1. Introduction
3.2 Plural Vasubandhus and Dignāga
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
suggestion regarding the pressure of Nāgārjuna's contemporary philosophical
atmosphere is so sympathetic that we cannot help but admit to its persuasion. Thus, the attempt of Svātantrika to weave Nāgārjuna's philosophical reworkings back into the context of epistemology (the Kātyāyana passage) with the help of the Buddhist theories of epistemology, especially Dignāga's, has proven very valuable to us, though this line has historically been overshadowed by the smart strategy of
Nāgārjuna that is so much more appreciated in the other line that it is inconceivable for them to also appreciate the value of epistemology and to admit to the question sidestepped by Nāgārjuna. On the contrary, the theories of Buddhist epistemology have to survive the challenges of Nāgārjuna, in the sense that they cannot maintain any determined absolute cause (be it objective or subjective) and any determined causal relation, and they, along with accounting for the impermanent nature of phenomenon, have to be able to positively account for the necessity of the relative causal network in phenomenon, if we want to maintain the philosophical unity of Mahāyāna Buddhism.
This is the wider context in which Dignāga presented his epistemology.
3.2 Plural Vasubandhus and Dignāga
The term hetuvidyā (因明) refers to the study of valid cognition in accordance with hetu 因 (cause and reason). Very early in ancient India, there already existed several different philosophical schools abundantly skilled at the utilization and development of proper inference, rhetoric and even sophistry to assist with their internal and external debates. The Indian study hetuvidyā was thus born in this intellectual womb.20 The study became more systematic when Gautama, one of the six
forefathers of the Nyāya school, resolved to investigate this world; later, Mahāyāna
20 Katsura (2010:1).
‧
Buddhism adopted and furthered the study, with the goal of helping with the attainment of the Wisdom of All Seeds (sarvathā-jñāna) of the Buddha.21 The essential difference in the development of Mahāyāna Buddhism from its
counterparties, which was especially arched in the hands of Dignāga, is the rejection of the ultimate reality (unconditional reality) of hetu; instead, the warrant of hetu is supplied ideally, in a different (cognitive) causality other than the causality that possesses physical efficacy, from the means of cognition (pramāṇa-s) in Dignāga,22 via which cognition has first become possible (realizable) and justifiable accordingly.
Such a warrant in Dignāga is so strict that “very few of our opinions can measure up to them” and “Dharmakīrti tended to see this strictness as a weakness in Diṅnāga's logic” (Hayes, 1988: 35), because the idealistic position has been so ignored by most of us and by Dharmakīrti that we do not know how to get over from the inductions to the strict universal claims in Dignāga's hetuvidyā from our purely empiricist
perspective with the assumption of the external, unconditional reality of hetu.
The outline of the study of hetuvidyā within Buddhism, as painted by Cheng Lü in his Yin ming gang yao (1977: 2 – 7; following Yijing's Nan hai ji gui nei fa zhuan 4.34), is a lineage from the Buddha through the Abhidharma texts, Nāgārjuna, Maitreya,
21 Cheng Lü in the introduction to his Yin ming gang yao, following Yogācārabhūmiśāstra (vol. 15),
described:“Yinming refers to the study of observation and inference. The determination the validity of observation and inference has to ground on 'yin', the cause or reason (因, hetu). 'Ming' (vidyā) generally refers to a systematic study or art in Sanskrit. Hence, the study of the observation and inference is termed as yinming (因明, hetu-vidyā).
In ancient India, there were plenty of schools and the debates propagated; consequently theories of the study were developed. The school of Akṣapāda maintained that one comes to cognize with means of cognition (量, pramāṇa);
the further analysis of the [employment and function of] the means of cognition contributes the formulae of inference. The comprehensive and knowledgeable Mahāyāna Buddhism also investigates all kinds of formulae of inference in the field of the study. Both of the study of Akṣapāda and the study of Buddhism are called 'yinming', but we divide them into inner study and outer study. The study of the school of Akṣapāda is outer study. The study of Buddhism furthering from the former is inner study. The outer study cannot but invokes endless debates; the purpose of the inner study aims at attaining Buddha's Wisdom of All Seeds (sarvathā-jñāna). 因明者，察事辯理之 學也。辯察據因，以判真似，故研其法要在研因。明處梵⾔，通稱學藝，遂⽬斯學以為因明。 在昔天竺，
宗計既繁，詭辯隨滋，漸⽣「因論」。⾜⽬學派，以量致知，辨析增詳，爰陳軌式︔佛家⼤乘，博學多聞，
復從明處，賅諸論法。凡是皆謂因明，⽽內外為異。因論之說，⾜⽬之學，外也。佛家因外⽽別詳之，內 也。外論不免究竟興諍，內學乃期⼀切種智（sarvathā-jñāna）.”
22 NMukh (Katsura's translation): “now this [pakṣadharma i.e., hetu] is only depending upon jñāpaka-hetu (證了因). (今 唯此依證了因故 Cf. PS II 6b: jñāpako 'dhikṛto yataḥ /). Only by means of the power of cognition it makes known the meaning/content of what is stated (但由智力了所說義); unlike kāraka-hetu (生因) it does not [make known the meaning of what is stated] by means of producing some use (or useful thing 用). (非如生因由能起用).” Zangyao: “今 此惟依證了因故，但由智力了所說義，非如生因由能起用.”
‧
Vasubandhu, and Dignāga till Dharmapāla, Bhaviveka, and Dharmakīrti.23 That is to say, Buddhist hetuvidyā is rooted in the Buddha and passed down in the hands of Abhidharma and Madhyamaka before flourishing in Yogācāra. Lü concluded that the complete set of formulations had already been laid out in Vasubandhu; the study was established by Dignāga and became communicable with the non-Buddhist after Jinendrabuddhi's promotion.24 Although the slot of Madhyamaka in the middle of the continuation of the lineage might trouble our contemporary scholars and the non-Chinese Buddhist thinkers, the lineage is indeed quite commonly acknowledged, from the perspective within classical Chinese Buddhism, and also quite reasonable, especially from the perspective of our putting Dignāga's epistemology in the position of responding to Nāgārjuna's rejection of the realistic epistemology – i.e, rejections of the unconditional reality of hetu – and the position of incorporating the thesis of freedom into the grand project of the Mahāyāna Buddhist treatise of the antinomy between freedom (the possibility of getting out of saṃsāra25) on the one hand and the phenomenal causal exhaustion (the universality of causal relations in all appearances) that is ascertained in the necessary role of two pramāṇa-s, on the other hand.
23 Cheng Lü divided the lineage into five period: (1) since the Buddha to Aśvaghoṣa: formulations of inference were used in the four Āgama sutras and the Abhidharma texts; (2) since Nāgārjuna to Pingala: formulations developed in the Mūla-madhyamaka-kārikā, Dvādashanikāya-Shāstra, etc.; (3) since Maitreya to Vasubandhu and his pupil Guṇamati: the set of formulations was laid out (論軌具備) in the Yogācārabhūmi-śāstra, Xian yang sheng jiao lun (顯 揚聖教論), Fang bian xin lun (⽅便⼼論), and Ru shi lun (如實論); (4) since Dignāga to Bandhu-prabha and
Asvabhāva: the study was established – representatively in the Nyāyamukha and Nyāya-praveśatāka-śāstra, Da sheng guang bainlun shi (⼤乘廣百論釋), and Prajñāpradīpamūlamadhyamakavṛtti ; (5) since Dharmakīti to Tianxhi (天喜, 1500 years after the Buddha's nirvana): revival of the study.
24 Cheng Lü's Yin Ming Gang Yao: “The emerging Mahāyāna Buddhism vastly refuted the non-Buddhist and
Hināyāna [theories of logic]. The formulations of logic in Madhyamaka theories were enriched. Since Maitreya, his junior fellow apprentice to Vasubandhu, the set of formulations was laid out in order to retackle the wrong view on emptiness. Dignāga's reformation [of the theory of logic] and Jinendrabuddhi's promotion [of that] made the [own theory of logic] communicable with the others' [theories] and became a new school. Afterwards, the theory was cleverly employed by Dharmapāla and Bhāviveka in order to ascertain the to-be-established and the to-be-refuted;
Dharmakīrti and Prajñākaragupta re-unfolded the profound meaning [of the theory] ⼤乘初興，盛破外⼩。中觀豬 論，法式稍詳。慈氏師弟，爰及世親，復㨀惡空，範規乃具。至於陳那改作，天主闡揚，遂能融洽外⾔，
成⼀家之說。然後護法清辨妙嫻⽴破，法稱慧護重顯幽微.”
25 In the vṛitti of PS(V) 1 (Hattori, 1968: 23), Dignāga wrote: “I express praise in honor of the Worshipful [Buddha] in order to produce in [the hearts of] men faith in Him who, because of His perfection in cause (hetu) and effect (phala), is to be regarded as the personification of the means of cognition (pramāṇa-bhūta).” Here, “cause,” as Dignāga himself clarified, refers to the perfection in “intention” (āśaya) and the perfection in “practice” (prayoga), whereas, “effect” refers to “the attainment of His own objectives (svârtha) as well as those of the others (parârtha).”
Among the three listed senses of the attainment, we especially note the attainment in the sense of “being beyond a return [to saṃsara], as one who is fully cured of a fever.”
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
From Cheng Lü's outline, we see that the tradition of the classical Chinese Buddhism regards Dignāga as the pupil of Vasubandhu and belonging to the school of
Yogācāra; that is, Dignāga in the Chinese tradition is regarded as an idealist who holds that reality is only appearance (i.e., the applicative scope of hetuvidyā is only the resulting cognition which is in the form of appearance and concept) without needing any transcendent external support (the unconditional reality of hetu). This tradition holds that Vasubandhu started out in the school of Sarvāstivāda, concluding and commenting on the theories of Sarvāstivāda in his Abhidharma-kośa, before developing into a different philosophical position in his Viṃśatikā-vijñapti-mātratā-siddhi and Triṃśikā-vijñaptimātratā-Viṃśatikā-vijñapti-mātratā-siddhi and thus founding the school of
Yogācāra.26 Frauwallner (1951) was the first to question this story and suggested that there were actually two Vasubandhus, one the Vijñanāvādin Vasubandhu (Asaṅga's brother) and the other the Sautrāntika Vasubandhu who retained this position up till his death,27 because the two positions were far too incoherent with regard to the texts and the philosophical systems. It is really not bizarre for a thinker to change his mind on a certain position as the ideas develop over time, especially so in the case of a revolutionary thinker. This is not to mention how reasonable it is that the
continuation between an original master and her/his also original pupils can be randomized either as a fact of repetition, revision or revolutionary development over the course of the real history. No matter which number in the end we will accept, the distinction between the early Vasubandhu and the later Vasubandhu remains sharply distinct; hence this historical question about the number should not worry us too much here. However, it is philosophically interesting to look into how Dignāga could be relocated in the transition between Sarvāstivāda and Vijñanāvāda, so that we can better see that the possibility of Dignāga (a follower of Vasubandhu) being a
Yogācāra thinker cannot be ignored.
26 Cf. Hayes (1945:96).
27 Singh (1984) believed that Frauwallner changed his mind and thought that the Sautrāntika Vasubandhu in the end became a Yogācāra thinker (1969), probably as the result of the attacks and revisions he received from P.S. Jaini (1957), Alex Wayman (1961) and Schimidthausen (1967). However, Hayes (1986: 168) considered Singh's account of Frauwallner's two-Vasubhandhu thesis to be “garbled.”
‧
We do not thus agree with Singh (1984, 1995) who claimed that the determination of the actual historical number of Vasubandhu and whose position was actually taken up by Dignāga is very significant in understanding Dignāga's philosophy. The
repercussions of Singh's suggestion, however, drew our attention to the new
scholastic development during the 80s when scholars became interested again in the significance of Dignāga's systematized epistemology in Buddhism, especially how it confronts the external challenges and prepares itself to reach out for communications as a developed philosophical union. Now the scenario is no longer that Buddhism scattered its critics or challenged the non-Buddhists in its youth but that it has evolved to a stage whereby it forms an array receiving external challenges and the revival of Brahmanism. Over the subsequent two or three decades after the issue was first raised by Singh, we have arrived at a much clearer understanding of the long-ignored dramatical difference between the popular Buddhist-realistic epistemology in Dharmakīrti and the overshadowed original epistemology in Dignāga. The
Sautrāntika assumptions that there should be some ultimate, causally effective reality behind the appearance, since the existence of the external object independent of our cogition is admitted, and that appearance is all that we could have as our
knowledge,28 are views that we can find presented and employed in Dharmakīrti29 but delicately neutralized and even objected to in Dignāga (we will elaborate on this matter and show how the neutralization can be performed in Chapter 4 with NMukh and PS(V)1). It is really such a shame that the delicacy was discontinued in the history of Buddhism, and the consequences were the deepening chasms among the Buddhist schools in the later development (Madhyamaka and Yogācāra) and, as the
28 Hayes (1945:97): “The key philosophical issue that is supposed to separate the Sautrāntika and the Yogācāra perspectives is that of whether or not the objects of experience actually exist independently of our awarenessof them.” Shastri (1964: 41): “According to this theory, external objects are not apprehended directly and
immediately, but through the cognitions of these objects. The objects transfer their forms to their cognitions, and the cognitions, having thus acquired the forms of the external objects, become their representatives. We have thus a representative perceptions of objects, and not a direct one. Hence the theory is called representationism. External objects, not being perceived directly, are only inferred from their cognitions to which they impart their forms.
Orthodox Indian writers, in their compendia of philosophical systems, have ascribed this theory to the Buddhist Sautrāntika school.”
29 Also see Hattori (1968: 90).
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
result of the failure to unite the schools within Buddhism, the discontinuation of Buddhism in the intellectual society in India in the forth-coming centuries.
Given Dignāga's status as the founder of the systematized Buddhist logic and epistemology that aim to construct a platform of internal and external scholastic communications, and as a philosopher who never proclaimed his own scholastic affiliation in his texts and was so dedicated to the unification of Buddhism to counter the revival of Brahmanism, the examinations of Dignāga's external academic
relations would naturally wield great significance and is probably also a necessary and reasonable approach to better understand Dignāga's own philosophy and his contributions to the entire tradition. However, we have to draw attention to a few deficiencies in Singh's suggestion so that we may better appreciate his attempt.
First, texts are determined, but philosophy is always dynamic and developing, especially when it comes to an issue that is questioning and confronting the
boundaries of our knowledge, when the determination of that which itself cannot be determined has to be presented in a determined way, and when the ineffable is to be sophisticatedly expressed, philological proofs should not be taken as the proofs for examining philosophy unless accompanied with very deliberate, careful and diversely systematical considerations on the core philosophical and systematic issues. In the case of Vasubandhu, this situation is more obvious, because the philosophical system was dramatically developed through time, whether or not there existed more than one Vasubandhu. Second, even the final determination of the scholastic lineages, if there can be any, does not imply a determined way for understanding any philosophical system embedded in the lineages; it only provides us with noteworthy suggestions.
Third, we should separate one's own scholastic affiliation from the pragmatic purpose of one's philosophical works; to treat the works of defense and the works of
compromise in the same way is obviously unwise. In the case of Dignāga, it is philosophically interesting to investigate his external relations to his teacher(s) and
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
his philosophical context, because this would provide us with helpful insights to better understand his pragmatic purpose and to know where to draw the bottom line of compromise. Precisely because of the value of determining the bottom line of compromise, we propose a suggestion opposite to Singh's conclusion: to understand Dignāga's works from the position of compromising with the Yogācāra Vasubandhu is a better option than from the position of defending the Sautrāntika Vasubandhu, because this is the highest common that Dignāga's works can target and this can yield the most optimized value of Dignāga's modest neutralization when not mentioning the un-common metaphysical assumptions in his extremely “radical” epistemology.
Suspending the historical question of the plural Vasubandhus, the philosophical grand pre-context of Dignāga is that: outsides, there were external realists adhering to the Nyāya-sūtra, who maintained both the unconditional reality of atoms and the
unconditional reality of cognitive faculties, in the emerging revival of Brahmanism;
insides, there were the Hināyāna Sarvāstivāda, Mahāyāna Madhyamaka, Sautrāntika and Mahāyāna Yogācāra. The disagreements between the Buddhists and
non-Buddhists had begun since the very beginning of Buddhism; non-Buddhists maintain pratītyasamutpāda (all phenomena arise dependently on multiple conditions) while non-Buddhists maintain the absolute cause in the world itself. Next, we have the rivalry between the Hināyāna Sarvāstivāda and Mahāyāna schools; the former
maintains the “ultimate reality” of atoms while the latter rejects it. The portion of the development within the school of existence from the Hināyāna Sarvāstivāda via Sautrāntika to Mahāyāna Yogācāra is very tricky, because, although each of the later schools developed and significantly revised the tenets of the former (from
Sarvāstivāda's admitting the ultimate reality of atoms via Sautrāntika's insistence that appearance is the necessary form of cognition and that atoms cannot be represented in appearance, to Yogācāra's strong rejection of the ultimate reality of atoms), it is easy to confuse the three schools together, in comparison to their major opponent the Madhyamaka, owing to the flavor of “the school of existence” that they seem to share
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
in common and also owing to epistemology as the common method they all adopt.
Thus, the facts have becomes so garbled that the Mahāyāna Madhyamaka and
Mahāyāna Yogācāra share a greater philosophical resemblance than that between the Hināyāna Sarvāstivāda and Mahāyāna Yogācāra – the rejection of the unconditional
Mahāyāna Yogācāra share a greater philosophical resemblance than that between the Hināyāna Sarvāstivāda and Mahāyāna Yogācāra – the rejection of the unconditional