One of my long time fascinations is with the Pali Kaccānagotta Sutta or Sanskrit Kātyāyana Sūtra. It survives in three versions: Pāḷi, Chinese, and Sanskrit. It is fairly well known that Nāgārjuna quotes a Sanskrit version of this text in his Mūlamadhyamakakārikā (MMK 15.7). It's less well known that a number of Mahāyāna Sūtras appear to quote this sūtra, as well. Long term, I would like to do a complete survey of how this text was used in Buddhism over time, but we can say that it forms an important link between Mahāyāna and Mainstream forms of Buddhism. Some very useful reading on this subject can be found in Salvini (2011). There is also some discussion focussed on MMK in Kalupahana (1986).
In this essay I'll translate and discuss a passage from the first chapter of the Aṣṭasāhasrikā Prajñāpāramitā Sūtra and argue that it recapitulates the argument about dharmas from the Kātyāyana. The implication here is that Kātyāyana provides a conceptual continuity link between trends of Buddhism. It represents a truth about experience that is widely acknowledged by different Buddhist schools of thought.
In my next blog essay I'll be exploring some important ideas about the history of the early Mahāyāna. One thing that has emerged recently is that Mahāyāna texts were almost certainly composed orally and in Prakrit. In the case of the Aṣṭa, we have physical evidence in the form of a birch bark manuscript, written in the Gāndhārī Prakrit in Kharoṣṭhī script and carbon dated to the first century CE (the mid-point for the probability curve is 79 CE). So the Sanskrit text is a translation. Aṣṭa might have been translated into Sanskrit as late as the 5th century CE. This undermines the claim of the Sanskrit version of Aṣṭa (or any Mahāyāna text) to be "the original". In some ways, the early translations into Chinese might better represent the original text, though this is debatable.
In my next blog essay I'll be exploring some important ideas about the history of the early Mahāyāna. One thing that has emerged recently is that Mahāyāna texts were almost certainly composed orally and in Prakrit. In the case of the Aṣṭa, we have physical evidence in the form of a birch bark manuscript, written in the Gāndhārī Prakrit in Kharoṣṭhī script and carbon dated to the first century CE (the mid-point for the probability curve is 79 CE). So the Sanskrit text is a translation. Aṣṭa might have been translated into Sanskrit as late as the 5th century CE. This undermines the claim of the Sanskrit version of Aṣṭa (or any Mahāyāna text) to be "the original". In some ways, the early translations into Chinese might better represent the original text, though this is debatable.
The passage that I want to explore is Chapter 1, section 19; Vaidya (1960). In Conze's translation (1973) this passage occurs on p.87-88. My translation is:
When that was said, the Bhagavan said this to Elder Śāriputra, "thus training, Śāriputra, the bodhisattva mahāsattva does not train in any dharma. What is the reason for it? For the dharmas do not exist in the way that the foolish, ignorant hoi polloi take them to exist."
Elder Śāriputra said, "How then do they exist, Bhagavan?"
The Bhagavan said, "They exist as though they don't exist. Not being found in that sense they are said to be unfound (avidye). The foolish, ignorant hoi polloi are engrossed in them. All dharmas imagined by them are non-existing. Having imagined them, they are obsessed by the two extremes. They don’t know or see those dharmas. Therefore, all dharmas they imagine are non-existing. Having imagined [the non-existing dharmas], they are engrossed in the two extremes; engrossed, they rely on the observed object as a basis and imagine dharmas in the past, in the future, and in the present moment. Having imagined them, they become engrossed in name and form. Non-existing all-dharmas are imagined by them. Imagining those non-existing all-dharmas, they do not know and do not see the path as it really is. Not knowing or seeing the path as it really is, they don’t depart from the triple realm and do not awaken to the highest truth. They go by the name “fools”. They do not develop faith in the true dharma. The bodhisattva mahāsattva does not become engrossed in any dharma, Śāriputra."
Typically, Conze manages to make this section paradoxical. He has dharmas both existing and not existing at the same time, which does not make sense on any terms. For Conze, such non-sense is a way of pointing to a transcendent, ineffable truth that words are incapable of communicating. Supposedly, the contradiction temporarily confuses the rational mind (as conceived) and allows the intuitive mind (as conceived) to make an intuitive leap to the transcendent truth. There are many false assumptions here about the nature of reason† and imagination.
† See for example: Reasoning and Beliefs. (10 Jan 2014)
The important point of the Kātyāyana is that existence (astitā) and non-existence (nāstitā) are not terms that can be applied to dharmas qua mental objects. The typical Mahāyāna explanation, following the Two Truths doctrine, is that dharmas both exist and do not exist. Kātyāyana makes sense, the Two Truths explanation does not. I believe that, in this passage from Aṣṭa, the Kātyāyana argument about dharmas is being recapitulated in much the same terms, and with the same warning about what happens if we do get caught up in the dichotomy. In other words, that this is, in fact, a tacit reference to Kātyāyana.
Perhaps it is worth rehearsing why the denial of existence and non-existence is accurate and not at all paradoxical. My starting point, as always, is to take the subject under discussion to be experience. Being naive realists, or what the text calls "foolish, ignorant hoi-polloi" (bālapṛthagjanā aśrutavanto), we have an experience and we imagine ourselves to be in contact with something real, be it internal or external with respect to our first-person perspective. Ignoring what the experience implies about the world of sense experience, ignoring matters of ontology, the focus of the Kātyāyana is on the experience itself. Is the experience of an object an existing thing or a non-existing thing, irrespective of the nature of the object? Clearly, the answer is that it is neither. An experience cannot be described in terms of existence or non-existence. It arises, lingers for a short time, and then passes away. But the experience itself is entirely internal to us. Two people may see the same object and agree on its characteristics. But their experience of it is individual and cannot be agreed on. And thus Conze, in affirming both existence and non-existence, has come to precisely the opposite conclusion because he seeks a transcendent truth behind the words; a noumenon of the text. Conze's Romanticism has a Platonic flavour to it.
The answer that an experience is neither existent nor non-existent is important, because it is the understanding of the nature of experience that has soteriological value. We say that "things" are arising and passing away, but the Buddhist texts seem to refer mainly (if not solely) to the arising and passing away of experiences. In the Kātyāyana, it says that only dukkha arises and only dukkha ceases. The same point is made in the Simile of the Chariot. Dukkha, here, is a synonym for unenlightened experience. This search for understanding is deprecated by Conze, by modern Zen commentators, and many Tibetan lamas, because they, too, believe in a transcendent truth that requires the suspension of reason (as they conceive reason). In the Spiral Path texts the experience of liberation (vimutti) is initiated by becoming fed up (nibiddā) with the objects of the senses, i.e., with the intoxicating play of experience. Suspension of reason is not a prerequisite for awakening in these texts.
The answer that an experience is neither existent nor non-existent is important, because it is the understanding of the nature of experience that has soteriological value. We say that "things" are arising and passing away, but the Buddhist texts seem to refer mainly (if not solely) to the arising and passing away of experiences. In the Kātyāyana, it says that only dukkha arises and only dukkha ceases. The same point is made in the Simile of the Chariot. Dukkha, here, is a synonym for unenlightened experience. This search for understanding is deprecated by Conze, by modern Zen commentators, and many Tibetan lamas, because they, too, believe in a transcendent truth that requires the suspension of reason (as they conceive reason). In the Spiral Path texts the experience of liberation (vimutti) is initiated by becoming fed up (nibiddā) with the objects of the senses, i.e., with the intoxicating play of experience. Suspension of reason is not a prerequisite for awakening in these texts.
Central to Buddhist soteriology is the fact that our sense of self, our first person perspective, is also an experience, and partakes in the nature of all experiences. Streams of sensory information converge and are woven together to create the persistent illusion of being a self. Though, of course, we know that the illusion of the first-person perspective can be broken by drugs, trauma, brain injury and, of course, by meditation. In this view, insights consist of seeing experience, particularly the first person experience, in such a light that it ceases to intoxicate and fascinate. The word for 'insight', vipassana, literally means to 'see through', not, as our translation suggests, 'to see into'.
In our naivete about experience, we imagine each experience signifies something real and we respond to it as though it were real. But, in addition to this, we are burdened with ideas about what constitutes happiness as the goal of our lives. The unenlightened, the bālapṛthagjanā aśrutavanto, believe, deep down inside, that happiness is about having pleasant experiences and avoiding unpleasant experiences. Maximising the former and minimising the latter seems to be the operational definition of happiness. If we can only arrange things so that we have the optimum amount of both, then we will be happy and free of unhappiness. For most of us this means living in an unsatisfactory compromise and a lot of self-delusion about how happy we really are. Our pleasures do not satisfy. Our pains are all too many and not the least of them is mortality!
The line of thought in the Kātyāyana is often mixed up with attempts to apply dependent arising to all kinds of other processes, particularly karma and rebirth. And I have showed how this leads to inconsistencies and incoherent statements about the nature of the world across a number of essays (see the Afterlife tab for a list). Many Buddhists end up believing that the impermanence of "things" (e.g., tables, chariots, or other physical objects) is the key teaching of Buddhism, when it's just a truism that everyone is already aware of (See Everything changes, but so what?). The Kātyāyana is one of the texts where the intent of the idea, by which I mean the application to experience and only experience, is apparent. And it was this intent that was, I argue, taken up by the Aṣṭa and by Nāgārjuna some centuries later. Although there are many loud voices arguing about what Nāgārjuna meant to say in his very confusing opus, with most of them seeing Mūla-madhyamaka-kārikā as having deep ontological implications, I say that, in citing the Kātyāyana in the way that he does, we might understand his ideas better if we take the domain of application to be experience and forget about ontology. Nāgārjuna makes better, if not perfect, sense if we take him to be someone commenting on the phenomenology of experience, rather than speculating about metaphysics.
In the Aṣṭa version of the idea, the author has chosen to use the words that are tricky to translate while retaining the connotations of the original. So in a key passage (Aṣṭa 1.19.4) the Buddha says to Śāriputra:
na hi te śāriputra dharmāstathā saṃvidyante yathā bāla-pṛthagjanā aśrutavanto 'bhiniviṣṭāḥ | āyuṣmān śāriputra āha - kathaṃ tarhi te bhagavan saṃvidyante? bhagavān āha - yathā śāriputra na saṃvidyante, tathā saṃvidyante evam avidyamānāḥ | tenocyante avidyeti |
"For the dharmas do not exist (na saṃvidyante) in the way that the foolish, ignorant hoi-polloi take them to exist (abhiniviṣṭāḥ)."
Elder Śāriputra said, "How then do they exist (saṃvidyante), Bhagavan?"
The Bhagavan said, "They exist as though (yathā) they don't exist. Not-being found (avidyamāna) in that sense (evaṃ), they are said to be unfound (avidyā)."
The last statement in the Sanskrit text is:
yathā śāriputra na saṃvidyante, tathā saṃvidyante evam avidyamānāḥ | tenocyante avidyeti |
Conze translates "As they do not exist, so they exist. And so, since they do not exist [avidyamāna], they are called [the result of] ignorance [avidyā]", employing his usual hermeneutic of obscurity. He also translates avidyamāna as "do not exist", but avidye as "ignorance", but by his own logic the latter ought to mean 'not existing'.
Saṃvidyante is a passive form from sam√vid. Conze translates as ‘exists’. BHSD defines it as "is found, exists’ (= vidyate ‘is found; often virtually = asti)." PED saṃvijjati2 ‘to be found, to exist’. MW ‘know, recognise; perceive; approve’. It's tricky because there are two homonyms √vid meaning 'to know' (cognate with our word 'wisdom') and √vid meaning 'to find'. The two are indistinguishable except by context. The same goes for vidyamāna, a present participle 'knowing, finding' (here negated by the prefix a-). The other word here is abhiniviṣṭāḥ (abhi+ni√viṣ) which has a range of meanings 'entered or plunged into; intent on, endowed with; determined, persevering). Conze (1973a) suggests "settled down in, is accustomed to suppose."
So Conze is treating almost all the verb forms as meaning "exists". And we ought to point out that if a Sanskrit author wished to assert the existence of something they can do so very directly with the verb asti or some variation on √vṛt. So we need to be alert here to connotations. I think that √vid as found is relevant here. So, to say that if we go looking for a dharma is it not found, is not the same as saying it means it does not exist. We certainly have experiences, and so, to that extent, they do sort of exist. But when we say they "exist", we mean only that we have an experience, not that some kind of really existent entity has arisen and persists. Clearly, the author of the Aṣṭa has something very like the Kātyāyana teaching in mind. And the consequence is similar in the sense that it leads to two extremes of thought: that dharmas either exist or do not exist and all the problems that this causes. And note that the Two Truth argument adopts both extremes rather than avoiding either of them. Compare Aṣta 1.19.7:
Saṃvidyante is a passive form from sam√vid. Conze translates as ‘exists’. BHSD defines it as "is found, exists’ (= vidyate ‘is found; often virtually = asti)." PED saṃvijjati2 ‘to be found, to exist’. MW ‘know, recognise; perceive; approve’. It's tricky because there are two homonyms √vid meaning 'to know' (cognate with our word 'wisdom') and √vid meaning 'to find'. The two are indistinguishable except by context. The same goes for vidyamāna, a present participle 'knowing, finding' (here negated by the prefix a-). The other word here is abhiniviṣṭāḥ (abhi+ni√viṣ) which has a range of meanings 'entered or plunged into; intent on, endowed with; determined, persevering). Conze (1973a) suggests "settled down in, is accustomed to suppose."
So Conze is treating almost all the verb forms as meaning "exists". And we ought to point out that if a Sanskrit author wished to assert the existence of something they can do so very directly with the verb asti or some variation on √vṛt. So we need to be alert here to connotations. I think that √vid as found is relevant here. So, to say that if we go looking for a dharma is it not found, is not the same as saying it means it does not exist. We certainly have experiences, and so, to that extent, they do sort of exist. But when we say they "exist", we mean only that we have an experience, not that some kind of really existent entity has arisen and persists. Clearly, the author of the Aṣṭa has something very like the Kātyāyana teaching in mind. And the consequence is similar in the sense that it leads to two extremes of thought: that dharmas either exist or do not exist and all the problems that this causes. And note that the Two Truth argument adopts both extremes rather than avoiding either of them. Compare Aṣta 1.19.7:
kalpayitvā dvāv antāv abhiniviśante abhiniviśya tan nidānam-upalambhaṃ niśritya atītān dharmān kalpayanti, anāgatān dharmān kalpayanti, pratyutpannān dharmān kalpayanti te kalpayitvā nāmarūpe 'bhiniviṣṭāḥ |
Having imagined [the non-existing dharmas], they are engrossed (abhiviviśante) in the two extremes (dvāv antāv); being engrossed (abhiniviśya), they rely on the observed object as a basis and imagine dharmas in the past, in the future, and in the present moment. Having imagined them, they become engrossed in name and form.
Note the recurrence of abhini√viṣ here, translated as 'engrossed' this time (and as "settled down" by Conze). Taking dharmas to be real, settling into a view, we make mistakes about the nature of experience and, by implication, suppose that sense experience can be ultimately satisfying. And this is categorically a mistake.
It has been argued that the Aṣṭa contains no direct reference to the Sarvāstivādin doctrine of sarva-asti (always existent), but Aṣṭa 1.19.7 might be just such a reference. Here, the deluded people imagine (kalpayanti) that dharmas exist in the past, future and present. This is precisely what Sarvāstivādins believe. If we recall the problem of Action at a Temporal Distance, this 'always on' feature of dharmas was the Vaibāṣika solution to the disconnect between action and result in time that the doctrine of karma requires. It earned the Vaibāṣikas the nickname Sarvāstivāda. However, after examining two of the early translations T224 《道行般若經》by Lokakṣema (179 CE) and T227 《小品般若經》by Kumārajīva (408 CE), both make the point about the two extremes, but neither of them have this passage about past, future and present. So we must conclude that it was interpolated into the Sanskrit text at a later date. So, if criticism of Sarvāstivāda was intended, it was not part of the original intention. Kumārajīva's translation of the dvāv antāv 'two extremes' is prosaically 二邊 'two extremes', whereas Lokakṣema has the more interesting 兩癡耳 literally 'two insane ears'.
Taking the text on face value, the criticism of the two extremes (existence and non-existence) is tilted towards criticising existence, presumably precisely because the existence view was prevalent at the time. If this interpretation is correct, then it may help explain the idiom in the next sentence (1.19.8)
Taking the text on face value, the criticism of the two extremes (existence and non-existence) is tilted towards criticising existence, presumably precisely because the existence view was prevalent at the time. If this interpretation is correct, then it may help explain the idiom in the next sentence (1.19.8)
tair asaṃvidyamānāḥ sarvadharmāḥ kalpitāḥ |
All dharmas imagined by them are non-existing.Kalpita is a past participle from √kḷp. The literal meaning is 'made, fabricated'. I'm presuming here that the fabrication is a mental one. There's not really a word for "imagination" in Sanskrit (one of many differences in how they understand mind). Again, the idea here seems to be that one has an experience and in the way of naive realism mistakes it for something more substantial than it is. And when we treat experiences this way it obscures the Buddhist path or, as Aṣṭa puts it, yathābhūtaṃ mārgaṃ na jānanti na paśyanti, 'Not knowing or seeing the path as it really is...'
Interestingly, in 1.19.12 the wrong view is seen as an impediment to the development of faith in the truth of the dharma (satyaṃ dharmaṃ). This suggests that the mistake is foundational and must be sorted out right at the beginning of the religious life. My sense is that most modern Buddhism is already lost in speculation about ontology and supernatural forces. As Justin Whitaker recently pointed out to me, most Buddhists and scholars still invoke some variation of "seeing reality as it is" when describing Buddhist soteriology. But reality implies existence. Whatever we see as it is (yathābhūta) cannot be described in terms of existence or non-existence and, therefore, is neither real nor unreal. Reality can have nothing to do with Buddhist soteriology, by definition. To be real, whatever it is would have to be permanently existing and I don't think I need to explain why that is a problem.
I hope I have showed that at the very least the author of Aṣṭa had Kātyāyana in mind as they were writing this section. I think this shows that at least at the beginning of producing the Prajñāpāramitā texts the authors saw the domain of application of the Dharma as experience. They were not caught up in the metaphysical speculations of the Ābhidharmikas. They were, however, caught up in their own metaphysical speculations about the nature of the Buddha, though that is a story for another time. The importance of this discovery is that it helps us to understand the apparently paradoxical texts of the Prajñāpāramitā literature. In flirting with paradox they were trying to describe an attitude towards experience that had a liberating effect. They did not set out to confuse the reader, but to draw attention to our suppositions about experience and reality. The former we can know and understand; the latter we can only make inferences about, based on the commonality of experience with reference to the same object.
A first step in reforming modern Buddhism would be to establish the domain of application of our theory and practice, and in such a way as our theory and practice were complimentary. Despite all the bitching from Buddhists about the Mindfulness Therapy movement, I think they have a much better handle on this focus and integration of theory and practice. Better to be working with experience in a shallow way than to have a deep engagement with the kind of ontological speculation that typifies contemporary Buddhists discourse, because the latter is not beneficial in any way while the former at least is mildly beneficial and creates a basis for progress.
I hope I have showed that at the very least the author of Aṣṭa had Kātyāyana in mind as they were writing this section. I think this shows that at least at the beginning of producing the Prajñāpāramitā texts the authors saw the domain of application of the Dharma as experience. They were not caught up in the metaphysical speculations of the Ābhidharmikas. They were, however, caught up in their own metaphysical speculations about the nature of the Buddha, though that is a story for another time. The importance of this discovery is that it helps us to understand the apparently paradoxical texts of the Prajñāpāramitā literature. In flirting with paradox they were trying to describe an attitude towards experience that had a liberating effect. They did not set out to confuse the reader, but to draw attention to our suppositions about experience and reality. The former we can know and understand; the latter we can only make inferences about, based on the commonality of experience with reference to the same object.
A first step in reforming modern Buddhism would be to establish the domain of application of our theory and practice, and in such a way as our theory and practice were complimentary. Despite all the bitching from Buddhists about the Mindfulness Therapy movement, I think they have a much better handle on this focus and integration of theory and practice. Better to be working with experience in a shallow way than to have a deep engagement with the kind of ontological speculation that typifies contemporary Buddhists discourse, because the latter is not beneficial in any way while the former at least is mildly beneficial and creates a basis for progress.
~~oOo~~
Bibliography
Conze, Edward. (1973). The Perfection of Wisdom in Eight Thousand Lines and its Verse Summary. San Francisco: City Lights.
Conze, Edward (1973a) Materials for a Dictionary of Prajñāpāramitā Literature. Suzuki Research Foundation.
Drewes, David (2009). Early Indian Mah ay ana Buddhism I: Recent Scholarship.Religion Compass 4/2 (2010): 55–65, 10.1111/j.1749-8171.2009.00195.x. https://www.academia.edu/9226456/Early_Indian_Mahayana_Buddhism_I_Recent_scholarship
Kalupahana, David J. (1986) Nāgārjuna, The Philosophy of the Middle Way: Mūlamadhyamakakārikā. SUNY.
Salvini, Mattia. (2011) The Nidānasamyukta and the Mūlamadhyamakakārikā: understanding the Middle Way through comparison and exegesis. Thai International Journal of Buddhist Studies.II: 57-95. https://www.academia.edu/1925584/The_NidÄnasamyukta_and_the_M_lamadhyamakakÄrikÄ_understanding_the_Middle_Way_through_comparison_and_exegesis
Vaidya, P.L. (1960) Astasahasrika Prajnaparamita. Darbhanga: The Mithila Institute. (Buddhist Sanskrit Texts, 4). http://fiindolo.sub.uni-goettingen.de/gretil/1_sanskr/4_rellit/buddh/bsu049_u.htm
Sanskrit text
Aṣṭa 1.19. (Vaidya 1960)
evamukte āyuṣmān śāriputro bhagavantam etad avocat – evaṃ śikṣamāṇo bhagavan bodhisattvo mahāsattvaḥ katamasmin dharme śikṣate? evam ukte bhagavān āyuṣmantaṃ śāriputram etad avocat evaṃ śikṣamāṇaḥ śāriputra bodhisattvo mahāsattvo na kasmiṃś cid dharme śikṣate | tatkasya hetoḥ? na hi te śāriputra dharmāstathā saṃvidyante yathā bāla-pṛthagjanā aśrutavanto 'bhiniviṣṭāḥ | āyuṣmān śāriputra āha - kathaṃ tarhi te bhagavan saṃvidyante? bhagavān āha - yathā śāriputra na saṃvidyante, tathā saṃvidyante evamavidyamānāḥ | tenocyante avidyeti | tān bālapṛthagjanā aśrutavanto 'bhiniviṣṭāḥ | tair asaṃvidyamānāḥ sarvadharmāḥ kalpitāḥ | te tān kalpayitvā dvayor antayoḥ saktāḥ tān dharmān na jānanti na paśyanti | tasmāt te 'saṃvidyamānān sarva-dharmān kalpayanti | kalpayitvā dvāv antāv abhiniviśante abhiniviśya tan nidānam-upalambhaṃ niśritya atītān dharmān kalpayanti, anāgatān dharmān kalpayanti, pratyutpannān dharmān kalpayanti te kalpayitvā nāmarūpe 'bhiniviṣṭāḥ | tairasaṃvidyamānāḥ sarvadharmāḥ kalpitāḥ | te tān asaṃvidyamānān sarvadharmān kalpayanto yathābhūtaṃ mārgaṃ na jānanti na paśyanti | yathābhūtaṃ mārgam ajānanto 'paśyanto na niryānti traidhātukāt, na budhyante bhūtakoṭim | tena te bālā iti saṃjñāṃ gacchanti | te satyaṃ dharmaṃ na śraddhadhati | na khalu punaḥ śāriputra bodhisattvā mahāsattvā kaṃcid dharmam abhiniviśante ||
PS. If anyone has a pdf of Conze's Sanskrit edition of Aṣṭa I'd love to get a copy.