In Light of the Apoha Theory of Śāntarakṣita and Kamalaśīla

[Full title: Communication between a Speaker and a Listener as the Seeing of a Double Moon—In Light of the Apoha Theory of Śāntarakṣita and Kamalaśīla by Kensho Okada]

Editor’s note: This contribution contains changes that the author has left to the editor’s discretion.

Introduction

We experience the practical use of words in our everyday activities 1) when a person under-stands/grasps an object through a word, and 2) when one person tells another something. The latter case is a person-to-person interaction in which a speaker intends to make a listener understand something by speaking word(s), while the listener infers the speaker’s intention by hearing and interpreting them. Communication is said to be established between the speaker and listener only when they are both present. People are, however, quite distinct from one another. Hence, even if the speaker and listener are present in the same time and space, they perceive altogether different things/images respectively. How can they communicate with each other under such circumstances? Also, what is the way in which they engage in conversation with one another?

Śāntarakṣita (ca. 725–788) and Kamalaśīla (ca. 740–795) present an explanation of this issue in the Śabdārthaparīkṣā chapter of their Tattvasaṃgraha (TS) and its pañjikā (TSP). This paper aims to examine their view to clarify how they answer the question at hand.

As is well known, Śāntarakṣita offers a threefold classification of the theory of the “exclusion of others” (anyāpoha):

1. Nominally bound negation/Implicative negation (paryudāsa)

(a) Exclusion having cognition as its essence (buddhyātmāpoha)

(b) Exclusion having a referent as its essence (arthātmāpoha)

2. Verbally bound negation/Non-affirmative negation (prasajyapratiṣedha)^[1]

Of these three exclusions, two of the paryudāsa variety and one prasajyapratiṣedha, Śānta-rakṣita asserts that the exclusion having cognition as its essence (= reflection, pratibimba) is the principal referent denoted by a word, and that the other two are understood by implica-tion. In this way, when reflection is admitted to be the principal referent, the aforementioned question regarding communication arises, for reflections manifest entirely differently in the cognition of individuals. How do Śāntarakṣita and Kamalaśīla resolve this issue?

1. Impossibility of verbal convention-setting

First, let us examine Śāntarakṣita’s statement, which is the origin of the question we seek to answer. In TS 870, he asserts that the 5 categories of particulars, genus, connection to genus, particulars qualified by genus, and the form of cognition are, ultimately, not suitable as the referent denoted by a word.^[2]

To begin with, arguing over particulars Śāntarakṣita and Kamalaśīla state the following:

tatra svalakṣaṇaṃ tāvan na śabdaiḥ pratipādyate /
saṃketavyavahārāptakālavyāptiviyogataḥ // TS 871 //

Of these, first of all, particulars cannot be understood by words. Because [particulars are] devoid of the pervasion of the time extending from verbal convention to verbal activity, [convention cannot be set with regard to particu-lars].^[3] (871)^[4]

TSP 341, 14–17 ad TS 871:

etad uktaṃ bhavati –samayo hi vyavahārārthaṃ kriyate, na vyasanitayā / tena yasyaiva saṃketavyavahārāptakālavyāpakatvam asti, tatraiva samayo vyavahartṝṇāṃ yuktaḥ, nānyatra / na ca svalakṣaṇasya saṃketavyavahārā-ptakālavyāpakatvam asti / tasmān na tatra samaya iti /

[By this,] the [following] is said: Indeed, verbal convention is made for [use at the time of] verbal activity,^[5] not through [mere] desire for [it]. Therefore, it is suitable to [set] verbal convention only with regard to that which pervades the time pervaded by verbal convention and verbal activity for [people] living daily [with use of language], not on anything else. And particulars do not pervade the time extending from verbal convention to verbal activity. Therefore, verbal convention [can]not be [set] with regard to them [i.e., particulars].

Here, from the viewpoint of verbal convention (saṃketa) and verbal activity (vyavahāra), Śāntarakṣita and Kamalaśīla examine whether particulars are suitable to be the referent denoted by a word and indicate that particulars cannot be the target of verbal convention. In the following verses, Śāntarakṣita asserts that (1) convention-setting with regard to particulars is useless for verbal activity^[6] and that (2) it is impossible to make verbal convention in regard to particulars.^[7] In both cases, on the grounds that particulars cannot be something on which verbal convention is established, Śāntarakṣita criticizes the view that particulars are the referent denoted by a word. In the same way, genus, connection to genus, particulars qualified by genus,^[8] and the form of cognition are not accepted as the word-referent due to the impossibility of convention-setting.^[9]

2. Opponent’s criticism

In the final part of the Śabdārthaparīkṣā chapter of TS/TSP, the impossibility of setting verbal convention and the uselessness of it are discussed once again.^[10]

As previously mentioned, Śāntarakṣita and Kamalaśīla admit that the principal referent denoted by a word is the reflection manifesting in cognition. Therefore, for them, the very reflection is the referent which sets with regard to verbal convention. Criticizing their view, an opponent indicates that the impossibility of setting verbal convention and the uselessness of it are also applied to the reflection that they accept as the object of convention.

The opponent says the following:

nanu cāpohapakṣe’pi kathaṃ saṃketasaṃbhavaḥ /
sāphalyaṃ ca kathaṃ tasya na dvayoḥ sa hi siddhyati // TS 1207 //
vaktṛśrotror na hi jñānaṃ vedyate tat parasparam /
saṃkete na ca tad dṛṣṭaṃ vyavahāre samīkṣyate // TS 1208 //

[Objection:] Even in the apoha theory, (1) how can convention[-setting] be possible? (2) How can it be useful, either? (1) For it [i.e., the apoha as reflection] cannot be established between two [persons, i.e.,] a speaker and listener [in the same way]. It is because [they can]not mutually cognize [one another’s] cognition. (2) And, that [reflection] which was perceived at the time of [setting the] verbal convention is not seen [any longer] at the time of verbal activity.

Here, adopting the same method as Śāntarakṣita, the opponent indicates that it can be neither possible nor useful to set verbal convention with regard to a reflection.

As far as the impossibility of the setting of verbal convention with regard to a reflection is concerned, according to the opponent, the speaker never cognizes the reflection manifesting in the listener’s cognition, nor does the listener cognize the reflection manifesting in the speaker’s cognition. Therefore, it cannot be established that their reflections are one and the same. This is the reason for the impossibility of the setting of verbal convention with regard to such reflections.

Regarding the uselessness of verbal convention-setting, the opponent asserts that, because the reflection perceived at the time of verbal convention and the one perceived at the time of verbal activity are totally different, it follows that setting verbal convention with regard to a reflection is not useful for ordinary verbal usage.^[11]

With regard to the former question, Kamalaśīla explains as follows:

TSP 447,1–7 ad TS 1208’ab:

kasmād ity āha –na hi jñānam ityādi / pratyātmasaṃvedanīyam evārvāgda-rśanānāṃ jñānam / na hy anyadīyajñānam aparo’paradarśanaḥ saṃvedaya-te / jñānād avyatiriktaś ca paramārthataḥ pratibimbātmakalakṣaṇāpohaḥ /

tataś ca vaktṛśrotror dvayor api kasyacid ekasya saṃketaviṣayasyārthasyā-siddheḥ kutra saṃketaḥ kriyate gṛhyate vā / na hy asiddhe vastuni vaktā saṃketaṃ kartum īśaḥ, nāpi12 śrotā gṛhītum, atiprasaṅgāt / tathā hi –śrotrā13 yat pratipadyate svavijñānārūḍham arthapratibimbakaṃ na tad vaktrā saṃve-dyate / yac ca vaktrā saṃvedyate na tac chrotrā, svasya svasyaivāvabhāsasya vedanāt /

[Reply:] Why? [Objection:] “[They can]not [cognize one another’s] cogni-tion” and so on. Ordinary persons can only cognize something to be cognized by themselves, for, being an ordinary person, no one knows the cognition [represented in] another’s [mind]. Moreover, the exclusion, having a reflection as its own nature, is ultimately not different from the cognition. Therefore, [the following question arises:] With regard to what [object] can the convention be set [by a speaker]? In what [object] can the convention be comprehended [by a listener]? It is because the same object is not established as the target of verbal convention at all between the two persons–that is, the speaker and listener. Indeed, if nothing real is established, the speaker cannot make any convention; nor can the listener comprehend it. This is because [if such a thing were possible,] an absurd consequence would follow. Namely, the lis-tener would comprehend the reflection of an object manifesting in his own cognition; but the speaker would be unable to cognize it. The listener cannot comprehend what the speaker cognizes. This is because [each] cognizes only his own manifestation.

Here, the argument encounters difficulty when a reflection is regarded as the object with regard to which verbal convention is set. Reflections are different for each individual. An ordinary person can only cognize the reflection manifesting in his own cognition. It is impossible for a reflection to move from one person’s cognition to another’s. Therefore, in reality, one can neither tell the content of his cognition to another, nor perceive that of another’s cognition. As a result, it may follow that there is not any object of verbal convention that a speaker and listener share at the time of verbal convention, and thus, the verbal convention itself cannot be established.

3. Śāntarakṣita and Kamalaśīla’s reply

Śāntarakṣita and Kamalaśīla sought to address this issue. How do they explain the manner of establishing verbal convention? Śāntarakṣita says the following:

svasya svasyābhāsasya vedane’pi sa vartate /
bāhyārthādhyavasāyo^[12] yo^[13] dvayor api samo yataḥ // 1209 //
timiropahatākṣo hi yathā prāha śaśidvayam /

svasamāya tathā sarvā śābdī vyavahṛtir matā // 1210 //
vyāpakatvaṃ ca tasyedam iṣṭam ādhyavasāyikam /
mithyāvabhāsino hy ete pratyayāḥ śabdanirmitāḥ // 1211 //

It is because, although [a speaker and listener] cognize their respective man-ifestations, they both equally determine [their respective manifestations] to be an external object. Just as one [person] whose eyes are affected by eye disease says to [another person] like himself “there is a double moon,” all verbal activities are thought to be the same. That [i.e., reflection] is accepted as being such a pervader, [i.e., being that which pervades the time extending from verbal convention to verbal activity,]^[14] on the basis of [persons] determining [so]. In actuality, these notions brought about by word(s) are [nothing more than] something manifesting erroneously.

Commenting on TS 1209, Kamalaśīla states the following: TSP 447,12–448,3 ad TS 1209:

na hi paramārthato jñānākāro’pi śabdānāṃ vācyatayā’bhīṣṭaḥ, yena tatra saṃketāsambhavaś codyate / yataḥ sarva evāyaṃ śābdo vyavahāraḥ svapra-tibhāsānurodhena taimirikadvayadvicandradarśanavad bhrānta iṣyate / ke-valam arthaśūnyābhijalpavāsanāprabodhāc chabdebhyo’rthādhyavasāyivi-kalpamātrotpādāt17 tat pratibimbakaṃ śabdānāṃ vācyam ity abhidhīyate jananāt, na tv abhidheyatayā / tatra yady api svasya svasyaivāvabhāsasya vaktṛśrotṛbhyāṃ paramārthataḥ saṃvedanam, tathāpi taimirikadvayasyeva bhrāntibījasya tulyatvād dvayor api vaktṛśrotror bāhyārthādhyavasāyas tulya eva / tathāpi vaktur ayam abhimāno vartate –yam evāham arthaṃ pratipādye tam evāyaṃ pratipadyata iti / evaṃ śrotur api yojyam /

ekārthādhyavasāyitvaṃ katham anayor vaktṛśrotroḥ parasparaṃ viditam iti cet, yadi nāma paramārthato na viditam, tathāpi bhrāntibījasya tulyatvād asty eva paramārthataḥ svapratibhāsānurodhena taimirikadvayavad bhrānta evāyaṃ vyavahāra iti niveditam etat / tenaikārthādhyavasāyavaśāt saṃketa-karaṇam upapadyata eva /

Indeed, ultimately, the form of cognition is also not admitted to be the referent denoted by words, so that the impossibility of [setting] convention with regard to it [i.e., the form of cognition] would be criticized. It is because all verbal activities are admitted to be something erroneous, just as in the case of two persons affected by eye disease [both] seeing double moons in accordance with their respective manifestations. Merely by awakening of the latent disposition which [is deposited] through objectless discourse, only conceptual [cognitions], by which [a manifestation] is determined as being an object, are produced by words. [Hence,] on the grounds that [conceptual cognition is] produced [by a word], that reflection [of conceptual cognition] is said to be the referent denoted by words; not on the grounds that [the reflection is, in reality, the referent denoted by a word]. That being the case, although a speaker and listener ultimately only cognize their respective manifestations, they still both equally determine [their respective manifestations] to be external objects, because [they] share the cause of the error, just as in the case of two persons affected by eye disease [both seeing a double moon]. Even so, the speaker thinks, “He [i.e., the listener] also understands the object that I understand.” So does the listener.^[15]

[Question:] How can the speaker and the listener know that they are determin-ing the same object?

[Reply:] Although they ultimately cannot know [it], such a verbal activity, which is nothing but erroneous, [is still established between them] in accor-dance with their respective manifestations, because [they] share the cause of the error, just as the case of the two persons affected by eye disease. It has already been explained. Therefore, it is quite possible that verbal convention is made by the force of [people] determining [their respective manifestations] to be the same.

Śāntarakṣita and Kamalaśīla accept that, in reality, a speaker and listener cognize only their own respective manifestations. According to them, however, a speaker and listener determine their own manifestations as external things in the same way. This same determi-nation is the reason for the establishment of verbal convention. That is to say, a convention is established on the basis of the thinking that “we perceive the same thing.” Based on this erroneous determination, two persons can communicate.

Moreover, there would appear to be no difference between the verbal activity of persons affected by eye disease and that of ordinary persons, given that all people cognize only their own respective manifestations. Śāntarakṣita and Kamalaśīla therefore assert that all verbal activity is erroneous.

4. Causal relation between word and intention

It is found that Kamalaśīla uses this example of “two persons affected by eye disease seeing a double moon in the same way” also in the Śābdavicāra section of the Pramāṇāntaraparīkṣā chapter and Śrutiparīkṣā chapter of the TSP. There, Śāntarakṣita denies that there is an invariable connection between a word and an external thing, and suggests that a causal relation is established between a word and speaker’s intention.^[16] While commenting on such statements by Śāntarakṣita, Kamalaśīla gives explanations with use of the example of “two persons affected by eye disease.”^[17]

Here, let us consider the views of Śāntarakṣita and Kamalaśīla as presented in the Śābdavicāra section of the TS/TSP.

There, as mentioned above, rejecting an invariable connection between a word and an external thing, Śāntarakṣita explains the manner in which verbal cognition is meant to be included in the inference.

vacobhyo nikhilebhyo’pi vivakṣaiṣānumīyate /
pratyakṣānupalambhābhyāṃ taddhetuḥ sā hi niścitā // TS 1514 //

From all words it follows that [speaker’s] intention is inferred. For it [i.e., intention] is ascertained to be the cause of a word through perception and non-perception.

Here, Śāntarakṣita states that an invariable connection having causality as its character is established between a word and an intention, and suggests that a word is a valid cognition with regard to intention. In some cases, however, an intention and an uttered word are known to be different. That is to say, it is just as in the case of some speaker uttering the word “Y” erroneously while intending to speak about “X.” In this case, it is assumed that the fallacy lies not in the word but in the erroneous speaker.^[18]

Furthermore, with respect to a word’s deviation from a particular intention, Kamalaśīla gives explanatory statements as follows:

TSP 541, 5–11 ad TS 1517–1518:

api ca, yadā sarva evāyaṃ śābdo vyavahāras taimirikadvayadvicandradarśa-navat svapratibhāsānurodhena bhrānta eveṣyate, tadā kathaṃ vivakṣāviśeṣe vyabhicārodbhāvanād aprāmāṇyaprasañjanaṃ syāt, tatra vivakṣāviśeṣe pā-ramārthikasya prāmāṇyasyānabhyupagamāt / tathā cāha –saṃketāpekṣayā tasya hṛdi kṛtvā prakāśanam / anumānatvam uddiṣṭaṃ na tu tattvavyape-kṣayā // iti /

Furthermore, when all verbal activity is accepted to be erroneous due to dependence on [respective] manifestations of one’s own, just as two persons affected by eye disease see a double moon [in the same way], how can [a word/ verbal cognition] be invalid on the basis of pointing out the deviation from a particular intention [of speaker]. For it is not accepted that [a word/verbal cognition has] ultimate validity with regard to a particular intention of speaker. The same thing is stated as follows: A word (/verbal cognition) is said to be an inference just when it is taken into consideration that a word makes manifest [an object] with dependence on verbal convention, not when with dependence on real truth.^[19]

Here, from the standpoint that all verbal activity is erroneous, Kamalaśīla implicitly indi-cates that the validity of words/verbal knowledge is not ultimate but conventional. Thus, as long as the erroneousness of verbal knowledge is considered, it follows that the validity of a word/verbal knowledge depends only on verbal convention, and has nothing to do with real truth.

For example, when one taimirika (person affected by eye disease) says to another taimirika “there is a double moon in the sky” and the latter agrees with the former, it follows that based on this agreement, verbal convention is established between them. In this case, in reality, the so-called “double moon” does not exist, and the word “double moon” would be nonsense. However, to the two taimirika the word “double moon” is thought to be useful.

Conclusion

To conclude, when reflection is admitted to be the referent denoted by a word, a speaker and listener cannot share the same real object, because they only cognize different reflections manifesting in their respective cognitions. They, however, think/believe that they perceive the same thing by the force of determination. Based on this same determination, the setting of convention is established. Therefore, according to Śāntarakṣita and Kamalaśīla, all verbal activity is erroneous, as it is the same as when two persons affected by eye disease both see a double moon.

Supplemental remarks

As discussed above, bearing in mind that a person cognizes only his own representation, Śāntarakṣita and Kamalaśīla suggest that all verbal activity is established in the same ways as two persons with eye disease equally see a double moon. Interestingly enough, similar views have been presented by Dharmakīrti in his Santānāntarasiddhi (SS) and by Vinītadeva in his Santānāntarasiddhiṭīkā (SSṬ).^[20] Dharmakīrti and Vinītadeva state the following:

SS 65 48,5–10:

bdag dang gzhan la rang rang gi snang ba nyams su myong mod kyi / rab rib can gnyis kyis zla ba gnyis mthong ba bzhin du rnam par shes pa de lta bu’i rgyu’i bag chags skye ba’i rang bzhin gyi khyad par thog ma med pa’i dus las don gcig’dzin par nges pa’i phyir ro //

[A representation is called a manifest action (vijñapti).] It is because although each person [i.e., a speaker and listener,]^[21] experiences his own representations, just as two persons with eye-disease perceive a double moon [in the same way], it is determined that the same thing is grasped by a special nature, which has as its nature the arising of the latent disposition [produced by] the cause of such cognition, from beginningless time.

SSṬ 49,17–50,18 ad SS 65:

’di ltar smra ba po yang’di snyam du bdag gis smras pa gang yin pa’di nyan pa pos rtogs par’gyur ro snyam du sems pa la / nyan pa po yang’di snyam du’dis smras pa gang yin pa de ni bdag gis rtogs so snyam du sems par’gyur te / … … ’di skad du smra ba po dang nyan pa po dag rang rang gi snang ba nyams su myong mod kyi / ’on kyang de gnyis kyi bya ba dang tshig gi rnam pa can gyi shes pa rgyu gcig las’byung ba gang dag yin pa de dag don gcig tu shes par nges pa’i phyir rnam par rig byed du tha snyad’dogs par byed do // don gcig tu’dzin par nges pa ni rnam par shes pa de lta bu’i rgyu’i bag chags skye ba’i rang bzhin gyi khyad par thog ma med pa las byung ba’i phyir ro // ’dir dpe ni rab rib can gnyis kyis zla ba gnyis su mthong ba bzhin te / rab rib can gcig gis rab rib can gnyis la ltos zhes zla ba gzhan zhig ston pa las des mthong ngo zhes smras pa na / de la ni ston pa po yang’di snyam du bdag gis’di la bstan to snyam du nga rgyal byed par’gyur la nyan pa po yang’di snyam du’dis bdag la bstan to snyam du sems mod kyi / ’on kyang de gnyis ni rang rang gi snang ba so sor nyams su myong bar zad do //

That is to say, the speaker thinks “the listener will understand what I say.” The listener, in his turn, thinks “I have understood what he said.” … Although the speaker and the listener both experience their respective representations, cognitions having the form of action and speech appear in both from the same cause, and these [cognitions] determinately grasp the same thing. Therefore, it is called manifest action (vijñapti) metaphorically. Because there exists a special nature, which has as its nature the arising of the latent disposition [produced by] the cause of such cognition, from beginningless time, the same thing is determinately grasped. “Just as in the case that two persons see a double moon” is the example [demonstrating] this. If a person affected by eye disease (A) indicates another moon and says to another person affected by eye disease (B) “Look!” and he (B) says “I see [it],” then, the speaker thinks “I have indicated it to him,” and the listener thinks “he indicated it to me.” Even so, they both only cognize their respective representations.

References and abbreviations

AJP Anekāntajayapatākā (Haribhadra): Anekāntajayapatākā by Haribhadra Sūri: with his own commentary and Municandra Sūri’s supercommentary, ed. H. R. Kāpadīā. Baroda 1940.

AVK Anyāpohavicārakārikā (Śubhagupta): D4246, P5744.

McAllister 2011 P. McAllister, Ratnakīrti’s Apohasiddhi: A Critical Edition, Annotated Translation and Study. Dissertation, Universität Wien. Wien 2011.

Mikogami 1993 E. Mikogami, Śubhagupta no Komyunikēshon Ron. In: Genshi Daijō Bukkyō to Daijō Bukkyō, ed. E. Maeda. Kyōto 1993 (in Japanese).

PV I Pramāṇavārttika, chapter I (Dharmakīrti): See PVSV.

PVSV Pramāṇavārttikasvavṛtti (Dharmakīrti): The Pramāṇavārttikam of Dharmakīrti: The First Chapter with The Autocommentary, ed. R. Gnoli. Roma 1960.

PVSVṬ Pramāṇavārttikavṛttiṭīkā (Karṇakagomin): Karṇakagomin’s Commentary on the Pramāṇavārttikavṛtti of Dharmakīrti, ed. R. Sāṅkṛtyāyana. Repr. Kyoto 1982.

PVṬ Pramāṇavārttikaṭīkā (Śākyabuddhi): D4220, P5718.

SS Saṃtānāntarasiddhi (Dharmakīrti): Santānāntarasiddhi, ed. T. Stcherbatsky. Petrograd 1916.

SSṬ Saṃtānāntarasiddhiṭīkā (Vinītadeva): See SS.

TS Tattvasaṃgraha (Śāntarakṣita): See TSP (G) and TSP (B). The verse-numbers follow TSP (B).

TS (D) Tattvasaṃgraha (Śāntarakṣita). Tibetan: D4266.

TS (P) Tattvasaṃgraha (Śāntarakṣita). Tibetan: P5764.

TSP Tattvasaṃgrahapañjikā (Kamalaśīla): See TSP (G) and TSP (B). Page-and line-numbers are given in accordance with TSP (B).

TSP (B) Tattvasaṃgrahapañjikā (Kamalaśīla): Tattvasaṅgraha of Ācārya Shāntarakṣita with the Commentary “Pañjikā” of Shrī Kamalaśīla, ed. S. D. Shastri. Varanasi 1968. TSP (G) Tattvasaṃgrahapañjikā (Kamalaśīla): Tattvasaṅgraha of Śāntarakṣita: With the Commentary of Kamalaśīla, ed. E. Krishnamacharya. Baroda 1926.

TSP (D) Tattvasaṃgrahapañjikā (Kamalaśīla). Tibetan: D4267.

TSP (P) Tattvasaṃgrahapañjikā (Kamalaśīla). Tibetan: P5765.

Footnotes and references:

[1]:

Cf. TS 1003, 1010, and 1014cd.

[2]:

Cf. TS 870: yataḥ svalakṣaṇaṃ jātis tadyogo jātimāṃs tathā / buddhyākāro na śabdārthe ghaṭām añcati tattvataḥ //; McAllister 2011: 181–182.

[3]:

Cf. TSP 341,13–14 ad TS 871: na tatra svalakṣaṇe samaya iti śeṣaḥ /

[4]:

Cf. PV I 92: śabdāḥ saṃketitaṃ prāhur vyavahārāya sa smṛtaḥ / tadā svalakṣaṇaṃ nāsti saṃketas tena tatra na // “Words express something upon which verbal convention has been agreed and it [i.e., verbal convention] is recalled for the purpose of verbal activity. At the time [of verbal activity] the particular [with regard to which verbal convention is set] no longer exists. Therefore, verbal convention is not set with regard to the particular.”

[5]:

According to Śākyabuddhi’s commentary, Dharmakīrti asserts that the purposes of verbal convention are 1) the understanding of form of cognition and 2) the apprehension of particulars. Cf. PVṬ D127a7–b5/ P151a3–b2 ad PV I 110d and PVSV 57,15–19.

[6]:

Cf. TS 873: tasmāt saṃketadṛṣṭo’rtho vyavahāre na dṛśyate / na cāgṛhītasaṃketo gamyate’nya (gamyate’nya G; (bodhyetā)nya B) iva dhvaneḥ // “Therefore, the thing perceived [at the time of] verbal convention is never perceived at [the time of] verbal activity. And, that which is not grasped by a verbal convention cannot be understood by a word. Just as with the fact that another thing [e.g. horse] cannot be [understood by the word ‘cow’].”

[7]:

Cf. TSP 344,1–3 ad TS 875–876: evaṃ tāvat svalakṣaṇe vyavahārānupapatteḥ samayavaiyarthyaprasaṅgān na samayaḥ sambhavatīti pratipāditam / sāmpratam aśakyakriyatvād eva na sambhavatīti pratipādayann āha–aśakyam ityādi / “In this way, because verbal action toward particulars is impossi-ble, it would follow that verbal convention is useless, [Therefore, the setting of] verbal convention with regard to particulars is impossible. This is, firstly, understood. Next, in order to explain that, because it is impossible to make [verbal convention], [the setting of verbal convention] is impossible, [Śāntarakṣita] states ‘aśakyam’ and so on.”

[8]:

Cf. TS 882: jātisambandhayoḥ pūrvaṃ vyāsataḥ pratiṣedhanāt / nānantarāḥ prakalpyante śabdārthās trividhāḥ pare // “Because genus and connection have been already rejected in detail before, the other three [which are enumerated] immediately after [particulars] are not supposed to be the referent of a word.” TSP 348,7–8 ad TS 882: evaṃ tāvat svalakṣaṇe samayāsambhavaṃ pratipādya jātyādiṣu triṣu samayāsambhavaṃ pratipādayann āha–jātisambandhayor ityādi / “In this way, to begin with, the impossibility of [setting] verbal convention with regard to particulars is explained. After this, in order to explain the impossibility of [setting] verbal convention with regard to three things–that is, genus and so on [Śāntarakṣita] states ‘jātisambandhayor’ and so on.”

[9]:

Cf. TS 884: buddhyākāraś ca buddhistho nārthabuddhyantarānugaḥ / nābhipretārthakārī ca so’pi vācyo na tattvataḥ // “Moreover, existing [only] in some cognition [as identical in nature], the form of cognition follows neither [external] things nor other cognitions, and does not accomplish desired purposes. Ultimately, that [i.e., the form of cognition] is also not the referent denoted [by a word].” TSP 349,2–4 ad TS 884: buddhyākāro hi tādātmyena buddhāv evāvasthita iti nāsau tadbuddhisvarūpavat pratipādyam arthaṃ buddhyantaraṃ vānugacchati / tataś ca saṃketavyavahārāptakālāvyāpakatvāt svalakṣaṇavan na tatrāpi samayaḥ sambhavati / “… Therefore, [the setting of] verbal convention with regard to it [i.e., form of cognition] is also impossible, because it does not pervade the time extending from verbal convention to verbal activity, just as particulars.”

[10]:

Cf. TS 1206ab: saṃketāsambhavo hy atra bhedādau sādhitaḥ purā / “Indeed, the impossibility of [the setting of] verbal convention with regard to particulars and so on has already been proved before.” TSP 446,8–10 ad TS 1206ab: aśakyasamayatvād ananyabhāktvāc ceti pūrvaṃ svalakṣaṇādau saṃketāsa-mbhavasya saṃketavaiphalyasya ca prasādhitatvāt / “It is because the impossibility of [setting] verbal convention and the uselessness of it with regard to particulars and so on have already been proved in such ways as ‘because [the setting of] verbal convention is impossible’ and ‘because of having no connection with the others’.”

[11]:

Commenting on the opponent’s view, Kamalaśīla explains the uselessness of the setting convention with regard to reflections as follows: Cf. TSP 447,7–11 ad TS 1208cd: ānarthakyaṃ ca pratipādayann āha–saṃkete na cetyādi / yat saṃketakāle pratibimbakam anubhūtaṃ śrotrā vaktrā vā, na tad vyavahārakāle’nubhūyate, tasya kṣaṇakṣayitvena ciraniruddhatvāt / yac ca vyavahārakāle’nubhūyate na tat saṃketa-kāle dṛṣṭam, anyasyaiva tadānīm anubhūyamānatvāt / na cānyatra saṃketād anyatra vyavahāro yuktaḥ, atiprasaṅgād iti / “In addition, in order to explain the uselessness [of setting convention with regard to reflection, the opponent] states ‘saṃkete na ca’ and so on. Some reflection has been experienced by a listener or by a speaker at the time of the verbal convention. [However, ] that [reflection] cannot be experienced at the time of verbal activity. This is because that [reflection], being momentary, has ceased to existence long before [the time of the verbal activity]. Moreover, that which is experienced at the time of the verbal activity cannot be that which has been perceived at the time of verbal convention. This is because a very different [reflection] is experienced at that time (= at the time of verbal activity). And, it is untenable that verbal activity regarding Y is [established] on the basis of verbal convention [set with regard to] X. This is because [otherwise] an absurd consequence would follow.”

[12]:

°ādhyavasāyo em. (cf. zhen pa D, P; bāhyārthādhyavasāyas tulya eva TSP 477,18); °ādhyavasāye G, B.

[13]:

yo em.; yad G, B; yad om. D, P.

[14]:

Cf. TSP 448,6–8 ad TS 1211: saṃketavyavahārāptakālavyāpakatvaṃ ca vaktṛśrotṛbhyām adhyava-sitārthapratibimbakasyāvasāyavaśād eveṣṭaṃ, na paramārthataḥ /

[15]:

Kamalaśīla has already stated the same thing in the following portion. Cf. TSP 358,9–12 ad TS 908: etad uktaṃ bhavati–paramārthataḥ svapratibhāsānubhave’pi vaktur evam adhyavasāyo bhavati–mayā’smai bāhya evārthaḥ pratipādyate / śrotur apy evaṃ bhavati–mamāyaṃ bāhyam eva pratipādayatīti / atas taimirikadvayadvicandradarśanavad ayaṃ sarvaḥ śābdo vyavahāra iti (cf. sgra las byung ba’i tha snyad’di thams cad de lta bu yin no D, P) /

[16]:

Cf. TS 1512 (Śābda): vacasāṃ pratibandho vā ko bāhyeṣv api vastuṣu / pratipādayatāṃ tāni yenaiṣāṃ syāt pramāṇatā // “How can be there any invariable connection between words and external things? If [invariable connection were to be admitted], words expressing them [i.e., external things,] would be valid cognition.” TS 2618 (Śruti): sākṣāc chabdā na bāhyārthapratibandhavivekataḥ / gamayantīti ca proktaṃ vivakṣāsūcakās tv amī // “It has already explained that words do not make [external things] known directly, because there is no invariable connection between external things [and words]. And, these [i.e., words,] are [nothing but] something indicating [the presence of the speaker’s] intention.” TSP 854,1–3 ad TS 2618: vivakṣām api na vācyatayā pratipādayanti, kiṃ tarhi, liṅgatayā sūcayanti / ata eva sūcakā ity uktam / tathā hi–śabdād uccaritād arthādhyavasāyī vikalpo jāyate, na vivakṣādhyavasāyī / “[Words] do not express [the speaker’s] intention as something denoted [by words]. How then? [Words merely] indicate [the speaker’s intention] as a logical mark. Therefore, [words are] said to be something indicating. That is to say, [when] a word is uttered, there arises a conceptual cognition which judges an [external] thing, not [a conceptual cognition which] judges [the speaker’s] intention.”

[17]:

In his TS 2620, Śāntarakṣita asserts that only when a person has already cognized a causal relation, can he know a speaker’s intention through words. Against this, the opponent objects that verbal convention cannot be a means for knowing a particular intention of a speaker. In response, Kamalaśīla criticizes the opponent’s view with use of the example of “two persons affected by eye disease seeing a double moon.” Cf. TSP 854,12–22 ad TS 2620.

[18]:

Cf. TS 1515–1516: bhrāntasyānyavivakṣāyāṃ vākyaṃ ced anyad īkṣyate / yathāvivakṣam (yathāvivakṣam B; tathā vivakṣam G) apy etat tasmān (etat tasmān B; etatta(kṣāsāmānye ta?)smān G [sic]) naiva pravartate // bhrāntābhrāntaprayuktānāṃ vailakṣaṇyaṃ parisphuṭam / vidagdhāḥ prakṛtādibhyo niścinvanti girām alam // vailakṣaṇyena hetūnāṃ viśeṣaṃ tāsu ye na tu / avagacchanti doṣo’yaṃ teṣāṃ liṅgasya nāsti tu // “[Objection:] It is found that as for an erroneous person, his statement and intention are different. Therefore, it [i.e., his statement] cannot be made in accordance with intention. [Reply:] There is clearly a difference between words used by an erroneous person and those used by a non-erroneous person. [However,] clever men [definitely] discern [this difference] through context and so son. The fallacy lies not in the logical mark, but in the persons who do not notice the difference among them [i.e., these words] due to difference of their causes.”

[19]:

The source for this citation is uncertain.

[20]:

For your information, similar descriptions can also be found in Śubhagupta’s Anyāpohavicārakārikā (AVK) and Karṇakagomin’s Pramāṇavārttikasvavṛttiṭīkā (PVSVṬ), as far as treatises on apoha theory are concerned. Cf. Mikogami 1993. Cf. AVK D198a1–2: rang gi blo la snang ba ni // rig pa yin yang phyi rol dang // gcig par’dzin pa’i nga rgyal skye // rab rib can du mtshungs blo bzhin // “Although [persons] cognize their own [respective] manifestations, they [erroneously] think ‘We comprehend the same external thing,’ just as the cognitions of persons suffering from same eye disease.” This verse is quoted in Anekāntajayapatākā (AJP 338,6–7 svabuddhipratibhāsasya saṃvittāv api jāyate / bahi-rarthagrahe mānas tulyataimirabuddhivat //); PVSV 60,26–61,1: … pratipattā pratipattim anusṛtya ete vṛkṣā iti svaparavikalpeṣv ekapratibhāsān ādarśya vikalpavijñāne vyavasthitas tadvijñānahetūn bhedena pratipadyetety uktim ataddhetubhyo bhede niyuṅkte. “… [One who makes the setting of a verbal convention] applies a word to what is excluded from those which are not its causes with the following hope: The cognizer [for whom the verbal convention is made such as ‘tree’] would be able to understand that these are trees [when he hears the word ‘tree’] in accordance with his understanding [of the verbal convention], having referred to [the object] which equally appear in the cognitions of himself and others, although being dependent [solely on his own conceptual cognition], by differentiating the causes of this [cognition from others];” PVSVṬ 241,10–21 ad PVSV 60,27: nanu vyāvṛttasya svalakṣa-ṇasya vyavahārakāle’nugamo nāsti. nāpi vikalpapratibhāsinaḥ sāmānyākārasya svajñānābhinnatvād vikalpāntare’nvayo’sti. nāpi vaktṛsaṃbandhinas tasya śrotuḥ śrotṛsaṃbandhino vā vaktuḥ pratītiḥ, anyacetodharmatvenātīndriyatvāt. na cāpratipanne samaṃ pratipādyapratipādakābhyāṃ saṃketaḥ saṃbhavatīty ata āha—svaparetyādi. … etad uktaṃ bhavati–yathaikas taimiriko dvicandraṃ dṛṣṭvā-nyataimirikāyopadiśan svadṛṣṭam evopadiśati, na paradṛṣṭam, apratyakṣatvāt / atha ca tasyaivaṃ bhavati–ayam eva mayā parasmai pratipādita iti / paro’pi ca svasaṃtānabhāvinam eva dvicandrākā-raṃ pratiyan ya eva pratipādakena mama pratipāditas sa eva mayā pratipanna iti manyate / tadvat pratipādyapratipādakayor buddhyākārasyādhyavasita-bāhyarūpasya bhede’py ekatvādhyavasāyāt saṃketakaraṇam / vyavahārakāle ca tasyaiva pratītiḥ, ekatvādhyavasāyāt / “[Objection:] An excluded particular does not continue to exist at the time of verbal activity. Also, since the form of the universal manifesting in one’s conceptual cognition is not different from this one’s own cognition, it does not occur in the other’s conceptual cognition. Neither one who listens to the speaker nor one who speaks to the listener cognize [the form of the universal manifesting in the others’ conceptual cognition], because it is not cognizable by sense organs. It is because it is the quality belonging to the other’s mind. When [the form of the universal appearing in the others’ conceptual cognition] is not cognized, the speaker (pratipādaka) and listener (pratipādya) cannot equally make verbal convention. [Reply:] [Dharmakīrti] states ‘svapara’ and so on. … [By this,] the following is said: When some person affected by eye disease sees a double moon and tells this to another person affected by eye disease, he describes only that which is perceived by himself, not that which is perceived by anyone else. This is because [he] cannot perceive [what anyone else perceives]. And he thinks ‘I explain it to him.’ Also, another person, understanding the form of the double moon that belongs to his own mind, thinks ‘I understand what he has explained to me.’ In the same way, although the forms of cognition, which are determined to be an external thing, are different between someone explaining [i.e., speaker] and someone to be explained [i.e., listener], conventions are made [on the forms of cognition] by the force of the determination of [these forms] being the same. And, it [i.e., the form of cognition] is cognized at the time of verbal activity, because [the form of cognition at the time of verbal activity is] determined to be the same [as that at the time of verbal convention].” (There is no corresponding description in Śākyabuddhi’s Pramāṇavārttikaṭīkā.)

[21]:

Cf. SSṬ 48,13–14: bdag dang gzhan zhes bya ba ni smra ba po dang nyan pa po’o //