Apropos Of Lindtners Two New Works Of Dharmakirti

Author(s): Ernat Steinkellner
Publisher: Ernat Steinkellner
Catalog link: https://jainqq.org/explore/269407/1

Page #1

This paper was written in 1987. I received no proofs.
32 APROPOS OF LINDTNER'S TWO NEW WORKS
OF DHARMAKİRTI
ERNST STEINKELLNER, Vienna
In two recent publications Christian Lindtner has drawn our attention to "quotations from Dharmakirti occuring in later philosophical works but not to be traced in any of his extant treatises". These quotations which are to be found in the Madhyamakaratnapradipa of Bhavya?, in Santaraksita's Tattvasiddhi.", in Jñānaśrīmitra's Sakārasiddhiśāstra and Kşanabhanrådhyāya and in various Jaina treatises have been attributed by Lindtner to two hitherto ūdknown works of Dharmakirti, a Tattvaniskarşa, on the basis of its mention in the Madhyamakaratnapradipa and the Säkärasiddhisástra, and a Laukikapramānapariksā on the basis of its mention in the Tattvasiddhi.
In view of these extant quotations the question of the authorship of the Madhyamakaratnapradipa is irrelevant : what is important is the very existence of these and other quotations from Dharmakirti which suggests that not all of Dharmakirti's once extant philosophical works were transmitted or even generally preserved in the memory of his tradition. On another occasion I examined the possibility of tracing the reason for the classification of Dharmakirti as a Madhyamika philosopher to one of these quotations, possibly from the Tattvaniskarşa.5 The verses collected by Lindtner so far and tentatively ascribed to the Tattvaniskarşa will probably be augmented by future research. It will be difficult to regard all of them as mere inexact variations of actually extant verses, although we know that some of Dharmakīrti's verses were changed intentionally, e g. for polemical reasons, by later Jaina authors. A study of these fragments will be attempted in the future and it is to be hoped that Lindtner's findings will increase the attention given to similar quotations and references to Dharmakirti that cannot be located in his extant treatises. Here I would like to confine myself to the question of whether the Laukikapramânaparikṣā, one of the two new works newly ascribed by Lindtner to Dharmakirti, can in fact be assumed to by him.
be
A fragment and the title of a Laukikapramânapariksa by Dharmakirti was found by Lindtner in the final parts of the Tattvasiddhi. He edited the fragment (yatra...... áfrayena) determining its frame with the words ācārya-Dharmakirtipadair api Laukikapramāṇapari kṣāyāmfti, and says that it is hardly possible to take this passage as a paraphrase as might Lperhaps be suggested by the Tibetan version.9 Foc-the-Tibetan translation shows-as

Page #2

$78
Prot. Bhattacharya Commemoration Vol.
Lindtner indicates-a terminative (......brtag par) in the case of the assumed name of the text, and neither a translation corresponding to iti (e.g. zes) nor to a word of saying (e.e hi pa). I think, however, that the frame of this fragment has to be looked at in another way
The absence of a Tibetan word translating iti at the end of the quotation does not pos a serious problem, although the many other quotations in this translation regularly end in gsuns pa, when Sūtras or Tantras are quoted, or zes bśad pa when Ācāryas or Šāstras ar quoted. 10 The iti, which is attested by the manuscripts, may have escaped the translators may have been disregarded intentionally in view of the possibility that they took the text in question not as a literal quotation but rather as a paraphrase.
However the beginning of the text's frame would have to be interpreted quite differenth to Lindtner's proposal. In accordance with the introductory mention of two other Ācāryas occuring earlier in the text which are parallel to the introduction of our text : slob dponchos kyi grags. pa'i žal sna nas kyis kyan, i.e. ācārya-Dharmakirtipādair api, the name followed by api can be considered as a sufficient introduction for the following statement, be it a quotation or a mere paraphrase. The word of saying completing this elliptical introduction is lacking of course in the Tibetan translation, but not in the Sanskrit Mss. where the preceding part of the sentence, not adduced by Lindtner, ends in uktam (cf. the text below) which is to be constructed with acårya-Dharmakirtipădair api. Thus the word laukikapramanaparik sāyām. taken by Lindtner as the name of the source from which the following text had been quoted, could in fact also belong to this text as its first word." The first sentence of the presumed quotation would then have to be read in a new way, i.e. as being introduced by a reference to the systematical frame within which the following statements have to be localized, and that frame is an examination of the worldly valid cognitions".
A distinction between two types of valid cognition was formulated by Dharmakirti at the very end of PVin I: an empirical one (samvyavahărika) and a transcendental onc (paramarthika). 12 LINDTNER 1984 : 158f. gives an interesting and plausible explanation for the fact that Dharmakīrti prefers this terminology, and not the pair laukikalokottara. Then why would he speak of a laukikapramana in the present case? The problem disappears when wc assume that this term was not formulated by Dharmakirti himself but by Santarakṣita who, as the author citing Dharmakīrti or referring to his words, uses a term synonymous to Dharmakirti's term but thereby neglects the possibly different approach intended by Dharmakirti's choice.
If we now look at the presumed quotation itself we have the choice of either taking the whole text up to iti as a quotation from a hitherto unknown text, to which śāntarakṣita13 added the systematical frame at the beginning, or taking the whole text as a reference to Dharmakīrti's and Dignāga's14 positions as formulated by Santarakṣita himself. In the latter case only the final iti remains as a disturbing element. But not too much so, since it could either be taken as an iti of merely concluding value, simply emphasizing the conclusion of this reference, or as

Page #3

Apropos of Lindtner's two new works of Dharmakirti
279
the addition of a scribe who understood the preceding text as a quotation. In both cases the absence of the word in the Tibetan translation would be normal.16 It is finally to by noted that none of the many quotations of verses from Tantras and Šāstras in this text is concluded by a quotative iti. Also some of the few prose quotations show no iti (K 5, 6; 14, 13), while others have it (K 5, 10; 6, 15; 8, 3). Thus the import of the presence or absence of an iti in our case for an argument in favour or against the preceding text's being a veritable quotation seems seriously decreased.
Looking at the text as a quotation would allow us to add another fragment to the small treasure of materials not to be found in Dharmakirti's extant works and also attribute it tentatively to the enigmatic Tattvanişkarşa, for the theme treated goes well with other materials collectable under this title.16
If, at the other hand, we look at it as a general paraphrase of a certain position of Dharmakirti and Dignaga formulated by Sāntarakṣita, we have to indicate its sources in the works of these authors in as much detail as is possible with such general statements.17
In any case we have to deal with the Sanskrit and Tibetan texts of this passage as well as with its translation because I believe that Lindtner's edition and interpretation can be improved upon.19
* For the Sanskrit text I referred to the same two manuscripts that were used by Lindtner and to the edition of the Kameshwar Nath Mishra which is based on four further manuscripts too.19 For the Tibetan text I used the Peking editi in (P 4531, 40b9--41a4) and the Cone edition?" (C 37a7-3763). For the variinis of the Narthang edition (N) I relied on the readings mentioned by Lindtner.
Sanskrit text:
kim ca nāmajātyadio-kalpanāsamāropitamb arthānvayavyatirekānukāric savikalpakam apid jñānam spastābhanruktam acõrya-Dharmakirtipādair api - laukikapramānaparikṣāyāmf yatra grāhyagrahakapratibhasabhedenasamjāyateh tatrāspastabham eva jñānami spastābhatayābhāti. tadyatha nāmajātyädil_kalpanārahitam jäänami spaștābhamı bhavati, grähyagrāhakasainvittin. bhedena bhidyate. tena nāmajātyādikalpanām āśrityācāryeņāpya aspastābhatā darsitā, na grähyagrahakasamvittip.bhedāśrayene?-ti".
a ojanyadio AB. b kalpanasamāropitam B : kalpānām asamāropitain A (gha): kalpanāsnāsamūropitam
kha : kalpanam samăropitam K. c Pavyatirekānukāri K (cf. Idog pa dan mthun pa) : Pāvyat irekāntakäri AB (kha, gha) d savikalpakam api A, K: savikalpakapario B.

Page #4

280
Prof. Bhattacharya Commemoration Vol.
n tel : jñānam
e jñānami spastābham AB : ses pa mi gsalba dan minon par ldan pa yin tel1 : is
[sic !] aspastábham K. f 'parik şaram AB. g pratibhāsao corr. (not transl. into Tib.): "pratibhāsi AB : pratibhåso K. b samjāyate corr. (Lindtner with kun tu bskyed pa NPC)”? : samjñāyate AB, K i °āspastābham AB, K: spastābham Lindtner. j tatra : de Tib.; aspaștābham eva jñanam not transl. into Tib. k nājanyádio AB.
jñānam (K Çf. ses pa): ajñānam AB. m spastābham K (cf. gsal bar sna; ba): spasta AB-3. n samvittio is here not well translated by riog pa which usually translates words for
"conceptual constrution", e.g. kalpanā. rtogs pa would be better. o api not transl, into Tib. p samvittio corr. (Lindtner), K : samvartti A : samvirti B.
asrayene corr. (Lindtner), K :asramene A: Säsrajena B. iti not transl. into Tib.
g r
Tibetan text :
'on kyai min dan rigs la sogs par brtags pa' i don grirjes su 'gro ba dan Idog pa dur. mthun pa'i rnam par rtog pa dan bcas pa' i ses pa mi gsal ba dan mon par ldan pa yin te slot dpon chos kyi grags pa' i cal sra nas kyis kyai' jig ten pa' i tshad mu brtag parc gan det gur. ba dan' dzin pa' i dbye bas kun tu bskyed pa de gsal bar snai ha nid du snan ba sted / 'di Itur min dai rigs la sogs par brtags pa ilaj bral ba nid kyi ses pa ni gsal bar snai bar gyur pa yin no gzun ba dan 'dzin pa'i rrogo pa' i dbye bas phye ba yin la de bas na min dan rigs la sogs par brtags pa la brten te / slob dpon gyis" mi gsal ba ñid du bstante gaun ba dan 'dzin pa yan dug par rig pa' i dbye ba la brten pa ni ma yin no //
a par corr. : pas CP b gyi P : gyis C c par C: par / N (?) P d de gsal bar snan ba ñid ilu snai ba ste NP : gsal bar snar ste C e rtogs would be preferable fbstan C : bsten NP
Translation :
"Moreover, that even a conceptual cognition which superimposed conceptual construc tions such as name and genus (and) agrees in presence and absence with something, has

Page #5

Apropos of Lindtner's two new works of Dharmakirti
281
distinct appearance," has been said by the reverend teacher Dharmakīrti too : Insofar as the worldly valid cognitions (laukikapramāna) are examined," when30 (cognition) occurs as divided into the reflection (pratibhasa) 31 of an [objective part which is) grasped and a [subjective part which is] grasping, then even a cognition of indistinct appearance shines as distinctly appearing:33 To wit.: A cognition devoid of conceptual constructions such as name and genus has a distinct appearance; 34 it is (nevertheless]35 differentiated as divided into the awareness of the grasped and the grasping." Therefore the teacher (Digpāga) has also38 shown that ca cognition) when it rests upon conceptual constructions such as name and genus is of indistinct appearance, 39 but not in that it rests upon the division into the awareness of the grasped and the grasping."
The whole passage can thus be accounted for even when we understand it not as a quotation from an unknown work of Dharmakirti, but rather as a paraphrase in general formulations given by śāntarakṣita with regard to a certain tenet of Dharmakīrti and Dignaga. Thus it does not seem necessary to interpret the expression laukikapramānaparikså as the title of a hitherto unknown work by Dharmakīrti.
Abbreviations and recurring literature
n
TSi
N
р
PS1
PV
Cone Tattvasiddhi : P 4531 Narthang Peking Pramāņsamuccaya, first chapter. In HATTORI 1968 Pramāņavārtika : Dharmakirti's Pramāņavārttika with a commentary by Manorathanundin. Ed. RÄHULA SĀNKRTYAYANA. Patna 1938-40 [but the verses are counted with Vetter according to the tables in VETTER 1964 : 1165] Pramāņavārtikabhashyam or Vārtikālankāraḥ of Prajñākaragupta (Being a commentary on Dharmakirti's Pimāņavārtikam). Ed. RAHULA, SĀNKRTYĀYANA. Patna 1953. Pramāņaviniscaya, Kapitel I : TILMANN VETTER, Dharmakirti's Pranāņaviniscayah, 1. Kapitel : Pratyakşam. Einleitung, Text der tibetischen Übersetzung, Sanskrit fragmente, deutsche Übersetzung. Wien 1966. Dignāga, On Perception, being the Pratyak sapariccheda of Dignag's pramāṇasamuccaya from the Sanskrit fragments and the Tibetan versions. Translated and annotated by MASAAKI HATTORI. Cambridge, Massachusetts.
PVBh
PVin!
HATTORI 1968

Page #6

282
Prof. Bhattacharya Commemoration Volume
JAMBOVIJAYA 1981 MUNI JAMBOVIJAYA, Jaipācārya-Śri-Hemacandrasūrimukhyasis
yābhyâm ācārya-Ramacandra-Guņacandrābhyam viracitāyām Drava
yalankara-svopajña-tīkâyām Bauddhagranthebhya uddbstähl' Studien V paṭhah.
zum Jainismus und Buddhismus. Gedenkschrift für Ludwig Alsdorf,
Hrsg. KLAUS BRUHN U. ALBRECHT WEZLER. Wiesbaden. LINDTNER 1980 Apropos Dharmakīrti— Two New Works and a New Date, Acta
Orientalia 41, 27-37. LINDTNER 1984 Marginalia to Dharmakīrti's Pramāņaviniscaya I-II, WZKS 28.
(149-175) 160-162. VETTER 1964
TILMANN VETTER, Erkenntnisprobleme bei Dharmakirti, Wien. VETTER 1966
PVin 1 - translation : cf PVin I
REFERENCES & NOTES
1. LINDTNER 1980 and 1984. 2. LINDTNER 1984 : 162. 3. Peking edition No. 5254. For the question of the authorship of this text cf. DAVID
SEYFORT RUEGG, The Literature of the Madhyamaka School of Philosophy in India, Wiesbaden 1981 : 66 and note 339; CHRISTIAN LIND INER, "Adversaria Buddhica", WZKS 26, 1982, (167-194) 174-182 and 182-184 for the arguments of Ejima Yasunori; D. SEYFORT RUEGG, “Towards a Chronology of the Madhyamaka School”, in: Indology and Buddhist Studies, Volume in Honour of Professor JW. de Jong on his Sixtieth Birthday, ed. by L.A. HERCUS a.O., Canberra 1982, (505-530) 530. D. Seyfort Ruegg further presented a paper at the VIIth World Sanskrit Conference, Leiden 1987, “The literaryhistorical problem of Bhāvaviveka / Bhavya” (to be published by Brill 1988 ?) in which he draws attention to a number of arguments against an early date for the Madhyama
karatnapradipa. 4. TSi : Peking edition No. 4531. Editions of the Sanskrit text of the Tattvasiddhi are
announced as forthcoming by Lindtner (LINDTNER 1980: note 33) and by Kameshwar Nath Mishra, Central Institute of Higher Tibetan Studies, Sarnath (in a paper read at the Vith World Sanskrit Conference, Philadelphia 1985). Lindtner has promised to present the evidence for the authenticity of the Tattvasiddhi as a work of śāntarakṣita in the introduction to his edition.
Prof. Kameshwar Nath Mishra's edition has already been composed in Devanagari and a copy of his text (29 pages) was put at my disposal for which I would like to express my sincere gratitude.

Page #7

Apropos of Lindtner's two new works of Dharmakirti
283
5. “Is Dharmakirti a Madhyamika ?" (Paper read at the VIIth World Sanskrit Conference,
Leiden 1987. To be published by Brill 1988 ?) 6. LINDTNER 1980 : 29, 31 and 1984 : 160f. and note 36. 7. No fragment in prose, except for the one examined below, has been found to my knowledge. 8. LINDTNER 1980 : 33-37. 9. LINDTNER 1980 : 35. 10. Cf. slob dpon Nagärdzuna' i zal sia nas kyan .....zes bsad pa dan / TSI 306 4f., or: 'Phags
pa' i lhas kyan ......zes bśad pa yin no // TSI 316 4f =K 5, 6: punas coktam ācārya-Nägarjunapádaih ...., and K 7, 14. : uktañ ca Aryadevena...... In neither case is an additional
mention of their works given. 11. It is also interesting to note that the sad (danda) after the word 'jig rten pa' i tshad ma
briag par in the Peking (and Narthang ?) edition is absent in the Cone (and Derge ?)
edition. 12. PVin I 100, 20-24. The Sanskrit fragment for the passage PVin I 100, 14-24 was identi
fied in the Nyāyabhūşana (NBhūş 57, 14-19) in LINDTNER 1984 : 156f. and by YAMAKAMI SHODO ("Nyāyabhūşaņa no kenkyū (1)", Kyoto Sangyo Daigaku Ronshū 14, 1984, [134-167) 136f.). A fragment of Dharmottara's commentary was found by Muni Jambūvijaya in the Dravyalamkarasvopajñatika (JAMBUVIJAYA 1981 : 137; also quoted
in LINDINER 1984 : note 23). 13. Or even by Dharmakirti himself, but then Lindtner's suggestion for an explanation of the
different terminology would be invalid or could be upheld only when we assume a histori
cal change in Dharm akirti's attitude without, however, being able to prove it. 14. The second part of the passage refers to Dignāga. This reference must then again be
taken as one made by Dharmakirti or directly by Saptarakṣita. 15. In the former case, because this kind of iti is often not translated, and in the latter, because
the copy used by the translators did not have the iti. 16. Cf. LINDTNER 1984 : 1611.; it would be the only fragment in prose, however, found
so før. 17. This will be done in my notes on the translation below, 18. Mainly because Lindtner in his attempt to make sense of the passage, used not only the
Tibetan text to emend the Sanskrit text, but also corrected the Tibetan text with the help of the Sanskrit text. We all do this, of course, but with some restraint and not when the received Tibetan text as such makes sense, or when it is clear that something else has been translated or that something has been translated incorrectly. In my edition below
I shall refer to Lindtner's textual proposals as well. 19. I would like to thank Christian Lindtner very much for putting copies of the relevant
folios at my disposal. Ms. A means folio 106a 1-4 (the Ms. referred to by Lindtner) and

Page #8

284
Prof. Bhattacharya Commemoration Volume
Ms. B means folio 19b 1-4. K means the text as edited by Kameshwar Nath Mishra (7. 26, 7-14) with variants indicated in his footnotes from his Mss. kha, gha. I added these variants in brackets not being able to identify for certain the Mss. with those of
Lindtner's at the moment. 20. Microfiche edition of The Institute for Advanced Studies of World Religions, New York,
Vol. 206, Tsu 26a 7-38b6. 21. Tib. mron par ldan pa I do, however, not understand. . 22Cf. also skye ba in the formulation of PVin I 88, 12f. : gan dag gi blo ñid gzun ba dan 'dzin
pa' i rnam pa ne bar ston par skye ba...... 23. Lindtner corrects spastam bhavati, but the Ms. reading spaștā bhavati must be taken as a
haplography for spastābham bharati. 24. According to PC; Lindtner gives na as the reading of N, but this must be simply an
unclear carving or printing of de. Lindtner proposes a correction la : de in accordance
with his reading yatra. But the Tibetan translates rather yat tat instead. 25. Lindtner corrects de la and adds ses pa ni in accordance with the reading tatra and with
the words spaștabham eva jñānam, not accounting for spaștábham eva, however. 26. Lindtner adds kyan for api of the Mss. 27. Io translating, especially the various terms, I follow Lindtner's words as far as possible. 28. This statement, that conceptual cognition which is normally of indistinct appearance can
said to be distinct in a certain sense, corresponds exactly to the meaning of the reference to Digoaga below. The Tibetan translation has mi gsal ba dan mñon par Idan pa yin for the predicate spaștābham and seems to misunderstand sāntarakṣita's intention in favour of the
more evident theorem. But I do not understand the second part of the Tibetan predicate. 29. This would be stated by Sântarakṣita in order to indicate the frame of reference for the
following general paraphrase of Dharmakirti's tenets in support of his statement that even a conceptual cognition (savikalpakam api jñānam) is of distinct appearance (spastābham). Dharmakirti's position is only meaningful as far as the worldly valid cognition is concerned, for the dichotomy of “grasped and grasping" is left behind when the transcendental cognition (päramärthikam pramānam) comes into focus (cf. LINDTNER
1984 : 157f.). 30. That is in the case of Yogin having brought his meditation upon a real or unreal object to
perfection (cf. PV II 285). Tib gan de would be “whatever” and seems to translate the generalized relative pronoun yat tat ("every”). It is quite possible that this is the better reading which was corrupted into yatra which then motivated the change to tatra below,
for which Tib, has only de for a possible tat or even tat fat. 31. pratibhasibhedao of the Mss. (“a division which reflects", Lindtner) is odd, because the
division itself does not reflect, i.e. lit.' have reflections".

Page #9

Apropos of Lindtner's two new works of Dharmakirti
.
285
32. Cf. PVin I 88, 12f. : gan dag gi blo ñid gzun ba dag 'dzin pa' i rnam pa ñe bar ston par
skye ba...... 33. Sāntarakṣita's formulation, that a cognition occurs. as divided into its objective and subjec
tive parts, means that only its nature is referred to. And this nature (svarūpa) cannot be connoted by words (sabdārtha). The cognition of its own nature is therefore direct (adhyak şa), i.e, with distinct appearance (spastabha). This is also the case when we refer to a conceptual cognition (savikalpakajñāna), which as such, i.e. as conceptualizing an object (artha), is of indistinct appearance.
Santarakṣita paraphrases here PV II 287 which is Dharmakirti's formulation of the idea of Dignāgas PS I 7ab (cf. below note 40). In these statements both authors seem to give an answer to the question how the originally conceptual cognitions of a Yogin can attain the quality of distinctness through the force of a perfect meditation (cf. Prajñākaragupta's introduction PVBh 331, 11-22). And Dharmakīrti says : sabdarthagrāhi yad yatra taj jñānam tatra kalpana / svarūpañ ca na sabdarthas tatrādhyakşam ato khilam // (PV II 287)= "A cognition which grasps the connotation of a word with regard to a certain (object) is with regard to this (object) a conceptual construction. But the nature of this conceptual cognition] is not connoted by words. Therefore this whole (cognition) is direct with
regard to this (its owo nature).” 34. With regard to the character of distinct appearance (spaștāvabhāsitva) in a cognition as
being in opposition to its conceptuality PV III 498ab may be adduced : tasya sapstāvabhasitvam jalpasamsargiņaḥ kutah /="How could this conceptual cognition) which is connected with words have a distinct appearance ?" and PV III 283ab from Dharmakīrti's explanation of yogic perception : na vikalpānubaddhasyasti sphuţarthāvabhāsită / --“There is no distinct appearance of something in a (cognition) that is connected with conceptual construction."
Distinctness and the absence of conceptual construction are closely connected in Dharmakirti's concept of perception (pratyak şa) : cf. PVin [ 40, 2 for the definition of perception and Pyin I 28 for the distinct cognition of the Yogin as a kind of perception (cf. also VETTER 1966 : 23f). For conceptual construction cf. PVin [ 4b, the source of
the examples "name, genus etc." is PS I 3d. 35. The Tibetan translation connects this sentence with the following, but in analogy to the
full reference to Dignāga following, an antithetical connection with the preceding sentence
is preferable (cf. also Lindtner's translation with “but......"). 36. The Tibetan translation with rtog pa is problematic. If corrected into rtogs pa we may
assume a word that can be synonymous with samvitti, but I can offer nothing as supported by the relevant source.
For the two parts cf. e.g. PV III 320a : tasmad dvirūpam asty ekam and the source of these two forms of cognitions in Dignāga's PS I llab with the Vrtti (cf. HATTORI 1968 : note 1.61).

Page #10
Prof. Bhattacharya Commemoration Volume It is also possible that all three parts of a cognition, i.e. objective and subjective parts and selfawareness are indicated here (cf. PV III 353 and 357, VETTER 1964 : 823 but this would be contrary to the Tibetan translation and of no import for the meaning of the reference in Santaraksita's context. santaraksita's paraphrase of Dharmakirti's standpoint ends here in my opinion. The following sentence adds a refcrence to Dignaga as a second support for his intentions and does not have to be taken as a reference to Dignaga made by Dharmakirti himself, as according to Lindtner. api is not translated into Tibetan. To my knowledge Digpaga does not consider the concept of distinctness (spastabhata) in his presentation of the terms of relevance here. It seems that santaraksita-following Dharmakirti-equates Digoaga's predication that some sort of cognition "is a perception" (pratyaksa)-e.g. yatraisa kalpana nasii tat pratyksam at the end of his Vitti on PS I 3d (cf. HATTORI 1968 : 25 and note 1.30)-as equivalent with the predication that "it is of distinct appearance", thus identifying the concept of being a distinct cognition with that of being a perception. In this way he can refer to several statements of Dignaga, where perception is defined as nonconceptual, e.g, PS I 3c (pratyaksam kalp napodham, cf. HATTORI 1968 : note 1.25), then the sentence adduced above and the end of his Vitti on PS 1,5 (sarve ir avikalpaka eva, cf. HATTORI 1968 : note 1.44). The whole paraphrase refers to PS I Tab and its Vitti, where Dignaga states that even a conceptual construction (kalpana) is aware of itself and that this awareness is perception (kalpanapi sva samvittavista, cf. HATTORI 1968 : note 1.51). But this is not the case when an external object is conceptualized, i.e. connected with name, genus etc. (narthe, vikalpanat, cf. HATTORI 1968 : notes 1.51 and 52). Saotaraksita reformulates this source in accorda que with his point, namely that even a conceptual cognition can be said to be distinct, and says-on the basis of PS I 7abthat Dignaga taught that a cognition is indistinct, i.e. not a perception, as far as it rests on conceptual constructions, but not as far as its own division into the two parts of an awareness of its objective and subjective parts or even into the three parts of object, subject and selfawareness (cf. above note 36) is concerned.