brjodh pa’i sgra). What are the objects of these two expression? Does those objects have different ontological status on or are they merely differentiated by our use of conventions? Gormapa refutes many scholars and his own theory is that both type of expression primarily refers to universals. They both express a single entity but differently. In the same chapter he further elaborates on the ideas of kind expression and collection expression (rigs dhang tshogs pa brjodh pa’i sgra).[37]

In the seventh chapter Sapan takes the issue of relations (sambhandha,brel ba). The topic of relation is very important within overall framework of Dharmakirtian’s tradition; the efficacy of reason adduce to bring the truth of proposition that proponent wants to communicate to his/her interlocutor depends upon the relation between the reason/sign and predicate of the proposition(rtags chos khi ‘brel ba). In other words ‘relations’[38] is the modus operandi of Dharmakirtian logic. Yet Dharmakirti and his followers like Sapan denies the reality to relations. They just think that it is a necessary construct to account for our use of signs (rtags). Sapan starts the chapter by observing that all negation(dgag pa) and affirmation(sgrub pa) is rooted in exclusion and relation (dgag shrub ‘gal ‘brel la stos pa). Sapan divides the chapter into three sections. Negating real relations. Affirming relations as conceptual imposition. And valid cognition that comprehends the relation. For Sapan relation is fiction; it is generally characterized phenomena that does not have ontic bearings. Buddhist scholars in general admits two types of relations. They are relations of identity and relations of causality (tadatmyasahbandha,bdag gcig ‘brel dhang tadutpatti, de byung ‘brel). He refutes both types of relations, denying them any reality. His arguments against both types of relations seems to be the result of the basic dual ontological assumption that he accepts following Dharmakirti’s leads; that of real specifically characterized phenomena and unreal generally characterized phenomena and the identity conditions of what constitutes real. Any relations presupposes two terms, that are object of the relation (‘brel yul) and one which relates (‘drel ba po). And since in Buddhist logic-epistemological system they admit only two relations,[39] that are relation of identical nature (bdhag cig gi ‘brel ba) and relation of cause and effect (dhe ‘byung gi ‘brel ba), these two types of relations has to be established or attain between two ontological categories. But relations built in-between generally characterised can’t be real since the basis itself is not a part of fabric of reality. And the relations between things cannot also be real for the following reason. If the relations is that of one identity, this cannot have real status since relations involves the notions of at least two entity and it does violates our notion of identity and difference. The relations of cause and effect also does not stand to the analysis. When the cause exist at time t, the effect is yet to come into existence, and when effect arises at time t+1 its cause is already disintegrated. This types of arguments made from the analysis of temporal sequence between cause and effect makes sense in the Buddhist framework or at least in the Dharmakiti’s system[40] which argues for the necessarily sequential nature of cause and effect. Therefore the relations between cause and effect is also mere construct.  So then how is relations constructed? Here again our conceptuality plays the role. The conceptuality of abstraction will abstract a given entity like vase into its being bulbous body, impermanent and compounded through the process of excluding from the class to which it does not belong. Then the conceptuality of coalescence will coalesce them into a single entity, which is what we label as relation of identity. Yet here again the status of or the validity of these two thoughts have to be  analysed. Like wise relations of cause and effect is also constructed by thoughts induce by valid perception.

Gorampa in his commentary explicates the etymology, definition and typologies of relations. Whether any Tibetan scholar/s admit the real relation is not clear. The cognition which construct relations has to be at least indirectly related with the real world of particulars. It seems that if we press the line of Sapans arguments and takes it to its logical conclusion, it will result to utter lack of relation between cause and effect. If this results then the question over relation between cause and effect in general, and Karma in particular rises. How to philosophically defend the arising of effect from a particular sets of cause and conditions in a world which utterly lacks real relations seems to be thorny. At the end of the chapter he develops a complicated process through which we can ascertained relation between two entities. This chapter is mainly based on the ideas develop in the first and last chapter of Commentary and Analysis of relations (brel ba brtag pa).

Chapter sevens deals with exclusion (virodha,gal ba). This is also a very important logical notions involve in the negation (dgag pa). Again Sapan refutes the view of Early Tibetan interpreter of their view with regards to this topic. It mainly revolves around the definition of exclusion. Early Tibetan defines exclusion as ‘two different phenomena which will never share a same locus’ (tha dhadh cing gzhi mthun mi srid pa); for Sapan and his close commentators this results in establishment of exclusion between a cause and effect, since the cause of something and its effect will never share identical locus, which leads to the absurd position of proposing both relations and exclusion between a cause and its effect. If this is accepted then it will result in going against both reasoning and scriptural authority. First happens because since now there is both relations and exclusions between cause and effect we can negate and affirmed one by other at the same time. Second is because it goes against a passage in Certainty of master works.

In the same chapter he enumerates typologies of exclusion, which is of two type, viz., exclusion of not being together (lhen gcig mi gnas ‘gal) and mutual exclusion (phen tshun spang ‘gal). Again both are sub-divided in to two sets. While the former can be attain only between two things (vastu,dngos po) later can be attain only between a thing and non-thing (avastu,dngos med). The former functions so as to negates the existence of its contradictory while later function so as to negates its identity as of its contradictory. For example a exclusion between conceptual certainty on the truth of impermanent of sound will negates the existence of distorted notions which holds sound as permanent on the locus of single person. Like wise the exclusion between permanent and impermanent will exclude a basis from either one of its contradictory; for example on the locus of sound impermanent will negate its being permanent. All of this is based on the understanding of concepts of identity and difference. All substantial difference results due to the difference in casual nexus. All conceptual difference results due to the appearance of given entity to conceptual thoughts. Likewise the identity is also due to the similarity in casual nexus and appearance to conceptual thoughts. This chapter is based on the chapter of inference for oneself of Compendium and Commentary and Drops of reasoning.

From the eight till last, it deals with ascertaining the valid cognitions that apprehends it (two objects of comprehension, namely specifically characterised phenomena and generally characterised phenomena). The first part of the eighth chapter takes up the issue of definition and definiendum (mtshan nyidh dhang mtshon bya). The chapter is divided into two parts, first dealing with the issue of definition and definiendum generally and then with regards to the definition and related issues of valid cognition. The main opponents were earlier Tibetan, whose theory on those topics are subjected to criticism. Arguing both from philosophical standpoint and also from Indian tradition he refutes them. The issue revolves around the hub of whether there is a definition for definition or not? Accepting first will leads to infinite regress while second will lead to giving up that definition is definiendum. Here Sapan follows the solution provided by Chapa that the name of definition should have a definition, while it’s meaning distinguisher need not. He also criticised the three faults[41] of definition as proposed by early Tibetan commentators. Sapan himself defines definition as ‘meaning distinguisher established by exclusion of direct contradiction’.[42] Here what is excluded or cancelled is class of phenomena to which it does not belong. Meaning distinguisher prevents definiendum and illustration from being definition. Here too we can see the negative process of apoha involve.

In the second half of the chapter valid cognitions are defined. Sapan begins by refuting Indian commentators views on the two passages of definition Dharmakirti’s gives in second chapter of the commentary. In the first line of the first verse of second chapter of Commentary he defines valid cognition as ‘valid cognition is consciousness which is non deceptive.’[43] The second is given in the third line of the fifth verse of the same chapter. Here Dharmakirti defines valid cognition as ‘valid cognitions are those that reveals unknown objects.’[44] Both Devendrabuddhi and Sakyabuddhi thinks that each gives a full definition of valid cognition, but the important point is that they have different meaning. Sapan’s main issues with this interpretation is that if we assent that this two definition are different then we need to posit two different valid cognition (two different valid cognitions and two types of valid cognitions are different). Another Indian scholar interprets the second definition as that of ultimately valid cognition (dhon dam pa’i tshad ma) and the first one as defining conventionally valid cognition (tha snyad pa’i tshad ma). This Sapan’s thinks will results in the fallacy of being not defining what counts as valid cognition generally. He also takes issue with Dharmotarra. Sapan himself and Gorampa too takes the two statements to be making a complete definition of what constitutes as valid cognition. These two statements define valid cognition in different way but with same import. Gorampa in his commentaries also raises issue with Tibetan interpretation of these lines. Sapan defines valid cognition as a consciousness that is non-deceptive. A consciousness can be count as valid if and only if it is non-deceptive in relation to action, agent and object. For early Tibetan right cognition (yang dhag pa’i she’s pa) is not necessarily valid cognition. Sapan refutes this and takes them as synonym. He proceeds to explain the etymologies, divides the valid cognitions into two types and refutes misconception with regards to the illustrations of two types. The concepts of internally valid cognition and externally valid cognition(rang nges tshad ma dhang gzhen nges tshad ma) are discussed too.[45]

It is also in this chapter that Sapan deals with his own understanding of affirmation and negations, the types of negations, and how they are superimposition/projection of thoughts. Affirmation and negation, if taken as a noun, can be generally understood as the object of perceptual consciousness and conceptual thought. On the other hand if taken as verb can be understood as the function of those two cognitions. Here it is to be understood in the second sense, as a function of our cognitions. Perception by and in itself does neither affirms nor negates. It does these through inducing a subsequent conceptual cognition which directly engages into the function of negation and affirmation. Conceptaul thoughts like inferential knowledge affirm and negates its objects directly without recourse to subsequently induced cognition. Here Sapan refutes one of the salient innovation of early Tibetan, that of explicit realisation and implicit realization (dngos rtogs dhang shugs rtogs). But what does affirmation and negation means? Affirmation or negation does not mean affirming or negating an existing thing on ontological levels; it means gaining certainty in the existence or non-existence of what is being affirmed and negated on the epistemological dimension.[46] This chapter is mainly based on the second chapter of Commentary and chapter on perception of both Commentary and Certainty. There seems to be no clear account of triad concepts of definition, definiendum and illustration in the seven works of Dharmakiti.[47]

In the ninth chapter, valid perception, first cardinal valid cognition in the Buddhist philosophy is taken up. It is also the longest chapter in the text. The whole chapter is divided into three parts, first dealing with right perception(mngon sum yang dhag), facsimiles perception  (mgon sum ltar snang) and results of valid cognitions (tshad ‘bras).

Within the first part he takes upon the definition given by early Tibetan of valid perception. They differentiate between perception per se and valid perception, which Sapan’s criticised as going against the Dharmakiriti’s tradition. But his main concern seems to be with the Tibetan attribution of an elements of cognitive contents (mngon sum nges pa can) to the perception, which supports the implications of real universal on the ontological levels and constructing activities to perception as oppose to Dharmakiti’s passive perception on epistemological side.[48]His own definition of the valid perception is non-mistaken cognition or non-mistaken perception free form conception(ma ‘khrul bai’ rig pa’m rtog bral ma khrul ba’i rig pa). He goes on to argue that our perceptions are devoid of conceptuality mainly based on the arguments developed by Dharmakirti in the first part of the third book of Commentary. Conceptual thoughts are necessarily apprehending its objects through the mediation of appearance of its universal based on the negative process of apoha. It is mostly mediated by language, exception being those conception of child untrained in the linguistic convention of a community sharing the same language, and invariably mistaken in taking it’s appearing objects as it’s conceived objects (snang yul zhen yul nyid du khrul ba).  Closely following Dharmakirti’s leads for Sapan perception is of four types namely; sense perception, mental perception, self-cognizing perception and yogic perception. Each is defined, its sub division classified and enumerated and goes on to show how each perception is valid and free from conceptuality (rtog bral dhang mi bslu ba).[49] The casual nexus and dynamics of those consciousnesses are also touch in this chapter. Sense perception arises through the interaction of three conditions.[50] They are free from conceptuality since its objects appears to it clearly.[51] In his analysis of mental perception he goes on to argue that it does not have independent continuum of its own. For him it arises at the same moment with sense perception, except it does not arises in the first moment of sense perception. Self cognisant cognitions are by nature of mere experience as oppose to physicality. It is the subjectivity of our mental world as opposed to the physicality(bems po las bzlog pa tsam).

In the parts dealing with the yogic perception, long and details information are given to support the possibilities of enlightenment, a state of mind free from any traces of delusions and ignorance. A textbook definition is provided and its types enumerated. His claims of existence of such a mental perception is mainly based upon the analysis of the teaching of various teachers. Here we first rationally analyse the Four Noble truth taught by Buddha. And once we get confidence in the truth and implication of general teachings of Buddha can safely assumes that he has a special types of mental perception that arouse out of deep meditation that we call yogic perception.[52]He says that a yogic perception that realises the sixteen aspects of the Four Noble truth in its entirety is possible because the basis of such a practice is stable, that practice itself will spontaneously get better and that the subjects contemplate on it(rten brten pa bsgom pa ngang gi khyad par du ‘gyur ba de bsgom pa). In order to prove that the stability of the basis of such a practice, which is the continuum of our mind he discusses existence of life before and after this one.The changeability/alterability of our perceiving subjective mind is established in this chapter. Underlying all of these arguments is Buddhist theory of causality. There are two causes, substantial and co-operating (nyer len dhang lhen cig byed rkyen). Clay is a substantial cause of clay pot, while the potter and his skills are co-operating cause. Nothing can be substantial cause of consciousness except preceding instant of consciousness (I heard some western Buddhist teacher to express this contention of Buddhism as, ‘law of conservation of consciousness’).[53]Ignorance as a misperception of reality is touched upon briefly and was contended to be the main or root source of all dissatisfaction. Main arguments underlying the possibilities of alterability of our mind is that, if we can manipulate the nexus of cause and conditions that give rise to particular effect, it will have irreversible impact on the effect. Here too he rejects the Tibetan understanding of reasoning process that establishes the independent of continuum of consciousness from matter. One surprising theme that he discussed here is not only our consciousness gets transfer to next life as propelled by karmic imprints but also our sense organs too leaves and acts as the substantial cause sense organs of next lives too.[54] Sapan also discusses the various types of antidotes to our afflictions differentiating between the temporary one and the permanent one, former merely suppressing our afflictions while later extirpating the afflictions. The later is discriminating awareness that realises the Four Noble truths as it is, pointing to the importance of truth dimension in the soteriological project. He rejects the skepticism of Indian thinkers pertaining to the possibility of mind free from afflictions and omniscient. His central idea is the theory of nature of mind as naturally luminous (sems kyi rang bzhin ‘od gsal ba)[55].

Then he proceeds to discuss the facsimiles perception. These part is important as it points to a different line of interpretation of Dignaga’s theory from that of Dharmakirti. The last part of the chapter discusses the results of valid cognition. Though Sapan’s briefly touches upon it, Gorampa in his commentaries develops it in great details based upon the third chapter of Commentary. These parts is important since it studies the representationalism of Sutrantika and Cittamatra school. Gormapa also analysed mind only school theories and the notion of emptiness as empty of subject-object dichotomy or duality. This chapter is mainly based on the chapter on perception of Compendium, Commentary and Certainty. 

In the tenth chapter, inference for self (rang dhon rjes dpag),and logical principle underlying the valid inferential cognition is taken up. A correct reasoning is defined as one with three features/complete with three modes.[56] Dharma of subject (establishment of the logical mark on the subject of the syllogism/minor term), pervasion and counter pervasion are three modes or the feature of the correct logical mark[(phyogs chos rjes khyab ldog khyab) (paksadharmata,anvayavyapti and vyatirekavyapti)]. The main theme of the chapter is to differnatite between correct or valid reasoning from false or spurious reasoning and the nature of propositions (dham bca’). When a reasoning that has these three feature is employed to support a proposition(pratijna,dam bca), it can lead to valid inferential knowledge(anumana,rjes dpag). The relationship between the subject of syllogism, predicate and logical marks are cogently discussed. The mode of establishments of the reasoning on the subject of proposition is quite straight forward. The nature of pervasion or the entailments of predicate by the reason or sign is more complicated in the absence of real properties or universals since individuals or real phenomena does not relate to each other. For example when the subject sound is establishes as impermanent by reason compounded, the first mode that is presence of being compounded on the subjects sound is clear. But what is the nature of reason and its relation or invariable concomitance (med na mi ‘byung gi ‘brl ba) with predicate? Whose compounded is the sign? Is it of subject? or of another phenomena then subject? Sapan and by extension Dharmakirti contend that it is general properties of being compounded, which does not really exist or it exists as a mere product of activities of our conceptuality and language. But then the question arises of how an unreal reason will leads to inferential knowledge which they contends capture real world? This seems to be the one of the central theme of Dharmakiti’s tradition, that is to support and give convincing account of inferential knowledge without jeopardising the basic Buddhist anti substance philosophy.

Next he goes on to the classification of valid reason into different kinds, sub groups are also given in details. Normally there are three kind of correct reasoning that leads to valid inference based on the types of invariable concomitance that exist between logical mark (linga,rtags) and the predicate of syllogism (sadhyadharma,sgrub byai chos), which are relationship of identity and of causality. The three types of correct reasoning are namely, correct reasoning of same nature (svabhavahetu,rnag bzhin gyi rtags), correct reasoning of effect (karyahetu,bras rtags) and correct reasoning of non- cognition (anupalbdhihetu,ma dmigspai rtags). The first two are reasoning of affirmation while the third is of negation. Sapan here too criticised Tibetan enumeration and of accounts of valid reasoning.

Any reason that does not fulfill three formal criteria is necessarily faulty one in as much as it does not lead to the correct knowledge. It will never be the basis of ascertainment. There exist three types of faulty reasoning, non established reason (ma grub pairtags), uncertain reason (ma nges pai’ rtags) and contradictory reason (‘gal bai’ rtags). While the first reason does not have first criteria, last two lack second and third criteria. All reason can be subsume under these two types. Any reason when forwarded can be check whether it is presence on the subject or not. If it is not then the reason is not established one. Once that is establish can be of three types; one that can establish the propositions, one that negates it and one that is uncertain with respects to it. First one is valid, second is contradictory and last uncertain reason.

The last parts of the chapter deals with the proposition, that is the combination of subject and predicate that any reason is to establish. Based on the Indian sources author gives a details account of different types of propositions.  A detailed exposition of the wrong propositions is discussed through the technical notion of bsal ba, by illustrating it through the Hindus philosophical tenets. An important aspect of this part of chapter is the discussion on validity of scriptural authority and conventional nature of our naming process. In this chapter also Sapans criticize his Tibetan adversaries of how their system of inference is astray away from Indian source. This is mainly based on the chapter of inference for oneself of Compenduim, Commentary and Certainty.

Inference for other (gzhan dhon rjes dpag) is the last chapter of the book. Here parties involve, these are defender (snga rgol), questioner (phyi rgol) and witness (dpang po), during formal debate are defined. The rules discriminating judiciously the winner and loser of any formal debate are determined. A valid mechanism for detecting strength and weakness in the participant are elicited. Major portion of the chapter is taken up by refuting the Hindu school account of punishment (tshar gcod) etc.[57]

Sapan also discusses the proof statements and refutation (sgrub ngag dhang sun ‘byin). He criticises the five limb proof statements of Nyayayika and also two limb practised by early Tibetan. The main mistakes of five limb is that it fails to state general concomitance between sign and the predicate (spyi khyab). Then he proceeds to discuss ad absurdum and own continuum (thal gyur dhang rang rgyud) and again critiques the certainty of numbers (grang nges) held by early Tibetan. One’s own continuum is used when the terms and principle of syllogism is co-accepted by both the participating parties. Consequence is a method employed by parties to expose limitation and contradiction in the opponent’s theory and logic etc. It is showing inner contradictions and hurling unacceptable consequences based solely on the principle and premises accepted by the opponents. He also writes in the same chapter about what types of ad absurdum implies owns continuum and of what sorts. At the end of the chapter he once again says how the early Tibetan scholars view of types of answer admitted during debates leads to self contradictory position. There is a brief discussion about the number and ways of responding during formal debates. They are of four types, contradictories, uncertain, unestablished and acceptance(gal ba, ma nges pa, ma grub pa dhang ‘dod.). Tibetan before Sapan enumerated only three ways.The chapter is mainly based on the Science of debating(rtsod pa’i rigs pa) and the Inference for other chapters of Compendiumm, Commentary and Certainty.

He exhorts ‘us’ to follow truth and shun away dogmatism. The colophon says that he composed his masterful texts after mastering treatises of two masters. It also says that the texts was composed at Sakya monastery.

In these eleven chapter he unpacked the whole edifice of Dharmakirti’s system. It is also very important source of information on the Tibetan development, innovation and incorporation of Indian pramana material. The text has inspired a large number of scholastics writing. Later Tibetan scholiast had developed a highly complicated system of commentary around the text. It is also probably few indigenous Tibetan text that have received attention and that has been commented upon by scholar of almost all denominations of Tibetan Buddhism.[58] It arguably is the most important commentary on Dignaga and Dharmakirtian tradition. It still remains as the best gateways to enter the Dharmakirtis mind, the vast and profound.




  • Treasury of valid cognition root texts and its auto commentary, Sakya Pandita Kunga Gyeltshen.
  • Clarification of seven texts, Gorampa Sonam Sengye.
  • Explanation of treasury, Gorampa Sonam Sengye.
  • Rays of Samantrabhadra, Gorampa.
  • Sun of Samantrabhadra, Gorampa.
  • Defeater, Panchen Sakya Chogdhen.
  • Commentary on treasury, Gyeltshab Dharma Rinchen.
  • Great Tibetan Chinese dictionary.
  • Blue annals by Go Lotsawa Zhonu Pal.
  • Treasury of knowledge, Jamgon Kongtrul.
  • Genral exposition of Middle way, Gorampa.
  • Among the Tibetan texts.
  • Recognising reality, Georges B J Dreyfus.
  • Foundations of Dharmakitis philosophy, John Dunne.

[1]In this short essay I will try to introduce Sakya Pandita’s interpretation of Dharmakirti’s thought  through his masterpiece ‘Treasury of valid cognition’ and the conflicts among his Tibetan successors and commentators.

[2] This article is mainly based on the notes that I took during my two years study of Dharmakirti’s tradition, in 2012 and 2013, at Dongsar Shedra. Thanks are due to Khenpo Pasang, Khenpo Choying Dorji, Lopon Sonam Dakpa and Geshe Lobsang Sakya, who introduce and instructed me in this intellectual lineage.  I also need to thank my class mates who help me to refine the understanding by engaging me in the monastic debate.

[3]  Tibetan monks students learn about the various tenets of different ancient Indian philosophical schools in a genre of literature known as ‘grub mtha’. Most of the Tibetan understanding of the tenets of non-buddhist schools  are base on the works like ‘compendium of reality’ (Tattvasamgraha, she kho na nyid bsdus pa) by Shantarakshita and Bhavaevikas ‘Blazing or reason’ (Tarkajvala, rtog ge ‘bar ba).

[4] Like the life of most of the Buddhist scholars of first millennium little is known about the personal biography of these two scholars. Though there are some Tibetan traditional accounts of their life, like the one by Bu ston rinpoche and Ta ra na tha, these accounts seems to be much embellished.

[5] Here and throughout the essay I am using terms borrowed from western philosophical vocabulary to express classical Tibetan and by extension Indian Buddhist abstract ideas, the terms which I have very naive understanding about. Therefore whenever possible I will give Tibetan equivalent in parenthesis and the primary meanings of the particular terms should be the one of Tibetan.

[6] This can be seen if we examined the order of the chapter of Dharmakirtis greatest work ‘Commenary on valid cognition,’(Pramanavartika, tshad ma rnam ‘grel). in the first chapter he sets out to formulate logical system and its underlying ontology that can be universally acceptable and then proceeds to substantiate the `buddhist principles of four noble truths in the second chapter, cornerstone of Buddhist religion and soteriology through the use of logic established.

[7] For a brief and succinct account of the of contributions made by these two scholars and history of pramana system can be found in the collected works of Pan Chen Sh’a kya mChog lDen (1428-1507) volume number 16 and 19.

[8] Both the treasury and three vows of Sapan are included among the eighteen texts of great renown(grags chen bco brgyad) which forms the content of curriculum of Saskyapa seminaries.

[9] In his Blue Annals ‘Gos Lo Tsa Wa mentioned that before Ngok, rMa taught Pramana is his tradition is known as Tshad Ma raying Ma. But his tradition seems to have dried up long ago.

[10] Sakya Shribhadra (1127-1225) and his disciples. According to the official biography of Sapan he received his monastic vow from this master.

[11] Traditionally seven great treatises and the auto commentary to the first chapter of Commentary were attributed to Dharmakirti.

[12] Pramana is a Sanskrit terms which has two meanings: means of correct cognition or correct cognitions itself. Tibetan equivalent is Tshad Ma. Dharmakirti himself defines pramana as”pramana is the cognition that is non deceptive,”(tshad ma bslu med can she’s pa P.V 2.1).

[13] Many scholars think that non-Buddhist tenets were inadvertently upheld by Ngok due to his being influenced by his teacher, Kashmiri scholar bsKal lDan rGyal  Po who is said to have been trained in the non-Buddhist schools of thought prior to his conversion into Buddhism.

[14] In the introductory part of his auto commentary Sapan makes this remarks,” On the right cognition, even those who claims to be correct expositors are seen to be erring therefore in order to refute these and to establish right cognition this is composed.’’

[15] ‘’yul gyi mtshan nyid blos rig bya,’’ RT 1.1.

[16] Yul, Yod Pa, gdzhi sgrub abd Shes bya are synonyms.

[17] This can be seen through their use of existence as reasoning to prove impermanence of sound. A valid reasoning should entail the predicate. And here the reason, existence necessarylly entails impermanence. But both of the author add that universals etc., are not impermanence.

[18] These are objective condition (dMigs rkyen), preceding condition (de ma thag rkyen) and dominant condition (bdag rkyen).

[19] An eye consciousness apprehending white conch shell as yellow is mistaken non conceptual consciousness.

[20] Sapan and his close followers like Gorampa reads Dharmakirti’s thought as expression of two Buddhist philosophical schools, Sautrantika and Cittamatra, in the doxographical hierarchy.

[21] Dharmakirti classifies perception into four types; sense perception, mental perception, self-cognition perception and Yogic perception.

[22] “gzhal by a rang mtshan gcig kho na.”TR1.11

[23] Gyaltshap Dharma Rinchen in his commentary, Essence of good saying, on Treasury qualifies Sapans root verse with ‘ultimately’.Similarly Gorampa too does not read the root texts and auto commentary literally. In his Clarification of seven treatises, he goes on to rebut those who reads Sapan literally through scriptural citations and reasoning, in the second case he hurls three undesired consequences.

[24] Chandrakirti on top of perception and inference adds another two to make it four. They are analogy and scriptural authority. Gorampa too accepts four valid cognitions in his interpretation of Prasangika madhyamika.

[25] Tibetan has two open canon, one purported to be direct teaching of historical Buddha and one the collection of later writings of Indian scholar. The first one is bk’ ‘gyur and later is bstan gyur. Although two canons were collected only after Sapans death.

[26] Gormapa attributes a similar hermeneutical principles to Shantarakshita in his General exposition of middle way philosophy and applaud it.

[27] Tibetan scholars does not agree about the doxographical schools to which the works of Dharmakirtis belongs to. Some thinks they belong to Madhymika while others think that they expound Mind only school. This also points towards existence of different views in the works of Dharmakirti and limitations of trying to subsume all tenets of buddhist philosophy into four school.

[28] “Blo yi mtshan nyid rig pa yin/The definition of consciousness is cognising.”TR2.1.

[29] “Mi slu ma grub tshad ma min.”TR2.18.

[30] Clarification, 70.

[31] This notion is expressed as all valid cognitions invariably eliminates superimpositions (Tsahd ma yin na sgro ‘dogs gcod pas khyab)

[32] “Non valid cognition is one on which non deceptiveness is not established.” TR2.18.

[33] In almost all of the four chapter of Commentary Dharmakirti faces the problems of universals. He tries to maintain the validity of reasoning without recourse to real universals.

[34] Anti realism in the sense of rejection or real universals, relegating them as a product of superimposition by our conceptual thoughts.

[35] Sapan makes this distinction in the fifth chapter verse number 14. There he says “when theoretically considering it is wise to make the distinction. When engaging it is achieve through mistaken.”

[36] These accounts are mainly based on the Gorampas Clarification.

[37] The accounts of these two expression in his Clarification and Explanation differs a bit.

[38] This relations attained between the correct reasoning and the predicate of propositions which it necessarily entails is known as ‘invariable concomitance’ (med na med ‘byung gi ‘brel ba).

[39] Sapan in his auto commentary gives few others types of relations admitted by Hindu phiosophers which can’t be include into two aforementioned one. They are relation of inherence (ldan pa’i ‘brel ba), relation of contact (‘phrod pa ‘dhu b’ai ‘brel ba), relation of assembly (‘dhu b’i ‘brel ba), relations of attributes (khyad pa byed chos kyi ‘brel ba), relation of connections (sbyor ‘brel) and relations of meeting (phrad pa’i ‘brel ba).

[40] Abhidharma tradition admits the existence of cause and effect at the same time. This is called co-emergent cause and effect (lhen cig ‘byung ba’i rgyu ‘bas).

[41] Three faults or fallacies of definition are, fallacy of impossibility(mi sridh pa’i skyon), fallacy of pervasion that is too extensive(khyab che b’i skyon) and fallacy of non-pervasion(ma khyab pa’i skyon).

[42] “dngos ‘gal gcodh pa’i don ldog grub pa.”Auto commentary on the TR9.15.

[43] “Tshad ma bslu med can shes pa.” PV2.1.

[44] “Ma shes don gyi gsal byed kyang.”

[45] Gormapa in his commentaries discussed these two types under the rubric of seven headings.

[46] Gorampa makes this point in his Clarification.

[47] Gorampa in his shorter commentary to the treasury answers Panglotsawa by quoting from Sutralamkara(mdo sde rgyen) to refutes him, who thinks that the concepts of the definitions, definiendum and illustrations are Tibetan elaborations.

[48]  Sapan criticism is of three fold; the difference between perception per se and valid perception is unfounded, the addition of novelty as the definition is obsolete and attribution of elimination of superimposition to perception is false.

[49] The refutation of theories that posits cognitively active perception is intimately related with the refutation of rejection of real universals.

[50] Gorampa in his commentary discusses the three types of conditions of sense perception in Mind only school too.

[51] In Dharmakirtian tradition objects does not really appear to its apprehending subject since they are cause and effect respectively; cause and its effects are necessarily sequential to each other. What appears to sense perception is the representation of its objects. The objects through the dynamics of its interactions with the physical sense organs creates a network which results in the origination of perception with the representation of objects. Vaibhaisika school does not accept this theory. They argue that what appear to perception is outside object, reflecting the common sense view of the tradition.

[52] Dharmakirti in the second chapter of his Commentary develops a systematic and rationally based soteriological principle through the same line of reflections.

[53]A lengthy argument is developed by Dharmakirti to prove the impossibilities of physical matter to act as the substantial cause of non-material conscious principle, an entity of qualitatively different constitution in the second chapter of Commentary.

[54] The sense organ seems to be different from those discovered by modern biology since they belong to  completely hidden phenomena in Buddhist scheme of classification of all phenomena.

[55] Dharmakirti himself use the term in the second chapter of Commentary though how of this concepts relates or resembles those developed in Tathagarbha tradition is unclear. Sapan too does not discuss it in depth.

[56] The definition given by early Tibetan is criticised by Gorampa in his commentary. For them valid reason is one which is three modes while for Sapan and Gorampa it is one endow with three modes, not three modes.

[57]  Recently a very detailed explanation of these parts of the texts has been written by Khenpo Sonam Gyatsho based on the sanskrit extant of Hindu philosophical texts.

[58] There are at least fourteen commentaries extant on this text. Except from Kagyupa scholar there exist commentaries from both Nyingmapa master and Gelukpa master, by Mipham Rinpoche and Gyeltshab je respectively.

Please Login to post a comment