1. Introduction
3.3 Yogācāra Dignāga and the Epistemological Turn in Buddhist Philosophy, Especially
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
3.3 Yogācāra Dignāga and the Epistemological Turn in Buddhist Philosophy, Especially Regarding the Non-temporality of Cognitive Causality and the Allowability for a Holistic Model of Epistemology in Contrast to
Proceduralistic Model in Time
As we have learned from the Kātyāyana passage in the Saṃyuktāgama 12.301,30 cited in Nāgārjuna's MMK 15.7, it is at the point of the “contact (觸, sparśa)” of the three parties, viz., (1) the cognitive faculty (根), (2) the cognized object (境) and (3) the consciousness (識), that cognition is realized. Regardless of the metaphysical assumptions one would like to assume for this scenario, this is the common understanding with which I believe no rational epistemologist would disagree.
External realists in India, notably Naiyāyikas, Sarvāstivāda and the Sautrāntika, maintain some realistic and causal account of the contact, meaning that the whole story is founded in the self-existing world where the self-existing atoms dwell, and that the atoms, aggregates of atoms, receptive faculties, understanding and the resulting cognition are linked via self-existing causal relations in a real causal chain or complex. Such a realistic and causal account of the contact showcases the
ontological positivism in epistemology. The distinct anti-external realism in Buddhism, inaugurated by Nāgārjuna, prompted his interlocutors to fixate on the contradiction between the causality among the (postulated) causes of cognition and the causality among the cognized things (the phenomenal causal exhaustion);
consequently, the external reality of the causes of cognition, including both the objective ones (atoms) and the subjective ones (pramāṇa-s), is rejected in order to preserve the phenomenal causal exhaustion (representatively in MMK 1).
Already in the Saṃyuktāgama 13.306, the Buddha said, “ visual consciousness arises from taking the vision faculty and color as its objects; the three combined together is the contact (觸, sparśa); together with the contact arise sensation (受, vedanā),
30 T02n0099_p0085c17(00) – p0086a03(10).
‧
apprehension (想, samjñā) and mental formation (思, saṅkhāra/samskāra/ cetanā) come as well.”31 The fact that cognition and the other mental activities become possible via the contact is thus a lucid common understanding among Buddhists.
Another common understanding, aside from the crucial role of contact, is the crucial role of manas in the relation between the contact and skandas. In the
Madhyamagama 58.21, it was said by the Venerable Mahākoṭṭhita that while each former five sense faculty possesses its own domain of mental formation and its own domain of objective appearance, manas, also in possession of its own domain, will apprehend the objective appearance of the other five and is that upon which the others rest.32 The weaker implication of this common understanding is that it is impossible for us to imagine any experience without being accompanied by the function of manas; the stronger implication of it is that the other dimensions of experience depend on the practice of the function of manas.
The development of various theories on how one should relate the contact and these skandas and how manas is related to the other factors of experience led directly to the split of Buddhism into different schools. According to Ven. Yinshun
(1949/2000:114), Sautrāntika thinkers, following the Saṃyuktāgama 13.30733, maintained that contact itself is exactly the consciousness that ensues at the point of contact, while the Sarvāstivāda maintained that contact and consciousness are two different parties in correspondence and that contact derives from the aggregate of the three parties, with contact itself also serving as the mental factor (⼼所,
31 T02n0099_p0087c26(02) – c27(02): “眼、⾊緣⽣眼識，三事和合觸，觸俱⽣受、想、思.”
32 T01n0026_p0791b16(01) – b17(03): “Each of these five faculties has its own domain of mental formation and its own domain of objective appearance (the domain of color, sound, smell, taste and touch). All of these domains of objective appearance are to be apprehended by manas. Manas is what the others rest upon. 此五根異⾏、異境界，
各各受⾃境界，意為彼盡受境界，意為彼依.”
33 T02n0099_p0088b01(00) – b06(00): “The conditions [of consciousness, say, vision-consciousness], vision faculty and color, come from the citta-caitta distinction (the provisional separation in our understanding of the two functions of the single mind: knowing and being known). The condition via which come consciousness, contact, and the co-arising sensation, apprehension, etc., are neither myself nor the other object (我所, mama-kāra), nor the reincarnating self (福伽羅, pudgala) …. This is the arising and ceasing; this is the way the suffering-skanda alter.
These (the epistemic terms) are postulated in thought and implemented in every aware being (眾⽣, sattva) 眼⾊⼆
種緣，⽣於⼼⼼法。識、觸及俱⽣受想等有因，非我非我所, 亦非福伽羅...是則為⽣滅，苦陰變異法，於 斯等作想，施設於眾⽣.”
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
caitasika/caitta) of the three parties.34 Regarding the relation between manas and the resulting consciousness, some maintained that the manas-consciousness at the
previous time point results in the present consciousness (past manas過去
意/immediate abdicating manas 無間滅意), others maintained that the appearing manas results in the appearing consciousness (appearing manas現在意)35, and still others maintained that both of the aforementioned claims are reconcilable. As early as the Hināyāna era, two preliminary models to understand the epistemic process had emerged: epistemic proceduralism and epistemic holism. In proceduralism, which sticks to the Saṃyuktāgama 13.306 alone, the purposive relations among the skandas and among the three parties of faculty, object and consciousness can be easily
demonstrated and understood in a procedure in temporal and causal sequence. Since the relations are explained in a procedure, time differences are tacitly assumed – even when the difference is zero (immediate), the cause's coming before the result is still assumed. Nonetheless, the time difference is always too short for the ordinary self-retrospection to look after, so they are simply bypassed without further investigation.
Naturally, the procedure of the skandas in time invites the causal involvement with the external causes that have impacted themselves and triggered their functioning, and hence the schools of proceduralism would maintain an ontology that assumes the ultimate reality of self-existing atoms in the ultimate self-existing world of space and time, in which the epistemic procedure itself also takes its own place in the spatial form (the subject's experience is internal while the objects are external) and the temporal order (the existence of the object is temporally prior to perception and then perception is temporally prior to comprehension). That is, the ultimate reality that is independent of cognition is assumed. However, if we consider the Saṃyuktāgama 13.306 together with 13.307, we see that even when the segments of the entire
procedure can be thought of in relations or stages, the separations are only reasonable
34 Ven. Yinshun: “Sarvāstivāda maintained that consciousness and contact are two [separate parties] and are
correspondent at the same time; hence, touch comes from the aggregate of the three parties, and [contact] is also the caitta that makes the three aggregated 有部以識及觸為⼆ ，又是同時相應的︔所以觸從三和⽣，又為令三和合 的⼼所.”
35 “Xian-zai 現在” may or may not imply the temporal condition “now.” Here, since the cause and the result are regarded as a single entity, to understand “Xian-zai 現在”as “appearing” is more appropriate than as “present.”
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
understanding in our thought alone that do not have any real impact upon the non-differentiated epistemic unity. Yet another implication from these understandings, that these postulated elements do exist and function independently from one another in some real causal relations, is simply wrong. Here, when examining the source of the idea for the pratyakṣa-anumāṇa distinction in Dignāga, especially in light of highlighting his redefinition: we have to have both (1) the undifferentiated direct experience and (2) the discursive understandings (indirect experience) of it in thought; each has its own source (the direct knowledge and indirect knowledge are sharply different in the two means), each is directly accessible in the consciousness, but each is not realized unless in the resulting cognition in the necessary cooperation with the other counterpart.
The theories of the proceduralism do not actually insist on the presumption of time (otherwise, the time differences would not be bypassed), because if it is granted that the procedure has to be real in time after all, the demand that the contact itself has to be caitasika becomes puzzling: the contact of the three parties are the cause and the result is cognition, while the contact itself is required to be caitasika as well, which is the object of the resulting cognition. We can pinpoint this difficulty in the
controversy over the role of manas in the resulting consciousness in the following conceptions: 過去意/無間滅意 vs. 現在意. This is also a problem that we can encounter when commenting on Dignāga's theoretical demand that there should be no separation among pramāṇa (the means of cognition), prameya (the object of
cognition) and phala (the resulting cognition). Perhaps the proceduralists had indeed attempted to remove the presumption of time differences with the very weak
argument that they are too short and thus ephemeral for our ordinary cognition to note, but the removal of the presumption was neither clear nor sufficiently
emphasized so that even when Dharmarkīrti (as well as his followers, in the common scholarly recognition during our times) attempted to develop his (their) interpretation
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
of Dignāga's epistemology upon a position inclining toward the Sautrāntika,36 strong proceduralism was still modeled in his epistemology without hesitation, with even the real causal account of the contact taken to be the warrant of the correspondence between the object known and the knowledge about it. On the other hand, the preliminary epistemic holism in Buddhism had not been able to provide sufficient arguments as to why the epistemic process can be thought of in separate parts but remains holistic itself, when the realistic causal account of the contact of the three parties was still tacitly assumed in the theories as well at the time.
The problem with time here is also troubling to us with regard to the practical goals of Buddhism. The order of the first two of the five universal mental factors (遍⾏
sarvatraga, namely, attention 作意 manasi-kāra, contact 觸, sensation 受,
apprehension想, and mental formation 思) had already been controversial during the Hināyāna era. Some observed and reported that attention comes after contact, while others held the opposite to be true. If we stick to the strong implication of the
Madhyamagama 58.211, namely that manas is that which the rest of the mental activities rely on, then we may maintain that attention (manasikāra) should come first; however, if we stick to the Saṃyuktāgama 13.306, then we would say that without the contact, nothing is possible for us to know, and then we would maintain that the contact should come first. Regarding the practical goal – to cease sufferings in saṃsāra, the way out is to stop taking the contact (to stop sensation, vedanā), as suggested in the Saṃyuktāgama 12.301 and MMK 15.7; if the contact is the very initial of everything else, ultimately speaking, then the practice is impossible, because autonomy would then become impossible in the sense that the function of manas in this order has to have been determined through the heteronomous causal account in the contact (where consciousness itself should not be possible yet, in the strict
procedualistic sense), and thus been conditioned by the external causal environment.
On the contrary, only when attention (manasikāra) is, or at least is allowed to be, the
36 For example, we continue to come across such an attempt even up to Jenendrabuddhi (PSṬ 70,11).
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
very initial of everything else, can the mental procedure be managed by the
autonomous consciousness, which itself is also a part of the contact (autonomous consciousness is not necessarily followed by the acceptance of the self, not even in the Kantian sense). Then, the proceduralism, together with its already-realistic causal account of the three parties and the five universal mental factors, becomes very unfavorable to the practical concern in Buddhism; it is thus no wonder that the epistemic system that is modeled in proceduralism is difficult to fit into the practical project of Buddhism, as we have learned from Stcherbatsky's report on the three schools of interpretation regarding Dharmakīrti's interpretation of Dignāga (1932: 39-46). On the contrary, the removal of the assumption of time differences (and the realistic causal account) in the strong holism allows us to understand the order of the five universal mental activities from a simple logical or conceptual point of view in thought alone, and this on the one hand preserves the possibility of the autonomy of the mind (not all the processes in the procedure have to have really been determined during the contact in time) and on the other hand stays concerted with the necessity of the contact as the necessary condition of the other mental activities. In this regard, the two common understandings (the necessity of the contact and the necessity of manas) can interlace their best theoretical performance. Nāgārjuna's direct rejection of the ultimate reality of any self-existing causes of cognition (both the atoms and self-exiting pramāṇa-s) was exactly the fruit and the force field that such a theoretical performance anticipated.
Nāgārjuna's rejection of the idea of the self-dependent cause of cognition in those schools and his preservation of the two aforementioned common understandings altogether should have radicalized Buddhist epistemology into an anti-realistic feasibility anticipating an epistemological turn. The rejection of the ultimate reality of the absolute cause of cognition, including both the self-dependent atoms and the self-dependent pramāṇa-s, ruled out the time presumption in the operation of the epistemic system for good, engendering the very strong version of Buddhist
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
epistemic holism. When, on the one hand, cognition is sourced neither from the self-dependent world nor from the self-existing s (the proceduralistic pramāṇa-s), and, on the other hand, the necessary condition of the contact and the other
necessary condition of manas have to be taken into account, the reality in the
resulting cognition cannot but be understood to be sourced from the way we cognize and experience the world. Cognition in this sense (when the principle of emptiness, whereby no external causal foundation is possible, is strictly followed) does not seek to represent the world in accordance with the way it is by itself being warranted via causal relations; on the contrary, it seeks to mold a world in accordance with the necessary way you can cognize at all. Objectivity is still publicly observable in the object cognized, not because it so stands by itself but because the object has to be cognized in it (objectivity) in a certain group of observers who share the same capacity for cognition. In this scenario, the determining relation between the
cognitive forms (the necessary ways through which any cognition has to be formed) and the forms of the object of cognition is then causal not in the sense of reality (the phenomenal causality) but in the sense of a different kind of causality – Dignāga's idea of formal conformity (PS(V) 1.9) is a great candidate for this idea. Human beings experience the world in three dimensions while ants experience it in flat, two dimensions and there exists no self-existing, cognition-free foundation that has to be common for all possible kinds of subjects of experience. Communicability depends on how much one kind of the subject of experience shares with the other in the cognitive forms, not on the ultimate or on the cognized (and hence not-yet-established) self-existents. The very end of the epistemological turn is the idealism of the cognitive factors, that they cannot be really (ontologically) prior to but only ideally (epistemologically) prior to experience – they do not really but ideally cause and condition their resulting cognitions, in another kind of causal relation, while the necessary conditioning of these cognitive factors upon the resulting cognition ensures the limited reality of experience, i.e., the valid objectivity in the resulting cognitions together with the phenomenal causal relations in them. And the holism contains both
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
the ideal cognitive factors and the undifferentiated nature of reality in experience which can be thought of discursively in parts. That should be the basic narrative of the two pramāṇa-s, when the Yogācāra Dignāga is reconstructed.
Whether Vasubandhu and Dignāga totally agreed with Nāgārjuna on the rejection of the atoms and pramāṇa-s, or in what sense there could be some agreement among them, is not directly clear from the literature; we shall leave the questions to the historians and philologists. But the epistemic idealism, established by Vasubandhu and thought to be followed by Dignāga, is exactly the outcome of, and the response to, Nāgārjuna's challenges to non-Buddhist epistemology. On the other hand, the success of the strategy of the communication platform contributed by Dignāga can then be drawn as such: when everyone agrees with the common understanding, viz., the ontological neutralism, the consequences of which should satisfy any rational epistemologist who is content with the validity of the phenomenal causal exhaustion and its valid relationship to knowledge, the inadequacy of the non-Buddhist
ontological positivism will be revealed so that the philosophical position of
Buddhism manifests the silent persuasion by way of better accounting for the limited reality and better sidestepping the contradiction between the phenomenal reality and self-existing reality, as Nāgārjuna had discharged his strongest arrow at this very issue. Remarkably, Dignāga in his Traikālyaparīkṣā refutes the ultimate reality of time difference (past, presence and future) upon the Mahāyāna Buddhism belief that
“all existences are empty and with no self-nature” (Nasu, 2015: 13; Cf. Hattori 1961);
both their holding an anti-realistic position on time and their championing the code that all existences are empty and with no self-nature together endow us with more philosophical grounds and greater impetus to align together Dignāga's epistemology and Nāgārjuna's criticism against epistemology.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
4 The Ontological Neutralism in Dignāga's Epistemology – Critical Reconstructive Interpretation of Related Passages in NMukh and PS(V)