app1_torella_apoha

• rt I: Anupalabdhi and Apoha in a Śaiva Garb**

• ffaele Torella

• the outset of an article published in 1988, I noted that the
• varapratyabhijñā-Vivr̥ti or ṭīkā (henceforth Vivr̥ti) would deserve a
• ominent place in any imaginary catalogue of the lost works of Indian
• ilosophical literature (Torella 1988: 137). Thirteen years later, in
• ich the Pratyabhijñā school of the Kashmirian Śaivādvaita has
• ntinued to be one of my preferred fields of research, I would still
• lly subscribe to that earlier statement. The role of Utpaladeva’s long
• mmentary on his Īśvarapratyabhijñā-Kārikā (ĪPK) as the real centre of
• avity of Pratyabhijñā philosophy has become, in fact, clearer and
• earer, though the later Śaiva philosophical tradition chose instead
• hinavagupta’s Īśvarapratyabhijñā-Vimarśinī (ĪPV) as the standard text.
• is brilliant and more compact and accessible text was copied again and
• ain over the centuries and became popular in south India too, where a
• mber of manuscripts in the principal southern scripts are still
• ailable. The success of a particular commentary is very often the
• direct cause of the decline of others, which are less and less read
• d, consequently, copied, until they are completely or almost
• mpletely lost. One could mention, among others, the particularly
• emplary case of the Yuktidīpikā on the Sāṁkhya-Kārikā that has come
• wn to us in a very few copies.

• need not repeat here what I have already said elsewhere on the nature
• the Vivr̥ti and its relation to the other commentaries.1 Suffice it to
• call that Utpaladeva is said to have composed the ĪPK and the concise
• ̥tti at the same time, and later on to have devoted an analytic
• mmentary to the complex of Kārikā–Vr̥tti, i.e. our Vivr̥ti (or Ṭīkā), in
• ich he discussed possible alternative views and rejected them, also
• dulging in occasionally quite long digressions on particular subjects.

• e only mention of an extant fragmentary manuscript of the Vivr̥ti
• curred in K.C. Pandey’s pioneering book on Abhinavagupta and the
• -called Kashmir Shaivism (Pandey 1963: 69-70; see also Torella 1988:
• 1). Though the original manuscript was apparently lost, a transcript
• it, by Pandey’s own hand, was found among the papers left by the
• arned scholar upon his demise and kept by his widow on behalf of the
• hinavagupta Institute, University of Lucknow. During my visit to
• cknow in 1987, Mrs Lila Pandey and the Trustees of the Institute were
• kind as to show me the notebook with the transcript, but did not
• low me to copy or photograph it, except for a small portion (10 pages
• t of 49), which I edited and translated in the above-mentioned
• ticle. My efforts in later years to secure the entire transcript were,
• spite the good offices of N. Rastogi and J.P. Sinha, totally
• successful. By now, Pandey’s transcript is to be considered
• finitively lost.

• e good fortune that sometimes helps scholars in their research,
• unteracting the obstacles set up by human unhelpfulness, took me in
• 98 to the India National Archives in Delhi in search of the
• nuscripts used by the paṇḍits of the Srinagar Research Department for
• e editions of the Kashmir Series. I was interested in those
• nuscripts because of the important marginal notes that some of them
• ntained, where I assumed to find quotations from lost Kashmiri works.
• en I chanced upon a Śāradā manuscript, entitled Pratyabhijñā-Vivr̥ti
• kt. ms No. 30), belonging to a small fund of manuscripts that the
• haraja of Jammu & Kashmir had transmitted to the National Archives in
• 48, I began looking through it with the same scepticism with which I
• d in the past looked through manuscripts with the same title all over
• dia (in all cases, the title had simply been the outcome of careless
• taloguing). Very soon, however, I realized that what I had in front of
• was precisely the original of Pandey’s transcript, as described by
• m on the very first page of his notebook.2 Once the transcript had
• come irretrievably lost to us, the original came to light again.

• e thirty-three folios of the manuscript cover the full text of ĪPK
• 3.6 through 1.5.3 along with Utpaladeva’s Vr̥tti and Vivr̥ti, plus
• hinavagupta’s Vimarśinī. The manuscript begins with the Vimarśinī’s
• ūmikā to ĪPK I.3.6 (saṁskāre saty api smr̥tiḥ [. . .], vol. 1, p. 103,
• 6) and ends abruptly with iha bodhas tāvad abhinnaḥ prakāśamātram eva
• asya paramā-<rthaḥ>, i.e. the first line of the Vimarśinī on ĪPK
• 5.4-5 (vol. 1, p. 135, ll. 5-6). Although the portion of the recovered
• xt is comparatively short, it proves to be particularly important for
• e study of Pratyabhijñā philosophy since some very crucial points are
• alt with there at great length. Among them, we find the nature of
• gnition, its being self-aware, the impossibility of a cognition’s
• coming the object of another cognition, the necessity for the various
• gnitions to rest on a single substratum for a connection between
• gnitions to be possible, and the necessity for cognition and its
• ject to share the same essential nature, i.e., “light”. Utpaladeva,
• d after him Abhinavagupta, progressively attain their final position
• means of a very interesting dialogue, mainly with the Buddhist
• amāṇa tradition and the Vijñānavāda, with the participation of the
• māṁsakas (Bhāṭṭas and Prābhākaras) on specific points.

• this contribution I am continuing, after an interval of thirteen
• ars, my above-mentioned article, which contained the edition and
• anslation of the full text of the Vivr̥ti on ĪPK 1.3.6 up to the
• ginning of the Vivr̥ti on ĪPK 1.3.7. The present article is devoted to
• e edition and translation of the full text of the Vivr̥ti on ĪPK 1.3.7.

• e third āhnika of ĪPK deals with the untenability of the opponent’s
• eses as regards the unification of cognitions, which alone renders all
• man experiences (vyavahāra) possible (cf. the relevant colophon of the
• marśinī [ĪPV I: 112]: paradarśanānupapattir nāma tr̥tīyam āhnikam). The
• eses of the opponents were presented in the previous āhnika (ibid.:
• : pūrvapakṣavivr̥tir nāma dvitīyam āhnikam). In order to better
• derstand the content of v. 1.3.6 of the ĪPK, we must read it together
• th the previous verse, to which it is closely connected (tr. Torella
• 02: 102-03):

• us, the functioning of the human world — which stems precisely from
• e unification (anusandhāna) of cognitions, in themselves separate from
• ch other and incapable of knowing one another — would be destroyed . .
• (I.3.6) . . . if there were no Maheśvara who contains within himself
• l the infinite forms, who is one, whose essence is consciousness,
• ssessing the powers of knowledge, memory and exclusion (I.3.7).

• paladeva’s final aim is to establish a single cosmic consciousness,
• e. Śiva, as the common background of all reality and, particularly, of
• l human experiences. He pursues this intention not by apodictically
• ating the Śaiva truth — for example, by resorting to the authority of
• vealed texts — but by critically examining diametrically opposed
• ctrines, i.e. those of the Buddhist pramāṇa philosophers. In doing so,
• recognizes the leading position of the Buddhist philosophy in the
• shmir of his day and at the same time shows that he is not afraid of
• allenging the great cultural prestige of the Buddhist doctrines.
• reover, in many cases he makes use, as far as possible, of their
• guments to preliminarily build up and then refine his own position.
• is way of carrying out the “duel” with the pramāṇa philosophers
• minds one of the attitude often found in oriental martial arts.
• stead of directly attacking his adversary, Utpaladeva seeks to exploit
• e adversary’s intellectual ability to finally turn it back upon him.
• e main technique on which Utpaladeva’s argumentation is based is the
• e well known in śāstric debate as prasaṅga-viparyaya. The Buddhist
• eas are not rejected at the outset but are rather apparently accepted
• d then pushed to their extreme consequences.

• a result, the Buddhist opponent is finally confronted with two
• ternatives: either to abandon his own specific theses or, if he is
• ill convinced of them (and, interestingly, this is the case not only
• r the Buddhist, but for the Śaiva philosopher himself), to change the
• rspective from which they are to be viewed. In other words, he should
• cept the overall theoretical Śaiva framework precisely to safeguard
• s own Buddhist ideas.3

• e main part of the Vivr̥ti passage edited here is devoted to the
• scussion of the Buddhist concepts of “non-perception” (anupalabdhi)
• d, less extensively, of “exclusion” (apoha) as the very core of
• nceptual thought. If one accepts the self-contained nature of
• gnition (as the Buddhist does), it becomes impossible to account for
• ther of them. On seeing an empty surface we would not be allowed to
• y that there is no jar there because the cognition of a certain object
• uld not be able to give rise to, at the same time, alternative
• gnitions later to be rejected. Yet the world of human knowledge,
• nguage and practical activity has among its pillars precisely the
• ssibility of taking something for “absent”, or of building mental
• ages by way of exclusion. Furthermore, granted that the Buddhist
• scription of how an individual cognition works in isolation is
• nsidered as basically correct, the capability of the individual
• gnitions to communicate with each other, to enter into a network, is
• quired. Such a “network” cannot be provided but by the single and
• itary consciousness in which all of them are rooted, in a word: by the
• va-nature which permeates reality, constituting its ultimate ground.

• th the topics of anupalabdhi and apoha will receive a separate and
• tensive treatment later on in the ĪPK (see I.7.7-11 for anupalabdhi
• d 1.6.1-5 for apoha). Here they are preliminarily discussed from a
• rticular viewpoint, i.e. as exemplary grounds for the necessity to
• stulate a single consciousness-principle (in the case of anupalabdhi)
• d a basic “universal” stuff in everything (in the case of apoha). The
• ddhists account for the establishment of the absence of a jar in a
• rtain place by saying that the cognition of the empty place, being
• are of itself as opposed to the cognition of the place together with a
• r, is also aware of the place as “devoid of a jar”. In fact, if a jar
• re present in a certain place, the jar would also be manifested in the
• gnition of the place since both the jar and the place share the same
• pacity [of being perceptible] (tulyayogyatārūpatvāt). This can be
• intained, the Buddhists go on, because particulars belonging to the
• me class (i.e. “visible” particulars) are grasped by one and the same
• gnition (here a “visual” cognition), and, consequently, in the case of
• jar being there, the cognition of the place should also contain a
• nifestation of the jar; but in the case at issue, this is not so.
• paladeva points out that we are dealing here with two distinct and
• parate cognitions, which are bound to remain so: the one can by no
• ans include or exclude the other, unless one intends to explain the
• tablishment of the absence of a jar in terms of inference, which is
• t assumed by the Buddhists (see Torella 2002: 143 n. 17). These and
• her arguments are taken into account in the text (see below). To
• mmarize, each cognition sheds light on itself alone; or, in other
• rds, each cognition has its own self-awareness (svasaṁvedana), which
• n by no means act as a common basis where two distinct cognitions may
• et. No single cognition, Utpaladeva concludes, not even perceptual
• gnition, can by itself establish the absence of something. The only
• y to fill the gap between cognitions and in this way give a reasonable
• planation of the establishment of “absences” which is so common in
• eryday life is to accept their resting on a single consciousness.
• stead of a multiplicity of unrelated svasaṁvedanas each corresponding
• a single cognition, Utpaladeva posits the one supreme
• nsciousness-principle as the common svasaṁvedana of all cognitions,
• ich in this way derive from it the capacity to enter into relation
• th each other.

• e starting point for the discussion of the apoha issue is clearly
• ated in Abhinavagupta’s Īśvarapratyabhijñā-Vivr̥ti-Vimarśinī (ĪPVV I:
• 2, 9ff.): Why, on the appearance of a particular object, should one
• gure out something different destined to be excluded later (nanu
• alakṣaṇe ’vabhāte kuto ’tadrūpam āśaṅkyate yad apohyate)?
• hinavagupta is not satisfied with Dharmakīrti’s answer, elaborated at
• ngth in the Svārthānumānapariccheda chapter of his Pramāṇa-Vārttika
• d Svavr̥tti: the correct knowledge of the particular (svalakṣaṇa) is
• ntinually threatened by erroneous superimpositions (samāropa), caused,
• g. by misleading similarities — such as the brightness equally found
• mother-of-pearl and in silver — which make the knower mistake one
• ing for another. The doubt which projects alternative objective forms
• be negated would arise precisely because the knower is aware of the
• ct that in principle many superimpositions are possible; thus a
• finite ascertainment becomes possible only after hypothetical mistaken
• rms are preliminarily rejected. Subsequently, Abhinavagupta takes into
• count the position of Śaṅkaranandana (ll. 18ff.), who in his
• ohasiddhi recognizes prasaṅga, i.e. the mere possibility that a thing
• different from itself (?) (anyathā-prasaṅga), as the cause of doubt.4
• th samāropa and prasaṅga are untenable for Abhinavagupta because, no
• tter whether they derive from beginningless nescience or from other
• uses, they ultimately lead to a regressus ad infinitum.

• paladeva puts forward his own version of how apoha works, taking for
• anted that apoha is indeed the pivot of all conceptual thought. His
• ew, evidently nourished by Buddhist ideas, is centred on the plurality
• causal efficiencies found in any object, an idea he substitutes for
• at of the plurality of wrong assumptions (vikalpa, samāropa), with
• spect to which Abhinavagupta will clearly explicate the fault of
• finite regress (anavasthā) (cf. below, n 68) already implicit in
• paladeva’s argument. But, more importantly, he makes an additional
• mark. On the level of māyā, on which alone discursive thought
• ikalpa) is at work, the objects should be at their highest level of
• fferentiation; then, how can their cognition also evoke “other”
• jects? Utpaladeva’s answer is rooted in Śaiva theology. The
• varatattva-level on which the whole manifested world is still enclosed
• the I (sarvam idam aham) remains so to speak in the background even
• the world of māyā, which is marked by full-fledged differentiation.
• is is especially true for the process of cognition, in which even the
• lly differentiated object ends up flowing into consciousness and being
• sorbed into it; moreover, as is often repeated by the Śaiva
• ilosophers, if the object were not essentially light, it could not at
• l shine in cognition (cf. ĪPVV I: 294, ll. 18-20: sa hy eko ’pi bhāvo
• stuto ’nyaiḥ saha tadātmaiva yadyapi māyāśaktyā pr̥thakkr̥taḥ, tathāpi
• ātatvād eva anyonyātmatām unmīlayati; see also p. 296, ll. 5-8). This
• tent basic non-differentiation of the object (from other objects and
• om the Self) is, according to Utpaladeva, what finally renders the
• ocess of apoha possible, the potential “openness” of the object (and
• e subject) being the very ground for the “doubt” about it. On this
• emise alone, the Buddhist apoha is accepted by the Śaivas, who take it
• a particular śakti figuring side by side with jñāna-śakti.

• xt5

• . 4v,16] vivr̥tau — yujyeta6 punar etādr̥śo vyavahāro yady eṣām
• ubhavasmr̥tyādīnāṁ cinmātrasāratvenaika evātmā syāt, yasyaitā
• śvātmanaḥ parameśvarasya jñānādikā vicitrāḥ śaktayaḥ,7 anyathā tu na
• ād iti ।।

• ced antaḥkr̥tānanta [f. 5r] viśvarūpo maheśvaraḥ ।

• ād ekaś cidvapur jñānasmr̥tyapohanaśaktimān ।। [1.3.7]

• imarśinī follows] [f. 6r,4] atra vr̥ttiḥ — cittattvam eva viśvarūpam
• o ’tiriktasyānupapatter aśeṣapadārthajñānānām anyonyānusandhānam
• yaiva jñānādiśaktayaḥ, mattaḥ smr̥tir jñānam apohanaṁ ceti hy uktam ।।

• vr̥tiḥ — prakāśarūpaṁ8 hi cittattvaṁ kartr̥tāmayam ādisiddham eva
• datiriktatvaṁ ca nīlasukhāder jaḍābhimatasya bhāva9jātasya svayam
• rakāśarūpatvaṁ syād atadrūpatve ca prakāśamānatānupapattiḥ, na hy
• vetarūpaḥ prakāśaḥ śveta iti bhavaty etac copapādayiṣyate10 । tad āha
• ttattvam evety ato ’tiriktasyeti ca । bhinnāvabhāsinām api cārthānāṁ
• dekarūpatve tadviṣayāṇāṁ jñānānām avivādasiddham eva tadātmakatvam, na
• tāni tato bhedena prathante, tathāpi11 viṣayabhedāvakalpita-12
• edānām api teṣāṁ vibhinnadeśakālāvacchede ’py
• acin-mayasvabhāvaviśrāntir eva । anyathā sarveṣām eva jñānānām
• asvasaṁvedanarūpāntarmukhacittattvamayasvabhāvābhāve
• valapradeśajñānaprakāśasamaye saghaṭapradeśajñānaṁ nāstīti kathaṁ
• [f. 6v]ścīyate, jñānānāṁ svaprakāśarūpatayā
• aṭādiviṣa-yavadekajñānasaṁsargitvābhāvāt ।

• valapradeśe hi ghaṭo yadi syāt, tat tatra pradeśajñāne
• lyayogyatārūpatvāt so ’pi prakāśeta tathā ca saghaṭapradeśajñānam etat
• āt, na kevalapradeśajñānam; ataḥ kevalapradeśajñānaṁ svātmānaṁ
• ghaṭapradeśajñānaviparītaṁ saṁvedayamānaṁ pradeśam api ghaṭarahitaṁ
• ṁvedayate iti ghaṭābhāvasiddhiḥ pradeśe syāt ।

• vatā kevalapradeśajñānena yadi nāma svātmā
• ghaṭapradeśa-jñānaviparītaḥ saṁviditas tato dvitīyam arthāntarabhūtaṁ
• ghaṭa-13 pradeśajñānaṁ svaprakāśarūpaṁ nāstīti kathaṁ sidhyet ।

• aṭo nāstīti kāmaṁ sidhyet, anyathā ghaṭāsadbhāve tasyaiva
• adeśajñānasya saghaṭapradeśajñānatāpattiprasaṅgāt । na ca tat
• ghaṭapradeśajñānam, api tu tadviparītam,14 atas tajjñānasaṁvedyo nāsti
• aṭa iti bhavet ।

• itīyaṁ tu saghaṭapradeśajñānaṁ nāstīti kena jñātaṁ yatas15
• jjñānasaṁvedyaḥ syād api tatra ghaṭaḥ, na ca tatra dvitīyaṁ
• ghaṭapradeśajñānaṁ yadi syāt tan16 na prakāśeta, na ca prakāśate ’to
• stīti vaktuṁ yujyate tad eva hi paryanuyuktam ।

• tham etaj jñānam aghaṭapradeśajñā[f. 7r]ne17 prakāśata iti tu tenaiva
• aprakāśena svātmā tathā jñāyeta, ghaṭe yadvad aghaṭābhāvarūpatā ।

• itīyaṁ tu saghaṭapradeśajñānam arthāntarabhūtam aghaṭa-18
• adeśajñānābhāvam api na na19 vyabhicarati yena tatprakāśād eva
• ghaṭapradeśajñānābhāvo ’pi jñāyeta, dvitīyārthāntarabhūtasaghaṭa-
• adeśajñānabhāvābhāvayor api tasyāghaṭa20
• adeśajñānapra-kāśasyopapatter ekatarā21navyabhicārāt ।

• ha saghaṭapradeśajñānaṁ ghaṭamātrajñānaṁ vā na pradeśajñānād bhinnaṁ
• mbhavaty ekajātīyarūpaviśeṣaviṣayasyāpi cākṣuṣasyaikatvaniyamāt,
• yathā ghaṭe ’pi pratiparamāṇujñāna-bhedāpatter ghaṭābhāsa eva na syād
• yādiyuktibhis tat kevala-pradeśajñānam evedaṁ ghaṭe sati
• aṭapratibhāsam api syāt, na ca tatheti tata eva kathaṁ na
• aṭajñānābhāvasiddhiḥ ।

• d evam iyannyāyānusaraṇād ānumāniky eva ghaṭābhāvasiddhiḥ syāt, na
• ātyakṣī. yadi vā samandhakāre22 rūpajñānaṁ nāstīti kathaṁ jñāyeta,
• rhy atraitad api vaktuṁ yujyeta23 +yadi rūpajñānaṁ syāt tad etat
• arśādijñānābhinnam eva syād+ [my tentative restoration: yadi
• pajñānaṁ tattatsparśādijñānābhinnam eva syād] iti rūpasparśajñānānāṁ
• innendriyajanyānām anekatvenaivopagamāt ।

• py uttarakālotpannavikalpaḥ24 smr̥tir vā sparśa-jñānamātrābhāsā
• pajñānābhāvāvyabhicāriṇī25 tam api sādhayati, yato26 rūpasyāpi tadā
• dbhāve yathā[f. 7v]sau sparśajñānābhāsā tathā rūpajñānābhāsā caiva27
• āt, na caivaṁ tadātmānaṁ sparśajñānamātrābhāsaṁ saṁvedayamānāpi
• pajñānābhāsavivekaṁ sādhayatīti । tato yady api svātmani
• pajñānābhāsaṁ vyavacchindyāt saṁvedyamānākāraviparītatvāt tathāpi
• rvaṁ sparśajñānamātraprakāśakāle rūpajñānaṁ nābhūd iti kathaṁ jānīyāt,
• arśamātrajñānasya tatra parisphurataḥ pūrvoktanyāyena
• itīyarūpajñānābhāvāvyabhicārābhāvāt28 । smārtaṁ hi vijñānaṁ
• rvānubhavasaṁskārajaṁ pūrvānu-bhavānusāreṇaiva vyavasthāpakaṁ bhavati,
• tu svātantryeṇa pūrvapravr̥ttānubhavabhedasākṣātkāri sad yena
• ātmany29 ābhāsamānākāraviparītaṁ rūpajñānaṁ pūrvasmin kāle
• avacchindan viṣayam api pūrvabhāvinaṁ vyavacchindyāt, yathā
• adeśamātrajñānaṁ svātmani saghaṭa-30 pradeśābhāsatāvyavacchedena bahir
• aṭam api vyavacchinattīty āhuḥ । tato vikalpajñānaṁ
• rvānubhavasaṁskārāt tadanusāri jāyamānam, yathā tatpūrvānubhavadvayaṁ
• ̥thakpr̥thakprakāśam ekaprakāśasaṁsargarahitam āsīt, tathānubhūtam31
• ubhavāmīti ca tat parāmr̥śati ।

• taś cobhayaparāmarśe ’pi
• tsvaprakāśasparśarūpajñāna-dvayamātravyavasthānaṁ syāt,32 na tv
• akālikatvena yena tad andhakārakālagatasparśajñānamātraṁ
• tkālikarūpajñānavimuktaṁ svābhāsopārūḍhaṁ [f. 8r] parāmr̥śad api
• dandhakārakāle rūpajñānaṁ nirākuryāt । yāvatā jñānānām
• yonyaprakāśarahitānāṁ svātmamātrapratiṣṭhānāṁ yathā na svayaṁ
• lyakālatāvyavasthāpanaṁ tathā vikalpenāpi tatsaṁskārajanmanā
• arūpamātraprakāśakānekajñāna-33 parāmarśakena naikakālavyavasthāpanaṁ
• kyaṁ pūrvānubhūtātiriktavastusaṁsparśisaṁvedanatvābhāvāt tasya, na
• pi pūrvānubhūtavastusaṁsparśitvam api vikalpānāṁ ghaṭata ity uktam ।
• d evam ekāntarmukhasvasaṁveda- narūpacittattvātmatāviśrāntiṁ vinā
• ānānām aindriya-kāṇām apy anyonyaṁ sahabhāvapr̥thagbhāvaprakāśo na syād
• y arthānām api sadasattāniyamaniścayo na syāt, tataś ca
• rvavyavahārocchedaprasaṅgaḥ । tasmād eṣām ekacittattvaviśrāntirūpam
• usandhānam avaśyābhyupagantavyam । tad āha aśeṣapadārthajñānānām iti
• ānānāṁ bahūnām anusandhānam eva cittattvam ity anena ekaśabdo ’pi
• vr̥taḥ ।

• r̥tyādau34 cānekānubhavānām anusandhānam eka-cittattvaviśrāntir evāta
• ābhedenātra cittattvam anusandhānam iti tasya caivam avibhedena
• kṣyamāṇanayenāśeṣabhāvanirbharasya35 tattadvicitrabhāvāvabhāsanāmayaṁ
• ānaṁ nāma śaktir eva । phalabhedāropitabhedo36 hi padārthātmaiva
• ktir ucyate sā ca śaktiḥ svecchā[f. 8v]vaśenaiva na punar
• yakāraṇasāmarthyasācivyena yasya phalati37 sa maheśvara iti sūtre
• thā nirdiṣṭaḥ । cittattvasya caikarūpatvam agre pratipādayiṣyate ।38
• r̥tir api pūrvāvabhāsitavastunaḥ pūrvāvabhāsitatvena vartamānakāle
• akhyāpanaśaktiḥ । deśakālaviśeṣāvacchedaśūnyatvena
• citrecchābodhitā39vabhāsabhedavāsanāmātropapāditavikal-
• rūpavijñānāpohanaśaktiḥ śaṅkitapratipakṣapratikṣepaparatvād vikalpānām
• padārthāntarasamparkarahite ’py arthe pratiniyataprakr̥tāv avabhāte40
• vidhārthakriyākāriṇy api viśiṣṭaikārthakriyārthina īśvarāṅgabhūta
• ānyonyātmatayā
• ṅkyamānatattadarthakriyākāritattatpratipakṣanirākriyāṁ vinā na tena
• avahartum alaṁ pratipattāra iti svabhāva evaiteṣāṁ41 padārthānām
• yonyabhedamayatvāt ।

• d viśvavyavahāraprajanaiṣā vikalpavyāpārāpohanaśaktir iti gītāsu
• agavatāviṣṭamaheśvarabhāvena parigaṇitā । tad āha mattaḥ smr̥tir iti ।
• ac ca jñānaśaktibhedāt ।।

• anslation

• vr̥ti: On the contrary, such a world of practical experience
• yavahāraḥ) will become possible (yujyeta)42 — if all the various
• gnitions — direct perception, memory, etc. — have only one intrinsic
• ture (ātmā),43 being in essence consciousness only. Of this [ātman],
• mely the Supreme Lord, whose essence is all, these cognitions, etc.,
• present the various and multiform powers.44 [The world of practical
• perience] would otherwise be impossible.

• rikā: “. . . if there were no Maheśvara who contains within Himself
• l the infinite forms, who is one, whose essence is consciousness,
• ssessing the powers of knowledge, memory and exclusion.” [1.3.7]

• ̥tti: The mutual unification of all cognitions of things is
• onstituted by] the consciousness principle itself (cittattvam) whose
• rm is all, since nothing distinct and separate (atiriktasya) from it
• admissible. The powers of knowledge, etc. only pertain to this
• onsciousness-principle]. For it has been said: “From me derive memory,
• owledge and exclusion.”45

• vr̥ti: For the consciousness-principle, whose form is light,
• nstituted by agency, is established since the very beginning. If the
• rld of objects, such as a patch of blue colour or pleasure,
• rroneously] assumed as being insentient, were separate from it, this
• uld [amount to saying] that its nature is not by itself light. But if
• hings] did not have light as their essence, their shining would be
• possible, for, as we are going to demonstrate,46 a light which is not
• ite47 cannot become white. This concept has been expressed [by the
• ̥tti] with the phrases “the consciousness-principle itself” and
• istinct and separate from it”. If the objects, though appearing as
• fferentiated [from consciousness], are to be recognized as constituted
• lely by consciousness, it is established beyond discussion that [also]
• e cognitions which refer to them are made of consciousness.48 For
• gnitions never manifest themselves separately from it; all the same,49
• ough they (i.e. the cognitions) may show [a certain level of]
• fferentiation somehow induced (avakalpita)50 by the differentiation of
• eir contents, yet, despite this delimitation due to spatio-temporal
• fferences, they ultimately do rest (viśrāntir eva)51 on their own
• ture consisting of a single and unitary consciousness (ekacit).52 If
• were not so, and all the various cognitions did not have as their
• sential nature the consciousness-principle, introverted, in the form
• a single and unitary self-awareness (eka-svasaṁvedana), how could
• e, at the moment when the cognition of an empty surface occurs, arrive
• the statement “There is no jar here” [lit.: “This is not the
• gnition of a surface with a jar”]? For cognitions, being by their very
• ture conscious of themselves [alone], cannot combine to form a single
• omplex] cognition (ekajñānasaṁsargitvābhāva)53 having as its object a
• r and so on.54

• uddhist:] In fact, if a jar were in an empty place, then also the
• ar] would be manifested in the cognition of the place since [both the
• r and the place] share the same capacity [of being perceptible]
• ulyayogyatārūpatva),55 and in this way it would be the cognition of a
• ace with a jar, not the cognition of an empty place. Therefore, the
• gnition of an empty place, being aware of itself as opposed to the
• gnition of a place with a jar, is also aware of the place as devoid of
• jar. In this way the absence of a jar in the place can indeed be
• tablished.

• aiva:] But if the cognition of an empty place is aware [only] of
• self as opposed to the cognition of the place with a jar, how would it
• possible to establish that there is not a second56 cognition — which
• another fact-event (arthāntarabhūta) — having the place with a jar as
• s object and being manifest by itself?

• uddhist:] It is none the less true that, some way or other, one should
• able to affirm that a jar is absent, for otherwise it would ensue
• at when there is no jar that very cognition of the place [without a
• r] would be the same as the cognition of the place with a jar.
• wever, that cognition would not be the cognition of the place with a
• r, but a cognition opposed to it;57 hence we would have: “There is no
• r being cognized by that cognition.”

• aiva:] But by which cognitive act is a second (dvitīyam) (i.e.
• stinct and separate)58 cognition of the place with a jar cognized as
• t being there, thanks to which [one could say that] the jar, being
• tentially cognized by that same cognition, should also have been
• ere? If such a distinct and separate cognition were there, it could
• t but be manifested (na ca . . . na prakāśeta), whereas, on the
• ntrary, it is not manifested. Therefore, one cannot reasonably say
• at it exists.59 In fact, this is precisely what has been questioned.

• uddhist:] But to him who might wonder how this cognition can be
• nifested when [only] the cognition of the place without a jar occurs,
• would reply that it is cognized by that (i.e. the latter’s) very
• lf-awareness, just as when there is a jar, its being there has the
• rm of the not-being-there of a non-jar.60

• aiva:] However, the distinct and separate cognition of the place with
• jar, which is “another” fact-event, is not also invariably accompanied
• the absence of the cognition of the place without a jar, so that
• mply from the manifesting of the latter61 one could also cognize the
• sence of the cognition of the place with a jar.62 For, even as regards
• e presence and the absence of a distinct and separate cognition of the
• ace with a jar which is “another” fact/event, there is no invariable
• ncomitance with either [the absence or the presence of the cognition
• the place without the jar], since the cognition of the place without
• jar is manifested [in both cases].63

• uddhist:] But the cognition of the place with the jar or the cognition
• the jar alone cannot occur separately from the cognition of the
• ace, since a visual cognition having particulars belonging to the same
• ass as its objects is by necessity unitary.64 If it were not so, also
• th reference to a jar the manifestation “jar” itself would not take
• ace because one would come to have a single and separate cognition for
• ery single atom.65 Therefore, from these arguments and others of this
• nd it ensues that this cognition of the place alone should, in the
• se of a jar being there, also contain the manifestation of the jar.
• t [in the case at issue] it is not so.66 Due to this very reason, the
• tablishment of the absence of the cognition of a jar cannot be
• estioned.

• aiva:]67 Thus, if we accept your argument as following such a line of
• asoning, the absence of a jar comes to be established by means of
• ference (ānumāniky eva), not by means of direct perception (na
• ātyakṣī).68 If, on the other hand, someone [i.e. a Buddhist] objected
• at on these conditions one could not [even] say when one is in
• ntinuous darkness that there is no cognition of a visible form, [we
• uld reply that] in this case this too69 would be admissible, provided
• at such a cognition of a visible form were not distinct from the
• gnition of a tangible form and so on — whereas [on the contrary] the
• gnitions of tangible and visual forms, being produced by different
• nses, are to be admitted as [differentiated and therefore] not
• educible to] one.70

• r can a conceptual elaboration, which takes place at a later time,71
• memory, in which only the cognition of a tangible form is manifested
• d which is [consequently deemed as] invariably associated with the
• sence of the cognition of a visual form, also establish the absence of
• e cognition of a visual form. For if at that time a visual form had
• so been there, then, just as [now] in this [memory] the cognition of a
• ngible form is manifested, in the same way also the cognition of a
• sual form should be manifested in it.72 In fact, although [memory], in
• e manner [outlined above], is aware of itself — referring to that time
• as containing the manifestation of the mere cognition of a tangible
• rm, it does not [contextually] establish a discrimination (i.e. the
• sence) of the manifestation of a cognition of a visual form.
• erefore, although it may [be admitted to] exclude in itself the
• nifestation of a cognition of a visual form as being opposed to the
• rm being currently cognized, none the less how could it (i.e. memory)
• ow that previously, i.e. at the time when only the cognition of a
• ngible form was manifested, there was no cognition of a visual form?
• r on the basis of the already stated reason, there is no invariable
• ncomitance of the cognition of a mere tangible form, which is taking
• ace there, with the absence of a distinct and separate cognition of a
• sual form. In fact, that [particular kind of] cognition that is memory
• mārtaṁ vijñānam) derives from the impressions left by a previous
• rception [and] performs the task of determining [its object] only by
• nforming to the previous perception, [that is,] not by autonomously
• vātantryeṇa)73 having access to the particular perception that
• curred in the past, so that, while excluding in itself the cognition
• a visual form in the past74 as being opposed to the form currently
• ing manifested, it might also exclude its object [equally] belonging
• the past.75 In the same vein, some [opponents] maintain that by
• cluding the manifestation in itself of the place with a jar the
• gnition of a mere [empty] place also externally excludes a jar. For
• e above-mentioned reason, just as two [distinct] former perceptions
• ch shining by its own light were not associated into a single light,
• the same way a discursive cognition (vikalpajñāna)76 — which derives
• om the impressions left by a former cognition and [consequently]
• nforms to it — is reflectively aware (parāmr̥śati) of them in terms of
• This has been] perceived” and “I am perceiving”.77

• erefore, although we have a reflective awarenesses of both
• ognitions], [none the less we do have] the determination only of those
• o cognitions — that of a tangible form and that of the visible form —
• ken in themselves (-mātra-), each being self-illuminating, but not the
• termination of their occurring at the same time (na tv
• akālikatvena), by virtue of which it (i.e. the memory/discursive
• gnition) — being reflectively aware only of the cognition of the
• ngible form which occurred at the time of darkness, free from any
• gnition of a visual form occurring at the same time and resting in its
• n self-manifestation (svābhāsopārūḍha) — might [be able to] negate the
• gnition of a visual form at the time of that darkness.78

• the contrary, [perceptual] cognitions, being unable to illuminate
• ch other and each being grounded in itself only, cannot by themselves
• tablish their occurrence at the same time. Similarly, the discursive
• gnition, too (vikalpenāpi), which originates from the impressions left
• them and has a reflective awareness of several [distinct] cognitions
• ch illuminating its own form alone, cannot establish their occurrence
• the same time, for it is not in the nature of [discursive cognition]
• be an awareness directly touching a distinct real entity which has
• rmerly been perceived. Nor, on the other hand, has it ever been said
• at the discursive cognitions are admitted to directly touch (i.e. have
• rect access to) the real entity formerly perceived. On these premises,
• gnitions, even the perceptual ones (aindriyakāṇām api),79 could not be
• nifested as mutually associated or separated, unless they rest on
• ving as their essential nature the consciousness-principle in the form
• a single and unitary introverted self-awareness. Consequently, one
• uld not determine for things, too, a restriction of their being
• utually] present or absent, whence the destruction of the world of
• actical experience would ensue. Therefore, we are obliged to admit for
• gnitions [some] unification which has the form of their resting on a
• ngle consciousness-principle. This [is what is meant by the Vr̥tti]
• th the phrase “of all cognitions of things”. The
• nsciousness-principle is that very unification (anusandhāna) of the
• ny cognitions: in this way, also the word “one” in the Kārikā (cf.
• acidvapuḥ) has been explained.

• so the unification of many perceptual cognitions in memory, etc. is
• thing but their resting on the single consciousness-principle; it is
• r this very reason that “consciousness-principle” and “unification”
• e put here [in the Vr̥tti] in a relationship of non-difference. And
• us, on the basis of the reasoning that will be stated below, to this
• nsciousness-principle, full of all things, there belongs a power, in
• rms of non-differentiation [from it], named “knowledge”, consisting in
• e manifestation (-avabhāsanā)80 of the various multiform things. In
• ct, when something which is in itself a [unitary] substance
• adārthātmaiva) is superimposed with a differentiation deriving from
• s different effects, we call it “power”.81 And by the word “Great
• rd” (maheśvara iti) the sūtra indicates him whose “power” yields fruit
• e to his will alone, i.e. without the intermediation of the capacity
• other causes. And the fact that the consciousness-principle is to be
• e and unitary in its nature will be explained later on.82 Memory, for
• s part, is the power to make an entity, which has been manifested in
• e past, known in the present as formerly manifested. We have the power
• exclusion (apohana-śakti) when a cognition in the form of conceptual
• aboration is generated exclusively by the latent impression left by a
• rticular manifestation, [a latent impression which is] awakened by the
• nderful (vicitra)83 will [of the Lord] as devoid of any specific
• atio-temporal delimitation. For these (i.e. “this kind of”) conceptual
• aborations have as their [specific] function that of rejecting
• visaged contrary options (śaṅkita- pratipakṣa).84 Also in respect to
• object being manifested without commingling with other realities
• nd] having a well-defined nature of its own85 — but at same time being
• pable of having a variety of possible causal efficiencies — the
• owing subjects aspiring to only one of these causal efficiencies — a
• ecific one — will not succeed in making use of such an object in the
• rious occurrences of practical life, unless we reject the various
• ntrary options, endowed with various causal efficiencies, which they
• e figuring out.86 For the object, which is not separated from the
• vel of īśvara, as it were, has as its essence that of all other
• jects.87 The above is the very essential nature of these88 realities
• they are characterized by mutual differentiation.89

• erefore, it is precisely this power of exclusion, whose sphere of
• tivity is represented by conceptual elaboration, that opens the way to
• e transactions of practical life.90 On account of this, [the power of
• clusion] has been mentioned separately in the Songs by the Blessed
• e, who possesses the status of the Great Lord. In fact, He said: “From
• memory [. . .]”.91 [All] these [powers derive] from a differentiation
• the power of Knowledge.

• Originally published in K. Preisendanz (ed.), Expanding and Merging
• rizons: Contributions to South Asian and Cross-Cultural Studies in
• mmemoration of Wilhelm Halbfass, Vienna 2007, pp. 473-90.

• * I wish to thank Prof. Karin Preisendanz for her useful comments and
• en editorial care.

• ferences

• xts

• G = Bhagavad-Gītā, in BhGŚ.

• GAS = Bhagavad-Gītārtha-Saṁgraha of Abhinavagupta, in
• īmadbhagavad-Gītā with Gītārtha-Saṁgraha, part I: Text, ed. S.
• nkaranarayanan, Oriental Research Institute, Tirupati, 1985.

• agavad-Gītā-Bhāṣya, in Śrīmadbhagavad-Gītā with the Commentaries
• imat-Śāṅkarabhāṣya (with Ānandagiri’s Vyākhyā), Nīlakaṇṭhī,
• āṣyotkarṣa-Dīpikā of Dhanapati, Śrīdharī, Gītārtha-Saṁgraha of
• hinavaguptācārya, Gūḍhārtha-Dīpikā of Madhusūdana (with
• ḍhārtha-Tattvāloka of Śrīdharmadattaśarma), ed. L.S. Panshikar, 2nd
• n., Motilal Banarsidass, Delhi, 1978.

• GS = Sarvatobhadra, in The Bhagavad-Gītā with the Commentary Called
• rvatobhadra of Rāmakaṇṭha, KSTS 64, ed. M.K. Shastri, Bombay 1943.

• P = Dharmottara-pradīpa of Durveka Miśra, in Dharmottara-Pradīpa
• eing a Sub-commentary on Dharmottara’s Nyāyabindu-Ṭīkā, a Commentary
• Dharmakīrti’s Nyāyabindu], ed. Pandita Dalsukhbhai Malvania, rev. 2nd
• n., Patna 1971.

• = Dharmakīrti, Hetubindu, ed. in Steinkellner 1967.

• K = Īśvarapratyabhijñā-Kārikā and -Vr̥tti of Utpaladeva, ed. and tr. in
• rella 2002.

• V = Īśvarapratyabhijñā-Vimarśinī of Abhinavagupta, ed. Mukund Ram
• astri, 2 vols., KSTS 22 and 33, Bombay 1918-21.

• VV = Īśvarapratyabhijñā-Vivr̥ti-Vimarśinī of Abhinavagupta, ed.
• dhusudan Kaul Shastri, 3 vols., KSTS 60, 62 and 65, Bombay 1938-43.

• -Ṭ = Nyāyabindu-Ṭīkā of Dharmottara in DhP.

• = Pāṇini’s Aṣṭādhyāyī, ed. in Sharma 1995.

• = Pramāṇa-Vārttika of Dharmakīrti, in Pramāṇa-Vārttika with the
• mmentary “Vr̥tti” of Acharya Manorathanandin, critically ed. Swami
• arikadas Shastri, Varanasi, 1968.

• SV = Dharmakīrti’s Svavr̥tti on Pramāṇa-Vārttika I, ed. in Gnoli 1960.

• = Vākyapadīya of Bhartr̥hari, in Vākyapadīya with the
• akīrṇakaprakāśa of Helārāja: Kāṇḍa 3, ed. K.A.S. Iyer, 2 vols., Poona,
• 63-73.

• kyapadīya, ed. in W. Rau, Bhartr̥haris Vākyapadīya: Die Mūlakārikās
• ch den Handschriften, Abhandlungen für die Kunde des Morgenlandes 42,
• esbaden, 1977.

• h = Vijñāna-Bhairava-Tantra with the Commentary partly by Kṣemarāja
• d partly by Śivopādhyāya, ed. Mukunda Ram Shastri, Kashi Sanskrit Text
• ries 8, Bombay 1918.

• udies and translations

• hnemann, G., 1980, “Identifizierung von Sanskrittexten
• ṅkaranandanas”, Wiener Zeitschrift für die Kunde Südasiens, 24:
• 1-98.

• oli, R., 1960, Pramāṇavārttikam: Chapter One with the Autocommentary
• erie Orientale Roma 23), Roma, 1960.

• ata, T., 1993, Prasaṅga und Prasaṅgaviparyaya bei Dharmakīrti und
• inen Kommentatoren (Wiener Studien zur Tibetologie und Buddhismuskunde
• ), Wien.

• asser, H., 2001, “On the Date and Works of Śaṅkaranandana”, in R.
• rella (ed.) Le Parole e i Marmi: Studi in Onore di Raniero Gnoli nel
• o 70° compleanno (Serie Orientale Roma 92), pp. 489-508.

• ndey, K.C., 1963, Abhinavagupta: An Historical and Philosophical
• udy, 2nd edn., Varanasi.

• arma, R.N. (ed. and tr.), 1995, The Aṣṭādhyāyī of Pāṇini, vol. 3,
• lhi.

• einkellner, E., 1967, Dharmakīrti’s Hetubindu: Teil I: Tibetischer
• xt und rekonstruierter Sanskrit-Text; Teil II: Übersetzung und
• merkungen, Wien.

• einkellner, E. and M.T. Much, 1995, Texte der erkenntnistheoretischen
• hule des Buddhismus (Systematische Übersicht über die buddhistische
• nskrit-Literatur 2), Göttingen. (Abhandlungen der Akademie der
• ssenschaften in Göttingen, Philologisch- Historische Klasse, Folge
• 214).

• rella, R., 1988, “A Fragment of Utpaladeva’s
• varapratyabhijñā-Vivr̥ti”, East and West, 38: 137-74.

• —, 2002, The Īśvarapratyabhijñā-Kārikā of Utpaladeva with the Author’s
• ̥tti: Critical Edition and Annotated Translation, Motilal Banarsidass,
• lhi (I edn. Serie Orientale Roma 71, IsMEO, Roma, 1994), repr. 2014.

• —, 2011, The Philosophical Traditions of India: An Appraisal,
• ranasi: Indica Books.

• ----------------------------------------------------------------------

• Cf. Torella 2002: xl-xliv.

• “Leaf 33. Size 7⅟2 x 9⅟2”. Lines per page 19 approximately. Syll. per
• ne 21. Handwriting almost recent. Written on local paper which does
• t seem to be old. Thirty-first leaf worm eaten, it is clear from the
• tebook pages 44-46. The ms begins line no. 1, top: śrīgaṇeśāya namaḥ;
• ne 2: saṁskāre saty api, etc. Ends abruptly as can be seen at page 49.
• aracter Śāradā”. This description perfectly suits the National
• chives manuscript, except for the mention of the worm-eaten
• irty-first leaf, which, at least as far as I can see from a rather
• int xerox-copy, would seem more or less intact. According to Prakash
• ndey (personal communication), the manuscript belongs to the latter
• lf of the eighteenth century.

• The technique of prasaṅga-viparyaya is also well known to Buddhists
• f. Iwata 1993).

• The ĪPVV seems to be the only work to quote from the Apohasiddhi;
• ong the five quotations (pp. 272, l. 14; 292, ll. 19, 21, 23; 293, l.
• ) the second one is certainly in prose, thus showing that the
• ohasiddhi, just like other works by Śaṅkaranandana, consisted of
• rikās mixed with prose passages. Only three out of the five quotations
• uld be traced either in its single extant Sanskrit manuscript
• ühnemann 1980: 193-94) or in the Tibetan translation (see Gnoli 1960:
• iii-xxiv); on Śaṅkaranandana, see Steinkellner and Much 1995 and
• asser 2001. (For a new interpretation of Śaṅkarānandana’s religious
• filiation see Torella 2011: 19 n.12) In a separate study under
• eparation I propose to examine the context of the Apohasiddhi from
• ich the above passages have been drawn. I cordially thank Ernst
• einkellner and Helmut Krasser for very kindly and generously putting
• my disposal all the available material on Śaṅkaranandana’s
• ohasiddhi, including a copy of the only available Sanskrit manuscript
• roduced from Rahula Sankrityayana’s negatives) and a transcript of the
• betan translation.

• The words underlined in the text are those literally cited in the
• VV. Punctuation is mine (that found in the ms is often misleading).
• e sandhi has been “normalized”.

• Ms yujyate. I follow Abhinavagupta’s reading (ĪPVV I: 273, ll. 6-7):
• d etat tātparyam avataraṇikāyām āha “yujyeta” iti ।

• At this point the ĪPVV seems to refer to other parts of the Vivr̥ti
• ich are not found in the ms (ĪPVV I: 273, ll. 15-16: śaktiyogalakṣaṇaṁ
• ti).

• prakāśarūpaṁ, my emendation (ms viśvarūpaṁ), following Abhinavagupta’s
• ading (cited and commented upon in ĪPVV I: 273, ll. 18-21) which seems
• re appropriate to the context.

• -sya bhāva-, my emendation (ms svabhāva-); cf. ĪPVV I: 273, l. 25.

• upapādayiṣyate, my emendation (ms pratipādayiṣyatāṁ); cf. ĪPVV I:
• 4, ll. 17-18: upapādayiṣyata iti prakāśātmā prakāśyo ’rtha
• yādisthāneṣu.

• tathāpi, my emendation (ms tathā hi); I follow the reading cited and
• mmented upon in ĪPVV I: 275, ll. 5-6 (s. n. 47 to the translation).

• avakalpita-, my emendation (ms -vikalpita-). Cf. ĪPVV I: 275, ll.
• 8: bhedas tenāvakalpito ’vajñayāprarohayogena kalpitaḥ.

• saghaṭa-, my emendation (ms ghaṭa-).

• ĪPVV I: 277, l. 17 seems to have one more quotation from the Vivr̥ti
• haṭam api) which is not found in the ms.

• Ms jñātam atas is to be corrected to jñātaṁ yatas. This was probably
• e reading of the exemplar, as we can surmise from the anusvāra above
• a- (subsequently deleted); the following ya- must have been misread as
• -, thus suggesting the deletion of the anusvāra.

• I correct ms tat to tan na, as meaning strictly requires.

• aghaṭapradeśajñāne, my emendation (ms tatsaghaṭapradeśajñānaṁ). This
• ems to be the reading required by the context. This would also tally
• th Abhinavagupta’s remark (ĪPVV I: 278, ll. 2-3): nanv
• haṭapradeśajñānaṁ prakāśata ity atredam eva tattvaṁ
• ghaṭapradeśajñānaṁ na prakāśata iti.

• aghaṭa-, my emendation (ms ghaṭa-), supported by Abhinavagupta’s
• otation and comments (ĪPVV I: 278, ll. 3-5: naitad ity āha “aghaṭa”
• i paryudāsavr̥ttyā hy atra vastvantaraprakāśo bhāti, na tu
• asajyapratiṣedhātmā saghaṭapradeśajñānaprakāśābhavaḥ).

• na na, my emendation (ms na).

• tasyāghaṭa-, my emendation (ms tasya ghaṭa-), supported by
• hinavagupta’s paraphrase (ĪPVV I: 278, ll. 13-14):
• valapradeśapratyayasyobhayathāpy upapatteḥ.

• -anavyabhicārāt, my emendation. This (or also -vyabhicārāt) is in
• ct the reading one would expect, instead of ms -avyabhicārāt
• katarāvyabhicārāt); see also Abhinavagupta’s comments ĪPVV I: 278, l.
• : avyabhicāro ’vinābhāvo nāsti.

• The original reading of the ms was sambandhakāre, then transformed
• to samandhakāre, which is in fact the correct reading (cf. ĪPVV I:
• 0, l. 23: santate ’ndhakāre).

• yujyeta, my emendation (ms yujyate).

• In the ms -vikalpa- has been later corrected to the correct
• ikalpaḥ.

• -avyabhicāriṇī, my emendation (ms -vyabhicāriṇī); cf. ĪPVV I: 281,
• . 18-19: [. . .] rūpajñānasyābhāvam avyabhicāreṇa bhavantaṁ sādhayet.

• yato, absent in the ms but cited by Abhinavagupta, apparently has to
• added here. The following phrases up to sādhayatīti are taken by him
• a quotation (from an unknown source). Cf. ĪPVV I: 281, ll. 20-21:
• tra hetutvaṁ granthāntarasyāha “yataḥ” iti.

• caiva, my emendation (the ms has cai-, corrected to nai- above the
• ne).

• -abhāvāt, my emendation (ms -bhāvāt). Utpaladeva is here stating the
• iva position on the matter which in fact excludes any correlation
• tween the two cognitions.

• The expected meaning (with the further support of Abhinavagupta’s
• mments in ĪPVV I: 282, ll. 10- 15) has induced me to modify at some
• aces the transmitted text which is evidently corrupt. In fact, the ms
• ads: ābhāsamānākāraviparītaṁ paricchindan pūrvasmin kāle viṣayam api
• pajñānaṁ pūrvabhāvi vyavacchindyāt. In particular, Abhinavagupta’s
• dātmyena asahamānā (ibid.: l. 12) can by no means refer to the
• ansmitted paricchindan.

• saghaṭa-, my emendation (ms ghaṭa-).

• The ms has an anusvāra deleted above -ta-.

• -vyavasthānaṁ syāt, my emendation (ms -vyavasthā na syān).

• By reading -prakāśakānekajñāna-, I restore what seems to have been
• e original reading of ms, which was corrected to -prakāśakāle jñāna-
• y changing -ne- to -le- and deleting -ka- by means of a horizontal
• roke).

• smr̥tyādau, my emendation (ms smr̥tyātmā); cf. ĪPVV I: 285, l. 6.

• The ms has nirbharasya, corrected to nirbhara-. The correct reading
• the original nirbharasya (cf. the paraphrase ĪPVV I: 285, ll. 21-25:
• tābhir yuktibhir ekasya [. . .] ityādivakṣyamāṇābhiḥ
• mastabhāvapūrṇasya jñānaṁ nāma śaktiḥ).

• phalabhedāropitabhedo, my emendation (ms phalabhedāropitadbhedo). The
• ssage is also quoted, with a slightly different, equally possible
• ading in Kṣemarāja’s commentary on the Vijñāna-Bhairava-Tantra (VBh)
• . 14, ll. 3-5): uktaṁ ca pratyabhijñāyāṁ “phalabhedād āropitabhedaḥ
• dārthātmā śaktiḥ”.

• The ms has pratiphalati, apparently corrected to phalati; the latter
• ading, which I have accepted, is also confirmed by ĪPVV I: 289, ll.
• 7: “phalati” phalaparyantā saṁpadyate, na punaḥ pratihanyate yasya.

• The following definitions of smr̥ti and apohana are incorporated in a
• rginal note found in one of the mss of Abhinavagupta’s ĪPV used for
• e Kashmir Series edition (Delhi National Archives, Sanskrit ms No. 5,
• 131, right margin): deśakālaviśeṣā- vacchedaśūnyatvena
• citrecchāprabodhitāvabhāsabhedavāsanā-
• tropapāditavikalparūpavijñānāpohanaśaktiḥ । pūrvā-vabhāsitavastunaḥ
• rvāvabhāsitatvena vartamānakāle prakhyā-panaśaktiḥ smr̥tiśaktir ity
• ayor bhedaḥ. Note 71 of the ĪPV edition (vol. I, p. 117) is based on
• is marginal annotation.

• The text as quoted in the ĪPV manuscript (see above n 38) has
• rabodhitā-.

• avabhāte, my emendation (ms avabhāti). The latter reading would not
• itself be impossible; Abhinavagupta’s paraphrase and comments,
• wever, consistently have avabhāte).

• If we rely upon ĪPVV I: 295, l. 25 (atrāha “māyīyānām” iti), we
• ould add māyīyānām (omitted in the ms) after eteṣām. However, this
• uld involve a slight redundancy, which is not usual in Utpaladeva’s
• yle. I prefer to surmise that eteṣām was omitted in the edited text of
• VV, leaving merely māyīyānām which glossed it.

• According to Abhinavagupta, here the meaning of the optative conforms
• Pāṇini’s Aṣṭādhyāyī (P) III.3.156: hetuhetumator liṅ, “Affix LIṄ
• tionally occurs after a verbal root when hetu ‘cause’ and hetumat
• ffect’ are denotated” (tr. Sharma 1995: 574). The fact that cognitions
• are the same essential nature is precisely what renders vyavahāra
• ssible.

• Cf. ĪPVV I: 273, l. 12: “ātmā” iti svabhāvaḥ.

• This means that they rest on the Lord, and are neither independent of
• m nor exceeding His own form (ĪPVV I: 273, ll. 14-15).

• See below n 91.

• See, e.g. ĪPK 1.5.3.

• That is, a cognition in which the colour white does not appear.

• For Abhinavagupta (ĪPVV I: 274, l. 25) this is a case of
• ṇḍāpūpīyanyāya, “the maxim of the stick and cakes”. A stick and cakes
• e usually tied together so that if something happens to the stick
• .g. being eaten by rats) the same happens to the cakes.

• Despite any “superficial” differentiation deriving from their
• ntents, cognitions too must have (at least) the same status of
• timate non-differentiation from consciousness as their objects (ĪPVV
• 275, ll. 5-7: “tathāpi” iti tenāpy ahamaṁśāvyatirekeṇa vartamānānām
• y arthaḥ).

• The preverb ava- is given a derogatory meaning (avajñayā) by
• hinavagupta: the objects are only able to shed an imperfect, i.e. not
• lly developed (aprarohayogena), appearance of differentiation onto
• eir cognition (cf. n 12 to the text).

• eva means here “restriction”. As the Vivr̥ti sets out to show, this
• sting of the various cognitions on one unitary consciousness principle
• the necessary condition for the worldly transactions to take place
• ccessfully (ĪPVV I: 275, ll. 14-17).

• In this context, eka (qualifying cit, jñāna, etc.) always oscillates
• tween “one, single” and “unitary”, usually expressing both meanings at
• e same time.

• The Buddhist position Utpaladeva is referring to is clearly
• rmulated in Dharmottara’s Nyāyabindu-Ṭīkā (NBṬ): 101, ll. 13-14:

• daikajñānasaṁsargivastvantaropalambhaḥ । ekendriya-jñānagrāhyaṁ
• canādipraṇidhānābhimukhaṁ vastudvayam anyonyāpekṣam ekajñānasaṁsargi
• thyate ।।

• Utpaladeva’s criticism is addressed to the Buddhists, who, just like
• e Śaivas, uphold the self-awareness of cognitions but, unlike the
• ivas, reject a common ground for them, represented by the unitary
• nscious subject. Interestingly, in order to present an exemplary
• atement on the self-awareness of cognition, Abhinavagupta resorts to
• artr̥hari’s Vākyapadīya, by citing a well-known verse from the
• tisamuddeśa (VP III.1.104):

• thā jyotiḥ prakāśena nānyenābhiprakāśyate ।

• ānarūpaṁ tathā jñāne nānyatrābhiprakāśyate ।।

• ncidentally, it is to be noted that pādas c and d differ considerably
• om the text in Rau’s and Iyer’s editions, which both read: jñānākāras
• thānyena na jñānenopagr̥hyate.) The weakness of the Buddhist position
• the matter appears quite clearly in their treatment of the crucial
• oblem of anupalabdhi (non-perception) and abhāva (non-being, absence),
• ich therefore becomes the object of a sustained analysis in the
• bsequent portion of the Vivr̥ti (ĪPVV I: 276, l. 24 to 277, l. 1:
• yavyavacchedair grahaṇavyavahāro ’pi svaparaviṣayo ya ucyate, so ’pi
• nirvahed ityāśayena saugatīyam abhāvasiddhiprakāram eva vicārayati).

• Cf. Hetubindu (HB), p. 22*, ll. 2-4 (Steinkellner 1967: Teil I, p.
• ): yatra yasminn upalabhyamāne niyamena yasyopalabdhir bhavati
• gyatāyā aviśeṣāt, sa tatsaṁsr̥ṣṭaḥ, ekajñānasaṁsargāt. See also the
• ntences immediately following the NBṬ passage quoted above (n 50):
• yor hi sator naikaniyatā bhavati pratipattiḥ । yogyatāyā dvayor apy
• iśiṣṭatvāt (NBṬ, pp. 101, l. 14 to 102, l. 1).

• As to the meaning of dvitīya (lit. “second”), see below n 58.

• Cf. ĪPVV I: 277, ll. 18-19: saghaṭapradeśajñānaṁ tad bhavet, na tv
• adviparītāghaṭapradeśajñānam.

• ĪPVV I: 277, l. 19: “dvitīyaṁ tv” iti vyatiriktam. In other words,
• e perception of an empty surface cannot evoke the possible perception
• a surface with a jar, to be subsequently negated. This is apparently
• at Abhinavagupta means to say with the laconic remark (ibid., lines
• . 19-21): na hy aghaṭapradeśajñānena jñānāntarasya bhāvo jñāyate, yena
• dabhāvo ’pi jñāyeta.

• Ibid.: 277, ll. 24-25: iti ca vaktuṁ na yujyata iti saṁbandhaḥ.

• According to the Buddhists, we do not need a separate cognition to
• count for the cognition of the absence of a non-jar when a jar is
• gnized: it is the self-awareness of the cognition “jar” which
• plicitly also cognizes the negation of its contrary. The jar itself
• s as its own form the not being a non-jar.

• That is, of the cognition of the place without the jar, i.e. an empty
• ace.

• In other words, the negation of the cognition of an empty space in
• e Buddhist argument is not to be taken in terms of prasajyapratiṣedha
• t of paryudāsa: it does not confine itself to negating a thing but
• firms the presence of another (see n 18 to the text).

• That is, of the space taken in itself, independently of the jar’s
• ing or not being there; cf. Abhinavagupta’s paraphrase (ĪPVV I: 278,
• . 13-14): kevalapradeśapratyayasyobhayathāpy upapatteḥ.

• Abhinavagupta develops the Buddhist’s arguments outlined by
• paladeva and casts them into full-fledged inferences based on various
• rms of anupalabdhi (cf. ĪPVV I: 278, ll. 14ff.).

• Cf. ĪPVV I: 278 l. 22 to 279 l. 1: kiṁ ca kevalapradeśo ghaṭo
• nantarūpaparamāṇumayaḥ saṁnihitānantāvayavo vā । tatra yāvanti
• avyāṇi, tāvanti cet pr̥thag jñānāni; tarhi teṣāṁ pr̥thak pr̥thak
• aprakāśatvenaikasamūharūpaprakāśābhāvāt pradeśajñānam eva na bhavet,
• āti ca tat pradeśajñānam iti.

• That is, when one is perceiving an empty space.

• From this point onwards it is no longer feasible to present the
• rger structure of the argumentation by demarcating the Śaiva and the
• ddhist positions, because the latter comes to be too closely embedded
• the former.

• The Buddhist explanation is seen as not conforming to common sense,
• ich explains the matter under discussion in a much simpler way (ĪPVV
• 280, l. 8: “evam” iti svakalpanayā lokapratītyananugr̥hītayā).
• reover, the Buddhist argument would lead to a regressus ad infinitum
• bid.: iyān ity anavasthāpādako nyāyaḥ).

• armakīrti himself was well aware of the risk involved in not
• counting for anupalambha directly, i.e. by explaining it within the
• ner dynamics of the self-awareness of a perceptive act: the absence of
• object would be arrived at by means of its non-perception, but the
• sence of its perception would need another anupalambha to be
• tablished, which would lead to a regressus ad infinitum.
• amāṇa-Vārttika [PV] IV.274-75:

• smād anupalambho ’yaṁ svayaṁ pratyakṣato* gataḥ ।

• amātravr̥tter gamakas tadabhāvavyavasthiteḥ ।।

• yathārthasya nāstitvaṁ gamyate ’nupalambhataḥ ।

• alambhasya nāstitvam anyenety anavasthitiḥ ।।).

• anorathanandin reads svasaṁvedanāt.

• rses 270-75 are quoted by Abhinavagupta in ĪPVV I: 279 l. 24 to 280 l.
• The same kind of objection is put by Dharmakīrti into the mouth of an
• ponent in his Svavr̥tti on Pramāṇa-Vārttika I.3 (PVSV: 4, ll. 15-17:
• rānupalabdher liṅgād asattāyām upalabdher abhāvo ’py anyayānupalabdhyā
• dhya ity anavasthānād apratipattiḥ syāt). The Buddhist explanation
• self would become acceptable (instead of coming to a dead end) if it
• re placed in a Śaiva perspective (ĪPVV I: 280, l. 13: yad bhavadbhir
• tam, tad upapadyate . . .). For the Śaivas, in fact, a single
• troverted self-consciousness is the common ground which unifies all
• gnitions (ibid. l. 10: asmaddr̥śā jñānānāṁ svasaṁvedanam antarmukham
• am eveti).

• Cf. ĪPVV I: 280, ll. 24-25: “etad api” ity aprātītikaṁ niṣphalaṁ ca
• d apīty apiśabdārthaḥ.

• If two things (e.g. a jar and a cloth) are the object of the same
• nse, they are cognized by the same cognitive act. But this is not the
• se with a visual and a tangible form. In the negative case it ensues
• at in the dark the absence of a tangible form cannot be understood as
• e absence of a visual form, since there is no single cognition in
• ich both a visual and a tangible form are manifested. Cf. ĪPVV I: 281,
• . 1-4: abhinnendriyajanyānāṁ hi ghaṭapaṭādīnām iva saṁnihitānām eṣa
• yamaḥ — ekajñānagrāhyateti । na ca rūpasparśau rūparasau vā
• mānendriyagrāhyau, yena tadviṣayam ekam eva jñānaṁ rūpasparśābhāsaṁ
• ṁbhāvyeta rūparasābhāsaṁ vā.

• If I understand correctly, this remark is meant to exclude the
• kalpa which immediately follows direct perception and transforms it
• to a determinate cognition.

• This inference would have as its hetu “the non-perception of the
• fect” (kāryānupalabdhi), whose validity is viewed by Abhinavagupta as
• ghly questionable already in itself, i.e. prior to his examination of
• e relevant arguments.

• Namely, through an illuminating power of its own (cf. ĪPVV I: 282, l.
• svātantryeṇa svena prakāśabalena). What memory can do is only to
• call the illuminating power belonging to the original perception, or,
• other words, it conforms to it or “imitates” it (ibid.: ll. 15-16:
• r̥tir hi pūrvam anubhavam anukaroti); however, this perception being
• mitated”, Abhinavagupta goes on, is to be related to the object of
• mory (ibid., l. 16: paraṁ smartavyasaṁbandhinam).

• That is, a cognition which had been hypothesized and then negated in
• e past.

• This crucial passage has been carefully examined and explained by
• hinavagupta (ĪPVV I: 282, ll. 10-15). Memory, while cognizing the
• evious cognition of a tangible form, contextually realizes that in the
• nge of its own cognitive act the cognition of a certain visual form,
• posed to that of the tangible form, would not be admissible. In this
• y, memory would be allowed to establish now the previous absence of
• e cognition of a visual form (and, through it, also the previous
• sence of that visual form) on the basis of the fact that if the
• gnition of a visual form had taken place it would have entered the
• nge of the memory cognition along with the cognition of the tangible
• rm. This argument, however, is based on a presupposition that is
• together unacceptable to the Śaivas for whom a cognition cannot
• ssibly become the object of another cognition. In the case at issue,
• e former cognition cannot become the “object” (viṣaya) of memory
• bid.: ll. 7-8: na tu viṣayatvenābhāsayati jñānānāṁ parasaṁvedyatvasya
• starato nirākr̥tatvāt).

• According to Abhinavagupta, this is a general designation also
• cluding memory (ĪPVV I: 282, ll. 24-25: “vikalpajñānam” iti
• mānyoktiḥ smr̥tim api saṁgr̥hṇīte). Abhinavagupta (ibid.: 283, ll. 1-4)
• so records a different interpretation by “others” who take vikalpa as
• particular form of memory, the only difference being that memory
• oper refers to a perception which occurred in the past and is now
• tinct (apart from the impressions left by it).

• That is, memory and discursive cognition respectively.

• Cf. the construction of the sentence by Abhinavagupta (ĪPVV I: 283,
• . 15-16): rūpajñānaviviktasvābhāsopārūḍhasparśajñānaparāmarśi-
• kalpajñānam iti sambandhaḥ.

• Utpaladeva is addressing again the Buddhist opponent: no single
• gnition, particularly not even the perceptual cognition, can establish
• e absence of something by itself alone. As a consequence, either the
• rception has to enter by itself into a net of relations to other
• gnitions (which goes against the Buddhist presuppositions) or the
• ddhist has to admit that absence can only be established by means of
• ference (which the Buddhist intended to reserve for the establishment
• the abhāvavyavahāra alone). But also the establishment of the
• hāvavyavahāra by means of inference would entail a regressus ad
• finitum (cf. ĪPVV I: 284, l. 25 to 285, l. 1: pradeśakaivalyavyavahāro
• y ānumānikatvād anavasthopahata iti).

• The “causing things to appear [as ‘this’]” concerns the limited
• bject (ĪPVV I: 286, l. 3: idam ity avabhāsamānā [read: avabhāsanā]
• śupramātr̥vartinī). The samarthācaraṇa component that Abhinavagupta
• ngles out in the causative form (“to make somebody/something able to .
• .”) is referred to the freedom of the Lord. The manifestation of the
• ject as external to consciousness (“this”) does not entail a real
• paration from the absolute light of consciousness (ibid., ll. 10-11:
• mahāprakāśāt pr̥thakkaroti), since it presupposes the contextual
• reation” of a māyic subject, from whose limited light alone the object
• pears as separate.

• According to Abhinavagupta (cf. ĪPVV I: 287, ll. 3-9), a single thing
• being, capable of producing various effects, does not lend itself to
• indicated, or taught successfully, through a single word or concept.
• actical reality and linguistic communication (vyavahāra) require the
• itary thing to be expressed in manifold ways on the basis of its
• rious efficiencies: one and the same fire can be indicated as
• urner”, “cooker” or “illuminator”. When the manifoldness of aspects is
• rne in mind, we speak of “power(s)”; when unity is borne in mind, we
• eak of “possessor of powers”. However, none of these two approaches,
• hinavagupta continues, can be claimed to be ultimately true, because
• th of them are a creation of the Supreme Lord (ibid.: ll. 8-9:
• rameśvareṇaiva tathā sr̥ṣṭatvāt). In their conception of śakti
• paladeva and Abhinavagupta are reminiscent of what Bhartr̥hari says
• out the many śaktis that any substance possesses, being the basis of
• e grammatical notion of sādhana; see particularly VP III.7.2:

• ktimātrasamūhasya viśvasyānekadharmaṇaḥ ।

• rvadā sarvathā bhāvāt kvacit kiñcid vivakṣyate ।।

• . Helarāja’s commentary thereon (VP: 232, ll. 4-5): ghaṭādayo bhāvā
• śvaśabdavācyāḥ । te ca tattadudakāharaṇādikāryasādhikānāṁ śaktīnāṁ
• mūharūpāḥ); see also kārikā 6.

• According to Abhinavagupta, Utpaladeva here refers to ĪPK I.6.11.

• What puzzles a hypothetical opponent in the “exclusion” process
• pohana) outlined here is the fact that the conceptual elaboration
• ises only from the impression left by the former perception of the
• ject, and yet does not possess the same character of spatio-temporal
• finiteness that belonged to its source (in fact, the conceptual
• aboration is in essence something general). The objection is made
• plicit in ĪPVV I: 291, ll. 8-9: nanu deśādyavacchinne ’nubhūte tatkr̥tā
• sanā kathaṁ tadanavacchinnaviṣayā [my correction; the edited text is
• parently corrupt: tatkr̥tayā vāsanā kathaṁ tadavacchinnaviṣayayā]
• abudhyate. According to Utpaladeva, this inexplicable fact is due to
• e unfathomable (ibid.: vicitrāparyanuyojyā) will of the Lord.

• According to Abhinavagupta (ĪPVV I: 291, ll. 13ff.), the peculiarity
• apohana-śakti with respect to the mere jñāna-śakti lies precisely in
• is figuring out of possible alternatives and then excluding them. In
• armakīrti’s formulation of the apoha theory, the alternatives to be
• jected are in fact various erroneous superimpositions (adhyāropa).
• wever, Abhinavagupta replies (ibid.: 292, ll. 15-17), the object of
• e superimposition is already a mental construct, which in turn had to
• established by way of the exclusion of other superimpositions, and so
• ; this entails that the simple ascertainment of a patch of blue colour
• uld not be arrived at even in thousands of cosmic eras
• ikalpasamāropaparamparānavasthānād yugakoṭibhir api na nīlaniścayaḥ).
• xt, Abhinavagupta takes into account the position of Śaṅkaranandana
• bid., ll. 18ff.), who in the Apohasiddhi substitutes anyathā-prasaṅga
• he mere possibility that a thing be different from itself? Cf. ll.
• -25: tatrāpi viśiṣṭe ’rthe pratibhāte ko ’nyathāprasaṅgasyāvasara iti
• ktavyam) for Dharmakīrti’s adhyāropa as the cause of doubt (see above
• 3). Both samāropa and anyathā-prasaṅga, however, are seen by
• hinavagupta as inevitably leading to a regressus ad infinitum.

• The possibility that an object may be different from what it appears
• be usually depends on its being mixed with something else (e.g. milk
• th water) or on its having the same features as something else.
• paladeva’s following explanation of the apoha process is intended to
• ver all possible cases, including the one most difficult to account
• r: the case of a distinctly characterized object which is not mixed
• th anything else.

• According to Abhinavagupta, this peculiar variation on the theme of
• oha, which, though not being at all incompatible with the Buddhists’
• sition, has, to the best of my knowledge, never been explicitly put
• rth by them, is to be traced back to some ācāryapāda, presumably a
• iva teacher (Utpaladeva himself?); cf. ĪPVV I: 295, ll. 18-20: tad
• tam ācāryapādair eva yāvatyo ’rthakriyās tāvatsaṁkhyo ’tatkāriṇāṁ
• ṇaḥ । tāvantas tatsamāropās tanniṣedhāś ca tatsamāḥ, etc.

• The unexpressed question that Utpaladeva is answering is highlighted
• Abhinavagupta (ĪPVV I: 294, ll. 6-7): How can a particular object be
• strictly individualized entity and at the same time the occasion for a
• mber of hypothetical alternatives?

• That is, the objects existing on the level of māyā. To sum up, they
• e primarily anyonyabhedamaya, and secondarily anyonyātma.

• In conclusion, the process of apoha can be successfully explained
• ly in a Śaiva context. Particularly, it presupposes that both the
• owing subject and the object are ultimately identified with īśvara and
• e viśvātmā — even at the level of māyā, although in a more virtual and
• ncealed form. In the Buddhist explanation the defect of anavasthā,
• c., is unescapable because vikalpana, samāropaṇa, etc. are separate
• d incapable of relating to each other by their own capacity alone
• PVV I: 296, l. 22: yena [referring to the Śaiva position]
• ̥thagvikalpanasamāropaṇānavasthādoṣāprasaṅgaḥ). This potential
• penness” of the object (and the subject) is at the root of the
• bject’s “doubt” about it. Of this “openness”, i.e. of the
• śvātmā-nature of the object, the subject maintains some kind of
• bliminal, inner awareness, even on the level of māyā. To sum up:
• like the Buddhists who view the “doubt” (i.e. the first step in the
• oha process) as deriving from the object alone, the Śaivas have it
• ise from the “greatness” of the subject as well (ibid.: 296, ll.
• -19: māyāpade ’pi ca tatprakāśasaṁskārasyeśvarasyānativr̥tter āśaṅkā
• amātr̥māhātmyād evottiṣṭhati). The solution of the problem of objective
• d subjective disparateness involved in the Buddhist theory as proposed
• Dharmakīrti, who maintains that the objects, though being essentially
• fferentiated, can be ascertained by the subject as unified owing to
• eir production of a single effect, e.g. a single cognition (PVSV: 60,
• . 16-17: bhāvāḥ prakr̥tibhedino ’pi jñānādikam ekaṁ kāryaṁ kecit
• rvanti nānya iti), does not seem satisfactory to Abhinavagupta. In
• ct, after quoting PV I.119 he refers to his subsequent indication of
• s shortcomings (ĪPVV I: 297, l. 1: tatra sphuritam api vastu
• ḥsamartham ity agre [to be added: sphuṭaṁ] bhaviṣyati).

• It is precisely for this reason, Abhinavagupta says (ĪPVV I: 297, ll.
• 8), that the power of exclusion has been given a separate status,
• ough exclusion is the very body of conceptual cognition and of the
• ecial case of conceptual cognition which is memory.

• Bhagavad-Gītā (BhG) XV.15b: “From me derive memory, knowledge and
• clusion.” This passage sounds somewhat cryptic. For Śaṅkara
• ītābhāṣya [BhGŚ]: 627, l. 4), apohana stands for “turning away from,
• parting from” (apāyanam apagamanam), i.e. the loss of knowledge and
• mory due to evil karma (the objects of apohana are jñāna and smr̥ti:
• dapohanam). According to Ānandagiri’s Vyākhyā thereon (p. 626, ll.
• -15), “knowledge” is the supernatural power to perceive what is beyond
• e normal reach of the senses; “memory” is the capacity to recollect
• at has been experienced in past lives. The commentary of Rājānaka
• makaṇṭha (possibly identical with Utpaladeva’s disciple) is the one
• at seems closer to the interpretation given by Utpaladeva. According
• Rāmakaṇṭha (Sarvatobhadra [BhGS]: 330, ll. 3ff.), memory, perceptual
• gnition and exclusion are the outcome of the outflowing of Power
• aktiprasara) in human beings. This Power, or more precisely this
• nsciousness-Power (saṁvedana-śakti) acts in two ways, in that it
• fines the being and the non-being of the various realities
• attatpadārthabhāvābhāvaparicchedakatva). While the “non-being” of
• alities is defined by means of apohana alone, their “being” is defined
• means of perceptual cognition (which makes something which was not
• ere before manifest) and by means of memory (which brings something
• at has already been perceived in the past again to the mind). In both
• ocesses apohana acts as prakāśanābhāva (absence of manifestation).
• ough these powers act in limited subjects, they ultimately flow from
• e one and absolutely real knowing subject (paramārthasaṁveditur
• asmāt). According to Abhinavagupta’s Bhagavad-Gītārtha-Saṁgraha (BhGA)
• . 231, ll. 10-16), jñāna is the manifestation of something new and
• rresponds to the “great” creation (mahāsr̥ṣṭi) of the universe;
• ohana, which represents the creation of the limited soul, is a
• nceptual thought, which acts by fragmenting all-inclusive reality;
• r̥ti has the function to make manifest again what, subsequent to the
• iginal cognition, has been reduced to the status of a latent
• pression.

• pendix II*

• udies on Utpaladeva’s Īśvarapratyabhijñā-Vivr̥ti