The 8th International Conference Buddhism & Australia
will be held on 7-9 February, 2019 in Perth, Western Australia.

Chinese Buddhist Encyclopedia Illustrations
Some of the Buddhist Illustrations created by Chinese Buddhist Encyclopedia
FREE for everyone to use

We would also appreciate your feedback on Chinese Buddhist Encyclopedia. Please write feedback here
Here you can read media articles about the Chinese Buddhist Encyclopedia which have been published all over the world.

Articles by alphabetic order
 Ā Ī Ñ Ś Ū Ö Ō
1 2 3 4 5 6 7 8 9 0

Defining and Redefining svalakṣaṇa: Dharmakirti's Concept and its Tibetan Modification

From Chinese Buddhist Encyclopedia
Jump to: navigation, search
Please consider making little donation to help us expand the encyclopedia    Donate Paypal-logo.jpg    Enjoy your readings here and have a wonderful day  

IMG 3386.JPG
Bu ads.jpg

Defining and Redefining svalakṣaṇa:
Dharmakirti's Concept and its Tibetan Modification
Chizuko Yoshimizu

Recent contributions to the Tibetan development of Buddhist philosophical systems have attracted considerable attention, not only because of their significant results but also because of their methodological consciousness that any intellectual tradition must be examined in light of its historical and cultural circumstances.

Continuity and discontinuity of thought as well as the characteristics of Tibetan interpretations firstbecome clear through a thorough investigation of both Indian and Tibetan traditions, and yet the significance of individual thought is finally to be considered in its contemporary context.

In this respect, the latest studies of the Tibetan development of Dharmakirti's (7c.) epistemology were most successful in indicating the consistency and inconsistency of Tibetan interpretations with Dharmakirti's original ideas.[1]

Special attention has been paid to the originality of dGe lugs pa thinkers. They indeed made several theoretical modifications to, reinterpretations and reevaluations of Indian original thought, especially with regard to logicoepistemological issues in both major fields of Buddhist philosophy, viz.,

the Madhyamaka system and that of Dharmakirti.[2] In order to gain a clear picture of the dGe lugs pa position on these Buddhist philosophical systems, we have attempted to reveal what might actually underlie their problematic commitments to traditional teachings, i.e.,to reveal its historical background, probable textual sources, possible misinterpretations and wrong transmissions of text,as well as particular aims and motivations they may have had in mind.

The present paper too is an attempt to clarify the way dGelugspa scholars redefined the concept svalakṣaṇa (rang mtshan) and to specify the reasons for this redefinition on the basis of the writings by the three main figures from the earlier period of the school, i.e.,Tsongkhapa Bio bzang grags pa (1357-1419),

rGyal tshab Dar ma rin chen (1364-1432) and mKhas grub dGe legs dpal bzang po (1385-1438).[3] I also wish to consider the question of how Dharmakīrti and these Tibetan thinkers understood the meaning of the individuality and reality of existents under the concept of svalakṣaṇa, since defining svalakṣaṇa is none other than defining what an individual and real entity is.

Through the following discussion,one willsee that both Dharmakirti and the dGe lugs pa thinkers define svalakṣaṇa not in isolation from, but in complete accordance with, their respective considerations of relating philosophical issues. As for the dGe lugs pa, however, it can be said that they aimed to comprehend such fundamental concepts as svalakṣaṇa from a wider perspective, namely they tried to formulate a version of individuality and reality which holds true not only for the Sautrantika tradition of Dharmakirti but for Buddhist philosophical systems in general, including Madhyamaka. I would like to focus on this point in the last part of the paper.

1. Dharmakīrti on svalakṣaṇa

Littie needs to be said about the considerable significance of the term svalaksa?ia which literally means 'own characteristic', and conies down to term for 'particular'or 'individual'.svalakṣaṇa is characterized by Dignaga (6c.) as the object of direct perception (pratyaksa), i.e.,the object of a cognition which is free of conceptual construction (kalpanāpoḍha)[4].A Dharmakirti added to this epistemological notion a clear ontological ground by identifying it with that which has causal efficacy (arthakriyāsāmarthya, arthakriyāśakti. don byed nus pa),

that is, an ability to produce an effect. He explicitly defines this alone (eva) as ultimately existent (paramārthasat) or as a real entity (vastu), in contrast to the 'universal' or 'common characteristic' (sāmānya or sāmānyalakṣaṇa).

The latter, in contrast to svalakṣaṇa, refers to the object of conception or of words that lacks causal efficacy and hence is considered to be merely conventional and unreal. [5]

We may be able to give the broad outlines of the development of the idea from Dignaga to Dharmakīrti, or from the epistemological to the ontological characterization of svalakṣaṇa, as follows: The fact that a thing is actually perceived by someone, sometime and somewhere indicates that this thing exists at that moment at that place, unless this perception is proven to be false by someone else. Since this thing causes a direct perception of its own image, it is admitted to be causally efficacious.

Furthermore, this thing must be allowed to be real, for unreal things such as a horn of a rabbit or. an abstract concept like 'eternity' cannot cause any direct perception.

In other words, the arising of a direct perception should properly presuppose the presence of something real as its object. Hence the object of direct perception proves to be existent in reality. In this way, a svalakṣaṇa to be cognized by a direct perception can be identified as a real entity.

The term svalakṣaṇa, as opposed to sāmānyalakṣaṇa or common characteristic, entails from the beginning that the phenomenon is individual, unique and distinct. Dignaga's description of svalakṣaṇa as the object of direct perception may well reflect the idea that svalakṣaṇa is a substantially individual thing, since it is the function of perception to make substantial distinctions among its objects.

To this extent, one could also say that svalakṣaṇa is a spado temporally individual and unique occurrence, which necessarily occupies a certain location in space and time, in contrast to a merely imagined object. The more strict spatiotemporal qualification of svalakṣaṇa can be derived from Dharmakīrti's definition of a real existent as having causal efficacy, if this qualification is linked to the theory of momentariness (ksanikatva), viz., that whatever is existent in reality is exclusively momentary.


It is theoretically consistent to interpret svalakṣaṇa as a unique and single phenomenon that occurs and disappears every single moment, since svalakṣaṇa is a real existent to be defined as that which has causal efficacy, although such a momentary thing is far beyond the range of perceptual object.

Besides svalakṣaṇa being distinguished.the one from the other in virtue of their distinct substances, we can also understand from the literal sense of the word that svalakṣaṇas are known to be unique because of their characteristics (laksana). Although it is beyond the function of perception to specify the features of an entity as, for instance, being a pot, being gold, being round, and so on, these kinds of unique features of one svalakṣaṇa can be perceived through its image as a whole and help to differentiate this svalakṣaṇa from other svalakṣaṇas.

Theoretically speaking, such a distinction of svalakṣaṇa by its nature too is grounded in its causal efficacy, because, according to Dharmakīrti, the difference of nature consists in the difference of causal efficacy[7] in the following manner: a svalakṣaṇa is known as individual and unique by its essential nature (svabhāva), since the essential nature of a real entity is determined by a particular ability of its cause to produce this entity, and this entity in turn arises being endowed with a particular ability that is its essential nature,[8]

Thus considered, it may be proper to say that Dharmakirti's identification of svalakṣaṇa as that which has causal efficacy provides a clear theoretical ground for both the reality and the individuality of the entity that is defined by Dignaga as the object of direct perception.

2. The dGe lugs pa on rang mtshan

The dGe lugs pa thinkers formulated the definition of rang mtshan according to their own interpretation of individuals and real entities. Let us look at the following definition, which mKhas grub proposes for rang mtslian, stillclaiming it to reflect a Sautrantika position:

"In their own system [of the Sautrantika), the definition of rang mtshan is the thing (dngos po) which consists (gnas), not being conceptually imposed, but from its own side [i.e., intrinsically], in its essential nature (rang bzhiri) uncommon [with other things]."[9]

Neither the object of perception nor causal efficacy is mentioned here as the definiens or the defining characteristic.

Nor is it possible to interpret the phrase 'consisting in its essential nature ' as implying 'consisting in its own causal efficacy' and the phrase 'not being conceptually imposed' as implying 'being directly perceived', once one takes account of the views peculiar to the dGe lugs pa with regard to rang mtshan and spyi.

One should firstrecall the dGe lugs pa position that rang mtshan is identical with a real existent which has causal efficacy (don byed nus pa), but not only rang mtshan is counted as real, nor is it determined for the object of direct perception alone, for they maintain that there exist real universals (sāmānya, spyi),[10] and that a rang mtshan appears in a conceptual cognition.

Even it is not contradictory that one and the same thing is rang mtshan as well as universal (spyi) in its different aspects.

They are not opposing notions but are relative.A pot, for instance, is a particular (rang mtshan) in relation to its property of being impermanent (anitya, mi rtagpa), but at the same time it is a universal as well in relation to its individuations, since the property of being a pot is common to all kinds of pots such as golden pots, silver pots, copper pots, and the like.[11]

Under this condition, the dichotomy between rang mtshan and spyi according to whether it is real or unreal, or whether it is cognized by direct perception or conceptual cognition is on no account conducive to clarifying the dGe lugs pa idea of individuality arid reality.

Nor can causal efficacy define the reality of rang mtshan. Although the dGe lugs pas accept the concepts 'that which has causal efficacy', 'that which is ultimately existent' and 'rang mtshan as synonyms in accordance with the statement of Pramdnavdrtiika III 3, they explicitly note that neither causal efficacy nor rang mtshan is taught by Dharmakirti as a definiens or a defining characteristic of ultimate reality, but just as an instance of those which are to be defined as such (mtshan gzhi).


That is to say, whatever is 'that which has causal efficacy' or 'rang mtshan' is a real entity, but it is not just this alone that is ultimately real, since there are universals that exist in reality.

Yet the dGe lugs pas maintain that the individuality of rang mtshan in the sense of 'consisting in its essential nature' is grounded in reality, as suggested in the aforecited mKhas grub's definition of rang mtshan, for the notion of 'not being conceptually imposed but from its own side' is adopted as the defining characteristic of ultimate reality (don dam bden pa, parmdrtkasatya) by Tsong kha pa:

"The definition of ultimate realityis that which is not merely conceptually imposed {rtog pas btags pa), but established from the side of the object itself (yul ranggi ngos nas)."[13]

A rang mtshan is a real existent insofar as it meets this condition. In the same way, the rang mtshan is established as an individual insofar as it is intrinsically abiding in its essential nature. The essential nature is, however, not necessarily confined to causal efficacy, since mKhas grub propounds the aforecited definition of rang mtshan, after having denied causal efficacy together with the spaciotemporal uniqueness as being the.defining characteristicsof rang mtshan, by saying:

"Such definitions of rang mtshan on which others insist as that which exists,without sharing (ma 'drespar) place, time and essential nature (yul dus rang bzhin) [with other things] and that which is causally efficacious are unacceptable."


Neither substantialindividuality nor causal efficacyis the definiens of rang mtshan either.[15] The uncommonness of essential nature is rejected here just because, in my conjecture, it lacks the qualification of being intrinsic (i.e.,rang ngos nas mthun mong ma yin pa'i rang bzhin du gnas pa) in contrast to mKhas grub's own definition, for, as will be discussed below, the non-intrinsic or conventional uncommonness of essential nature is also accepted by those who refuse the real existence of rang mtshan.

Accordingly, for the dGe lugs pas, rang mtshan is an entity that is individual and unique in reality solely because of the intrinsic abiding in its essential nature.

What then is the essential nature that determines a thing as an individual or rang mtshan? Let us consider this question with the example of 'golden pot' (gser bum), which the dGe lugs pas use for rang mtshan when explaining the theory that a rang mtshan appears to a conceptual cognition.


Since we have closely analyzed this problematic presentation in our previous studies,[17] I would just like to reconsider what it means to say that 'golden pot' is an example of rang mtshan.

First, one should note that such an example of svalakṣaṇa would not be acceptable to Dharmakīrti.

Even not appealing to the theory of momentariness, the word 'golden pot' (gser bum) cannot directly refer to any substantiallyindividual entity, which is the object of direct perception, but according to the apoha theory it solely refers to the universal. 

For the dGe lugs pa thinkers, however, 'golden pot' is an example of a particular (rang mtshan), and 'pot' is a universal (spyi). In Tibetan, this example is always simply given as 'gser bum\ i.e.,'golden pot', which is not accompanied by a demonstrative pronoun, nor by an indefinite article,nor by a suffix designating the plural.

That is, neither 'this or that golden pot' (gser bum di/ de), nor 'some golden pot' (gser bum zhig), nor 'golden pots' (gser bum rnams/ dag) is specificallyintended. Since the Tibetan language has no definite articles and only rarely use the indefinite zhig,

the expression 'gserbum signifieseither a golden pot or the golden pot in the sense of a generic singular (viz.,a golden pot or the golden pot in general), which is to be cognized as such by its properties of being a pot and being gold.

These properties are, on one hand, essential characteristics of a golden pot, whereby a golden pot is distinguished from other things such as silver pots, copper pots, glasses,tables, and so on. On the other hand, they are also common properties to all golden pots, viz.,18-carat golden pots, gold-plated golden pots, small golden pots, big golden pots, and so on.

That is to say, any individuation or differentiation among individual golden pots is not, and cannot be, indicated by the expression 'gser bum\

The fact that this example is nevertheless repeatedly applied to rang mtshan means that it completely meets the conditions of rang mtshan for the dGe lugs pa.

That is to say, a golden pot consists in the essential nature of being a pot and being gold from its own side independent of any conceptual construction. To sum up, the essential nature in perspective of the dGe lugs pa does not actually differ from common properties, which are identical with real universals to be signified by generic concepts.

Despite the fact that their understanding of rang mtshan obviously deviates from that of Dharmakīrti, the dGe lugs pa scholars seem to have formulated such an idea of essential nature on the basis of Dharmakirti's own words in Pramāṇavārttika I 40.

It is even not far from the truth to speculate that mKhas grub's definition cited previously is a modification of Pramāṇavārttika I 40. Let us compare them with each other:

Pramāṇavārttika I 40) "Since all things (sarvabhava) by nature consist in their respective essential nature (svabhāva), they are distinguished from their homogeneous and heterogeneous [things]."[18]
(mKhas grub's definition) "The definition of rang mtshan is the thing (dngos po) which consists (gnas), not being conceptually imposed, but from its own side [i.e.,intrinsically], in its essential nature (rang hzhiri)uncommon [with other things]."

The similarity in expression is evident.

Taking the subject of Pramāṇavārttika I 40, 'all things' (sarvabhava), to be identical with svalakṣaṇas in the sense of real existents,[19] the dGe lugs pa interpreters understand this verse to be intended to teach the mode of existence of real entities (dngos po'i gnas lugs).'[20]

In this regard, it seems reasonable to assume that they took this verse to describe the essential characteristic of svalakṣaṇa and adapted it to their own definition of rang mtshan.

To conclude this section, I would like to propose the following tentative illustration of Pramāṇavārttika I 40 with the example of 'golden pot' in accordance with the dGe lugs pa interpretation: "The svalakṣaṇa such as a golden pot consists in its essential nature of being a pot, being gold, being impermanent, and so on.

Therefore itis different from such homogeneous things as a silver pot as well as from such heterogeneous things as a table, space, etc."[21] So would the verse be elucidated by the dGe lugs pas.

Reasons for redefining svalakṣaṇa

From the theoretical point of view, the dGe lugs pa interpretation of svalakṣaṇa apparently goes beyond the range of sound interpretation. It is not exaggerated to regard it as a systematic revision of the Sautrantika doctrine.

This revision is, however, certainly an outcome of various external and internal factors. Such a realistic position as the dGe lugs pa thinkers have is actually considered to have originated with some Indian scholars and have been carried over by Tibetan gSang phu tradition.[22]

The lack of semantic interest may also be described as a general tendency of this Tibetan scholastic tradition.

Of course one should also clarify,in addition to this historical background, the theoreticalgrounds for the dGe lugs pas' redefinition of svalakṣaṇa. We will devote the last section of the present paper to this inquiry.

mKhas grub explains the reason for his rejection of causal efficacy as a defining characteristic of real entity as follows:

"The dBu ma thai 'gyur ba (i.e.,the Prāsaṅgika-Mādhyamika) maintains that rang mtshan is the main (subject) to be negated {dgag bya) through the logical reason (rtags) to investigate the ultimate (reality).Accordingly, he maintains that the ultimate reality consists in the negation of that very concept (don Idog) of rang mtshan asserted by substantialists (dngos smra ba),

Hence, whatever is asserted by the substantialists as the very concept of rang mtshan is [none other than] that which the dBu ma thai 'gyur ba asserts to be unestablished as a [real] basis (gzhi ma grub) even according to verbal conventions (tha snyad du yang), for such [things] as that which (exists) not sharing (ma 'drespa) place, time and essential nature [with other things], and that which is causally, efficacious are, on the contrary, accepted by the dBu ma thai 'gyur ba too [according to verbal conventions].

Therefore, these [things] are the instances of that which is to be defined [as rang mtshan} (mtshan gzhi) but are not the definiens of rang mtshan here in the case (skabs 'dir) [in which the Sautrantika tenet is treated]."[23]

Insofar as rang mtshan is a real entity, the 'concept of rang mtshan' or the defining characteristic thereof must, on one hand, correspond to the condition of real existent, the establishment of which the Prasangika-Madhyamika refutes even conventionally. In other words, the concept of rang mtshan is, for the dGe lugs pas, from the beginning determined as the object of refutation (dgag bya) from the Madhyamaka point of view, since the core of the Madhyamaka ontology consists in negating such a substantial or real existent.

On the other hand, the 'concept of rang mtshan' or the defining characteristic thereof may not correspond to that which the Prasangika-Madhyamika accepts on the conventional level, for, supposing that such a thing be the defining characteristic of rang mtshan, it would follow that the rang mtshan itself must be conventionally accepted by the Madhyamika too, which, however, contradicts his position in which the real existence of rang mtshan is not acknowledged, neither ultimately nor conventionally.

Moreover, it is also an important thesis for the dGe lugs pas that, in the Prāsaṅgika- Mādhyamaka system, all causal relations as well as causal efficacy are conventionally established.

Hence the dGe lugs pas exclude causal efficacy from the defining characteristics of rang mtshan and ultimate reality.[24]

In relation to these Madhyamaka positions, the dGe lugs pas evaluate the ontplogical views of other schools, viz.,Sarvastivada, Sautrantika and Yogacara, as being substantialist,for the reason that the latter assert such substantial or real entities as being vastu (dngos po) or svalakṣaṇa {rang mtshan), because they are 'not merely conceptually imposed but established from the side of the objects themselves' (rtog pas btags pa tsam ma yin par yui rang gi ngos nas grub pa),[25] In this manner, in order not only to include universals in the domain of real existents, but also to hold the consistency with the Madhyamaka ontology, the dGe lugs pas redefine even the most important concept of Dharmakīrti's tradition.

What the dGe lugs pa scholars thereby finally aimed at is,in my opinion, a systematization of the Buddhist philosophical teachings of the four main traditions,i.e.,the Sarvastivada, Sautrantika, Yogacara, and Madhyamaka.

For the dGe lugs pas, the question of what is a real entity or what is the reality should be answered not within the narrow scope of one tradition, but in a range of knowledge that extends over the entire historical development of Buddhist philosophy.

In other words, the dGe lugs pas intended to connect the different systems, which had developed separately in different periods in India, by reinterpreting them systematically from one common perspective.

What they actually did, however, is to reevaluate the teachings of other schools in light of the Prasangika-Madhyamaka of Candrakirti (7c), which they estimated as the highest among Buddhist philosophical systems.

This kind of attempt to systematize various philosophical thoughts in light of the Prasangika-Madhyamaka doctrine, indeed, can be seen in several discussions in the dGe lues pa exegeses.[26]

In its historical aspect, it is to be considered as a result of the fact that Candrakirti's system had won a certain popularity among Tibetan Buddhists by the period of Tsong kha pa. At the same time, however, this attempt in turn resulted in accelerating the reevaluation of Buddhist philosophical traditions in the eyes of Tibetan thinkers.

Firmly bound to tradition, but also creative, Tibetans intensively engaged themselves in the development of Buddhist philosophy.

Itis a remarkable phenomenon in Tibetan intellectual history that they rediscovered and reinterpreted many Buddhist philosophical concepts.

Redefining svalakṣaṇa is one of Tibetan challenges to the traditionalsystem of Indian Buddhist philosophy.

In this regard, it remains a fascinating task for us to discover and analyze their philosophical commitments and their underlying motives.

In this fashion, we can better establish the significance of the Tibetan developments in the history of the transmission of Buddhist thought.


  1. Cf. e.g., Georges Dreyfus, Universals in Indo-Tibetan Buddhism," in Sh. Ihara and Z. Yamaguchi (eds.), Tibetan Studies, Proceedings of the 5th Seminar of the International Association for Tibetan Studies, Narita 1989, Naritasan Shinshoji, 1992, vol. 1, pp. 29-46; Id. (ed. in collaboration with Sh. Onoda), A Recent Rediscovery: Rgyal-tshab's Rigs gter mam bshad, a Facsimile Reproduction of Rare Blockprint Edition, Kyoto, 1994; Id., Recognizing Reality, Dharmakirti's Philosophy and its Tibetan Interpretations, Albany, 1997; Tom Tillemans, "On the So-called Difficult Point of the Apoha Theory," Asiatische Studien/ Etudes Asiatiques 49-4, 1995, pp.853-889; Chizuko Yoshimizu, "Tsori kha pa on don byed nuspa"in E. Steinkellner et ai.(eds.), Tibetan Studies, Proceedings of the 7th International Seminar of the International Association or Tibetan Studies Graz 1995, Wien, 1997, Vol.2, pp. 1103-1120; Id., "Gelukuha ni yoru Kyōryōbu Gakusetsu Rikai (1) Nitaisetsu" (The dGe lugs pas' Interpretation of the Sautrantika System (1): The Theory of the Two Kinds of Reality), Jou rnal of Naritasan Institute for Buddhist Studies 21, 1998, pp. 51-76; Id., "Pramanavarttika I 40 no kaishaku ni tsuite" {On the Interpretation of Pramanavarttika I 40), IBK 47-2, 1999, pp. (97)- (101); Id., "Dṝśya and vikalpya or snang ba and btags pa Associated in a Conceptual Cognition," in Sh. Katsura (ed.), Proceedings of the 3rd InternationalDharmaktrti Conference. Hiroshima 1997, Wien, 1999, pp. 459-474; and Id., "Geluku-ha ni yoru Kyōryōbu gakusetsu rikai (2) Fuhen jitsuzai ron" (The dGe lugs pas' Interpretation of the Sautrantika System (2) The Theory of Real Universals), Bukkyo bunka kenkyu ronsyu (Studies of Buddhist Culture) 4, 2000, pp. 3-32.
  2. As for the dGe lugs pa interpretation of Candrakirti's negative position on the logical method to investigate reality, cf. e.g., Shiro Matsumoto, "Tsong kha pa no jiritsu ronshd hihan" (bTsong kha pa's Criticism of the Independent Argument), in Z. Yamaguchi (ed.), Chibetto no bukkyo to shakai {Buddhism and Society in Tibet), Tokyo, 1986, pp.475-508; David Seyfort Ruegg, Three Studies in the History oj Indian and Tibetan Madhyamaka Philosophy, Part 1, Wien, 2000; Kodo Yotsuya, The Critique of Svatanlra Reasoning by Candraklrti and Tsongkhapa, Stuttgart, 1999; Chizuko Yoshimizu, Die Erkenntnislehre des Prdsahgika-Madhyamaka nach dem Tshig gsal ston thun gyi tshad ma'i rnam bsad des 'Jam dbyans biad pa'i rdo rje, Wien, 1996; and Id., "Tsong kha pa's Reevaluation of Candrakirti's Criticism of Autonomous Inference," in G. Dreyfus and S. McClintock (eds.), The Svatantrika- Prāsaṅgika Distinction, What Difference does a Difference Make? Wisdom Publication 2002, pp. 257-288.
  3. This paper was presented at the University of Lausanne on the 26th of March, 2002, and originally written on the basis of my previous two Japanese articles, i.e., "Gelukuha ni yoru Kyōryōbu Gakusetsu Rikai (1)" and "(2)" as well as a German paper, "Das Individuelle und das Wirkliche bei den dGe lugs pa: Grundbegriffe buddhistischer Philosophic in tibetischer Modifizierung", read at the University of Munich on the 22nd of November 2001. For this revised version, I would like to thank Prof. Tom Tillemans for his valuable suggestions regarding both contents and English expressions.
  4. Pramāṇasamuccayavṛtti ad Pramāṇasamuccaya I 2 (Masaaki Hattori, Dignāga, On Perception. Cambridge, 1986, p.79 n.1.14): na hi svasāmānyalakṣaṇābhyām anyat prameyam asti. svalakṣaṇaviṣayaṃ hi pratyakṣaṃ sāmānyalakṣaṇaviṣayam anumānam iti pratipādayiṣyāmah; Pramāṇasamuccaya I 3c (Hattori, op.cit., p.82 n.1.25): pratyaksaṃ kalpanāpoḍham.
  5. Praviāṇavārttika I 166abc and its Svavrlti (in The Pramāṇavārttikam of Dharmakīrti, the First Chapter ivith the Aulocommentary, Text and Critical Notes ed. by R. Gnoli, Rome 1960): sa pāramārlhiko bhāvo ya evārthakriyākṣamaḥ. idam eva hi vastvavastuyor lakṣaṇam yad arthakriyāyogyatā Jyogyatā ca.; Pramānavārttika III 3 (in "Pramaṇavarttikakarika, Sanskrit and Tibetan," ed. by Yūsho Miyasaka, Ada Indologica 2, pp. 1-206): arthakriyāsamartham yal tad atra paramārthasat / anyat saṃvṛtisat proktam te svasāmānyalakṣaṇe //
  6. Cf: e.g - Pramaṇavarttika I 269ab: sattāviātrānubandhitvān nāśasyānityatā dhvaneḥ; Heiubindu (in Dharmakīrtis Helubinduḥ, Teil I: Tibetischer Text und rekonslruierter Sanskrit-Text ed. by E. Steinkellner, Wien 1967) 4*, 6f.: yat sat tat kṣaṇikam eva, akṣaṇikatve 'rthakriyāvirodhāt tatlakṣaṇam vastuṭvaṃ hīyate.Regarding the theoretical link between impermanence and real existence in the proofs of momentariness, cf.e.g.E. Steinkellner, "Die Entwicklung des Kṣanikatvānumāna bei Dharmakīrti", WZKSO 12/13 (1968-69), 1968, pp. 361-377; Chizuko Yoshimizu, "The Development of sattvānumāna from the Refutation of a Permanent Existent in the Sāutrantika Tradition", WZKS 43, 1999, pp.231-354; and Id., "Kōjō na mono wa naze munōryoku ka―Setsunametu ronshō no riron teki haikei (Why is a Permanent Thing Inefficacious?― The Theoretical Background of kṣanikatvānumāna)'", IBK48-1, 1999, pp. (196)-(200).
  7. One may take Pramaṇavarttika I 40 (cited in n.18 beiow) to state the uniqueness of svalakṣaṇa in this sense, as the dGe lugs pas do, if one supposes that the subject of this verse (sarvabhāva) refers solely to svalakṣaṇa.
  8. Cf. E. Steinkellner, "Wirklichkeit und Begriff bei Dharmakīrti," WZKS 15, 1971, pp.179-211, p.l83f., 188f. and Yoshimizu, "Kōjō na mono wa naze munōryoku ka", p. (197)f.
  9. Yid kyi mun sel(in mKhas grub rje's gSung 'bum Tha, lHa sa Zhol version) 21b2f. (tr. Dreyfus, Recognizing Reality,p.117): rang lugs la / rang mtshan gyi mtshan nyid rlog pas btags pa min par rang ngos nas thun mong ma yin pa'i rang bzhin du gnaspa'i dngos po'o //A similar description occurs in rGyal tshab's Thar lam gsal byed (in rGyal tshab rje's gSung'bum, Cha, lHa sa Zhol version) 45b3f. with regard to the subject (i.e.,sarvabhāva) of Pramāṇavārttika I 40 (see n.18 below), where rGyal tshab identifies as rang mlshan as 'the thing which consists, not being conceptually imposed but from its own side, in itsessential nature uncommon [with other things]' (rtogpas btags pa tsam min par rang bzhin gyis gzhan dang ma 'dres par rang gi ngo bola gnas pa, cf.Yoshimizu, "Geluku-ha ni yoru Kyōryōbu gakusetsu rikai (2), p. 23).
  10. The dGe lugs pas differentiate spyi (sāmānya) from spyi mtshan (sāmānyalakṣana). The latter signifies solely unreal, unconditioned and imagined object like space (nam mkha', ākāśa). Cf. Tillemans, op.cit.,p.865f., Dreyfus, Recognizing Reality,p.181, and Yoshimizu, "Tsori kha pa on don byed nus pa", p. 1114 n. 39.
  11. As for the relation between rang mtshan and spyi for the dGe lugs pas, cf. e.g. Yid kyi mun sel 33a4 (tr. Dreyfus, Recognizing Reality, p.181 and Yoshimizu, "Geluku-ha ni yoru Kyōryōbu gakusetsu rikai (2), p.15 n.19): rang mtshanyin kyang rang gi gsal ba la rjessu 'gro byed pa'i spyi yin par mi 'gal zhing/ Cf. also Dreyfus, Recognizing Reality, p.l73ff. and note .16 below.
  12. Regarding the synonyms of rang mtshan, see the explanations by dGe lugs pas cited in Yoshimizu "Gelukuha ni yoru Kyōryōbu Gakusetsu Rikai-(1)", pp.58 and 63 n.9. As for their commitments to Pramāṇavārttika III 3, see Thar lam gsal byed 210b6f.,mNgon sum le'u tik(in The Collected Works of Tsongkhapa 22 of bKra shis lhun po version ed. by Ngawang Gelek Demo, New Delhi 1978) 17a5f., Yid kyi mun sel44b 1 ff.(cited and translated in Yoshimizu, op.cit,p.65 n.14). Concerning the problematic Tibetan translation of arthakriyāsamartham in Pramāṇavārttika III 3 as don dam don byed nuspa and itsinterpretations, cf. e.g. mNgon sum le'u tik 166bl-4, Thar lam gsal byed 211a3ff. and the references cited in Yoshimizu, op.cit, p.61 n.8. rGyal tshab states the opinion in his Thar lam gsal byed 211a3ff. (cited and translated in Yoshimizu op.cit, p.62) that the qualification of 'being ultimate' is made to causal efficacy in Pramāṇavārttika III 3 in order to eliminate the 'false'conception that the causal efficacyis solely conventionally (kun rdzob tsam du) accepted. This 'false' conception most likely belongs to the Mādhyamikas, as willbe discussed later.In this regard, itisinteresting to remark that Se ra Chos kyi rgyal mtshan propounds the definition of ultimate reality as that which ultimately has causal efficacy (do?i dam par don byed nus pa'i chos) in his Grub mtha (in Textbooks of Se ra Monastery ed. by Tshulkhrim Kelsang & Shunzo Onoda, Biblia Tibeticaseries,Kyoto 1985) 4b3 (cited in Yoshimizu, op.cit.,p.64 n.U).
  13. Tshad ma'i brjed byang (in The ]]Collected Works of Tsong kha pa)] 22 of bKra shis lhun po version ed. by Ngawang Gelek Demo, New Delhi 1978) 34alf.: don dam bden pa'i mtshan nyid rtogpas Mags pa tsam ma yin par yul rang gi ngos nas grub pa / Cf. also mNgon sum le'u tik 17a6f.: don dam bden pa'i mtshan nyid rtogpas btags pa la ma Itos par rang gi ngos bos dpyad bzod du grub pa / Parallel definitions by other dGe lugs pas are cited in Yoshimizu, op.cit.,pp.53 and 64 n.ll. The dGe lugs pas presumably define the two kinds of realityon the basis of Dharmakirti's own words in Pramāṇavārttika I 68-91, especially 68-70, as I have previously discussed in Yoshimizu, op.cit.,pp.52-57 and "Dṛśya and vikalpya", p.460 n.5.
  14. Yid kyi mun sel21a5f.: gzhan dag yul dus rang bzhin ma dres par gnas pa dang / don byed nus pa sogs rang mtshan gyi mtshan nyid du 'dod pa mi 'thad do// One should note the fact that the similar definitions of rang mtshan appear in the sDe bdun 'jug sgb Yid kyi mun sel(in The Collected Works of Tsong kha pa 27 of bKra shis lhun po version ed. by Ngawang Geleg De mo, New Delhi 1977) 3b6, which is a glossary of terms, concepts and their definitions ascribed to Tsong kha pa, but probably descended from Phya pa's tradition of gSang phu monastery. Yet it seems more plausible to assume that mKhas grub denies the traditionallyacknowledged definitions, which Tsong kha pa and he himself have learned from their teachers, rather than to jump to the conclusion that mKhas grub thereby rejects Tsong kha pa's view, because, as will be seen below, mKhas grub gives his own definition with a clear consciousness of the theoretical consistency with Tsong kha pa's fundamental ontology as well as his understanding of causal efficacy. Thus considered, the fact that the old type of definition of rang mtshan is found in the sDe bdun 'jugsgo might support the originality of mKhas grub's definition, as I have suggested in Yoshimizu, "Geluku-ha ni yoru Kyōryōbu gakusetsu rikai (2) ", p.24f.
  15. The substantial distinction according to place and time mentioned here by mKhas grub is, however, on no account concerned with momentary existents, since he himself describes the difference of place and time as a rough incompatibility of location such as east and west and morning and afternoon. See Yid kyi mun sel 33alff.: snga dro'i ha ba phyi dro med pa dus ma 'dres pa'i don yin gyi /. . . shar la regpa'i rdzas des nub la ma reg pa Ita bu / yul ma 'dres pa'i don yin gyi / . . . khra bo la yod pa 'irang bzhin de ser skya la med pa sogs / rang bzhin ma 'dres pa'i don yin gyi / Cf. also a parallel explanation in rGyal tshab's Thar lam gsal byed 451a-4 and the discussion in Dreyfus, Recognizing Reality, p.H7ff. Moreover, it must be noted that the momentariness itself is differently understood by the dGe lugs pas as having a certain duration. Cf. e.g. Yid kyi mun sel 34a2, 34b5ff. (cited in Tillemans, 17 op.cit, p.884, Yoshimizu, op.cit., p.17 n.23) and the discussion in Dreyfus, op.cit., pp. 109-114.
  16. This explanation occurs for the first time in Tshad ma'i brjed byang 19a3f.: rtogpa layulji liar snang thing'jug pa'i tshul ni / gser bum bum par 'dzin pa'i rtog pa la gser bum yang bum par snang zhing rang gi dngos kyi gzung bya de'ang bum par snang la snang ba'i ngo na de gnyis gcig tu 'dres nas snang zhing snang ngor so sor dbyer med pas snang btags gcig tu bsres pa zhes bya ste snang ba rang mishan dang btags pa sgra don no // Cf. also Yid kyi mun sel 35a3ff., ICang skya's Grub mtha' (in mDo sde pa Chapter of Peking version, Buddhist Philosophical Systems ed. by Lokesh Chandra, Śata-Piṭaka-Series 233, New Delhi 1977) 74blff., and Thar lam gsal byed 59b5-60a3. This passage indeed has raised discussions among scholars because of its remarkable assertion that a rang mtshan appears to a conceptual cognition. For the details, see the references cited below in n.17.
  17. As for the close analysis of this passage, cf. Tillemans, op.cit., p.866, Dreyfus, Recognizing Reality, p.323, Yoshimizu, "Dṝśya and vikalpya," p.466 and "Geluku-ha ni yoru Kyōryōbu gakusetsu rikai (2)", p.llf.
  18. Pramāṇavārttika I 40: sarve bhāvāḥ svabhāvena svasvabhāvavyavasthileḥ/ svabhāvaparabhdvdbhyaṃ yasmād vyāvṛttibhāginaḥ/ /
  19. See Yid kyi mun sel 41b6f and Thar lam gsal byed 45b3f. (cited in Yoshimizu, op.cit., p.22f.). It is Śaṅkaranandana who interpreted 'all things' to refer to both individuals and universals {Pramāṇavārttika D152b6). However, this does not necessarily suggest that Śaṅkaranandana asserts the existence of.real universals, for, to my knowledge, he expresses nowhere such a realistic view. He includes universals into 'all things' presumably in a hypothetical sense in accordance with Dharmakīrti's postulation 'sati vet in Pramāṇavārttikasvavṛtti 25, 12. For this issue, cf. Yoshimizu, "Pramāṇavārttika I 40 no kaishaku ni tsuite", p.(101) n.10 and "Dṝśya and vikalpya", p. 463 n.19.
  20. See e.g., Tharl lam gsal byed 46b 1. Pramanavarttika I 40 introduces together with 41abc (tasmād yalo yalo 'rthānām vyāvrttis tannibandhanāḥ/ jātibhedāḥ prakalpyante) the idea that such concepts of properties as 'being impermanent' (anityalva) and 'being produced' (kṛtakatva) are formulated on the basis of the essential nature {svabhāva) of things, although the real existence of universals, which are identical with or different from particulars, is unacceptable. Dharmakīrti is thereby demonstrating that an inference based on the essential property as a logical reason {svabhāvahetu) is valid for establishing the reality of entities such as their being impermanent. In fact, he opens with this verse the long discussion of the apoha theory. rGyal tshab, however, interprets this apoha section of Pramanavarttika I as contributing to the establishment of the two kinds of reality(see Yoshimizu, op.cit., pp. 460-463, 470 Appendix 2). As regards Pramāṇavārttika I 40 in commentarial tradition, cf. also Dreyfus, Recognizing Reality,p.118, Yoshimizu, "Pramāṇavārttika I 40 no kaishaku ni tsuite" and "Dṝśya and vikalpya," p. 463 n. 19. For the dGe lugs pas, the question of how one can establish reality by means of inferences, if the meaning of words is mere elimination of others (anyāpoha), overlaps with the question of how the Mādhyamika can prove the non-substantiality and emptiness by means of empty words (cf. Yoshimizu, op.cit., p. 462 and "Geluku-ha ni yoru Kyōryōbu gakusetsu rikai (2), p. 28).
  21. It is interesting to note that both rGyal tshab and mKhas grub offer a similarelucidation in their respective commentaries on Pramāṇavārttika I 40, as I have pointed out in Yoshimizu, "Pramāṇavārttika I 40 no kaishaku ni tsuite",p.(101) n.10 and "Geluku-ha ni yoru Kyōryōbu gakusetsu rikai (2)", p. 22f
  22. Cf. Dreyfus, Recognizing Reality,pp.193-200 and the references cited in Yoshimizu, "Dṝśyaand vikalpya" p. 459 n.l.
  23. Yid kyi mun sel2Ia6-21b2 (cited and translated in Yoshimizu, "Gelukuha ni yoru Kyōryōbu gakusetsu rikai (2)," p. 19): dbu ma thai 'gyur ba / rang mtshan don dam dpyod pa'i rtags kyi dgag bya'i gtso bor 'dod pas / dngos smra ba 'dod pa'i rang mtshan gyi don Idog de bkagpa don dam bden par 'dod pa yin la / de'i phyir dngos smra bas rang mtshan gyi don Idog tu gang 'dod pa de / dbu ma thai 'gyur ba tha snyad du yang gzhi ma grub par 'dod pa yin la / yul dus rang bzhin ma 'drespa dang don byed nus pa sogs dbu ma thai *gyur ba yang khas len pa'i phyir ro// des na de dagskabs 'dirrang mtshan gyi mtshan gzhiyin gyi mtshan nyid min no //
  24. Cf. the discussions and textual sources cited in Yoshimizu, "On rah gi mtshan nyid kyis grub pa 111,"Journal of Naiilasan Institute for Buddhist Studies 16, 1993,pp.l29, 132f and ibid., 17, 1994, p. 327, n.67.
  25. This kind of real entity can be properly identified with 'that which isintrinsically established' {ranggi mtshan nyid kyis grub pa) in opposition to the unreal, mere conceptual existent (blags yod) or that which is postulated by names and signs {ming dang brdas rnam par gzhagpa). The expressions 'rang gi ngos nas grub pa and Wang gi ngo bos grub pa', which they use in their definitions of rang mtshan and ultimate reality, are no doubt synonyms of the former, i.e., 'rang gi mtshan nyid kyis grub pa\ Also the expression 'rtog pas btags pa' means the same as the 'btagsyod'. Cf. Helmut Tauscher, Die Lehre.von den zwei Wirklichkeiten in Tson khapas Madhyamaka-Werken, Wien, 1995, p. 124 n.262 and Yoshimizu, "Tsong kha pa's Reevaluation," Appendix.
  26. Cf. e.g., the synthesis of the Madhyamaka ontological doctrine of non-substantiality and the logicoepistemological system of Dharmakīrti's tradition by dGe lugs pa scholars, which is the object of studies such as Yoshimizu, Die Erkenntnistehre des Prāsaṅgika-Madhyamaka and Seyfort Ruegg, Three Studies in the History of Indian and Tibetan Madhyamaka Philosophy.


by Chizuko Yoshimizu