1. Introduction
1.3 Backgrounds of the Problems
1.3.2 Kant: Critical Epistemology and Transcendental Idealism
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
the mind in Samādhi is not obtained from different conditions”50; “(One has to) use the Buddha's teachings as mirrors to reflect one's own mind and use one's own mind as the light of wisdom to illuminate the hidden meanings between lines of the
sutras,”51 etc. Thus, it was a huge surprise to us that none of the contemporary scholars of Buddhist epistemology has taught us that Buddhist epistemology should be free of any ontological presumption – the mind itself or the consciousness itself to be out of the question of ontology – and that we should be cautious about the
ontological assumptions when understanding such Buddhist epistemic terms. To the best of my knowledge and from my survey, no modern scholar has suggested reading Buddhist epistemology as a “radical one,” not even the Dignāga scholars.
Hence, it seems eminently reasonable to reconstruct an alternative understanding of Dignāga's epistemology as a case of critical epistemology.
1.3.2 Kant: Critical Epistemology and Transcendental Idealism
Kant's Transcendental Idealism was first officially introduced and elaborated in KrV and Prolegomena. Kant concluded that the idealists from the pre-Socratic Eleatic School to Bishop Berkeley were all in agreement with the one single formula that “all cognition through the senses and experience is nothing but sheer illusion, and only in the ideas of the pure understanding and reason is there truth”52 (Prolegomena: 374).
In face of the two trends of idealism before him, the “dogmatic/visionary idealism” of Berkeley and the “sceptic/problematic idealism” of René Descartes, Kant proposed a
“third trend” in the development of modern idealism in the history of the western philosophy. Kant labelled both of the aforementioned trends in idealism as
50 “三昧證知心，非從異緣得” in Da Pi Lu Zhe Na Cheng Fuo San Mei Jing, 大毘盧遮那佛三昧經.
51 “以聖教為明鏡，照見自心。以自心為智燈，照經幽旨” in Yuan jue jing da shu shi yi chao, 圓覺經大疏釋義
鈔 by Zong-mi (宗密).
52 Ellington's translation.
‧
“empirical idealism” (meaning essentially that the sensational experience is not ultimately true at all), and named his own proposal in KrV “transcendental idealism”
with a proviso that it was so termed only in contrast to the “empirical idealism” but with no intention, as he exclaimed in the second edition of KrV (especially
“Refutation,” KrV: B274-9) and in the Prolegomena, to take any charge of the idealistic accounts. Nonetheless, it was telling that he assigned to his philosophy the name “idealism.”
The rejected “idealistic accounts,” as voiced in the “concluded formula” above, represent the distrust of the sensational aspect of cognition. Traditionally, idealists trust reason and understanding, and, because of (a) the causal dependence of the operative senses and (b) the uncertainty (indeterminate nature) of our sensation, believe that our sensational cognition is only illusory and ultimately not true.
Berkeley was particularly concerned with the former (a) and Descartes with the latter (b). Acting on his concern, Berkeley concluded that space and the things in space that have to arise from our sensational cognition are never to be ascertained not to be the mere “imaginary entities,”53 while the condition that makes space and the things in space possible is but a non-entity.54 On the other hand, Descartes chose to hold on to the “indubitable” inner experience (thinking) with some deliberate but problematic assumptions on the correspondent existence of the outer existence (inferred thing in itself). On the contrary, by casting doubts on such metaphysical assumptions as the untested ontological status of space in itself and the cognized objects in themselves, Kant believed in “fact,”55 the given knowledge, and that fact counts sensation, as well as understanding and reason, as a necessary component. Further, he believed that
53 Einbildungen, KrV: B 274.
54 Unding, KrV: B 274.
55 Tatsache, KrV: A 84 / B 116. “Die Rechtslehrer, wenn sie von Befugnissen und Anmaßungen reden, unterscheiden in einem Rechtshandel die Frage über das, was Rechtens ist (quid juris), von der, die die Tatsache anghet (quid facti), und in dem sie von beiden Beweis fordern, so nennen sie den ersten, der die Befugnis, oder auch den Rechtsanspruch dartun soll, die Deduktion.” Kant used “factum” and “Tatsache” to mean fact. However, while
“Tatsache” referred to the “something done” in KrV, “factum” was used in KpV in analogy to “Tatsache” to refer to the truth about reason: “fact of reason” – “Das vorher genannte Faktum ist unleugbar” (KpV: 5:32). However, we can observe in KU (§91) that the term “Tatsache” included both the narrower theoretical sense of “something done”
in KrV and the practical sense in KpV.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
each one alone without the coordination with the others must but collapse and cannot yield any empirical reality. We can see that this attitude also prevailed among the Nāgārjuna and Buddhist epistemologists. Nāgārjuna, in MMK and
Vigrahavyāvartanī, always upheld the inconsistence between the necessary,
exhaustive logical consequences of a thesis and the fact to argue for the falsehood of the thesis56 without needing any metaphysical reference or any reference to the
ontological status of thing in itself. Dignāga, on the other hand, insisted that we have to establish two and only two pramāṇa-s, a sensational one and a conceptual one, while the sensational pramāṇa enjoys a privileged epistemic priority with regard to the validity of cognition. Besides, Dignāga established two and only two pramāṇa-s because we factually always have two and only two aspects in the cognized object, the particular aspect and the universal aspect.57 The sensational pramāna as a valid source of knowledge demonstrated Dignāga's epistemic optimism about sensation, while the rationale for the two and only two pramāṇa-s in fact showcased his trust in fact, as we have seen in Kant and Nāgārjuna.
Kant's epistemic optimism, especially with regard to sensibility, can be clearly
discerned from “Transcendental Aesthetic” in KrV. Basically, Kant argued there that in order to meet the exhaustion of space and time as valid sensational aspect of
cognition in all possible experiences, the transcendent reality of space and time (viz., a concept of the reality that space and time might be posited to have when they alone were detached from any sensational objects) must be rejected and we have to accept their transcendental ideality, meaning that we have to take the sensational aspect of the cognition of space and time to be mere forms of representation (i.e., to be at the level equal to the status of understanding)58. Remarkably, we can find Nāgārjuna and
56 “However the fact is not like this (na asti, 是事則不然).”
57 PS(V) 1.2.
58 Kant writes “Unsere Behauptungen lehren demnach empirische Realität der Zeit, d.i. objektive Gültigkeit in Ansehung aller Gegenstände, die jemals unsern Sinnen gegeben werden mögen” (KrV: A35/B52). This passage may be understood that Kant maintains the “empirical reality of time (empirische Realität der Zeit)” itself, as Kemp Smith understood (1929: 78): “What we are maintaining is, therefore, the empirical reality of time, that is, its objective validity in respect of all objects which allow of ever being given to our senses”. Yet, the original reads differently: “Unsere Behauptungen” as aforementioned thus leads us to admit “empirische Realität der Zeit” which
‧
Dignāga both maintain similar positions on the aesthetic condition of space and especially time, namely, the anti-realistic position. In MMK 19, Nāgārjuna concludes that time, just like external things (things in the form of space), does not exist
ultimately and has to be restricted to phenomenon (Cf. Siderits and Katsura, 2013:
207-211). Dignāga in his Traikālyaparīkṣā refuted the ultimate reality of the time difference upon the Mahāyāna Buddhist code that “all existences are empty and with no self-nature” (Nasu, 2015: 13; Cf. Hattori, 1961).
Here, we may run into a paradox. Transcendental idealism calls for the
uncognizability of the thing in itself and places cognition (and thus experience) in a position of mere representation (the result of cognition). However, owing to the concern about the uncaused result (Kant himself also had to admit that sensibility must be affected in some ways by objects),59 people naturally understand the transcendental ideality of space and time by adding a postulate of external yet uncognizable causal relation between the thing in itself and the sensibility and a postulate of the uncognizable thing in itself. According to Allison (1983/2003: 450), from Kant's contemporaries, including Pistorius, Eberhard, Jacobi, Maimon, and Aenesidemus-Schulze60, to Kant's most influential introductory figure into the
English academic circle Strawson (1966) and his followers Prichard, Guyer, Langton, etc., this interpretation (with the two additional postulates) has remained the standard mainstream “treatment” of such a position. However, his treatment unavoidably invited a paradox. All that we can know is but representation, while we still have to
however does not refer to the empirical reality of time as by itself alone but to “objektive Gültigkeit in Ansehung aller Gegenstände, die jemals unsern Sinnen gegeben werden mögen.” Cambridge edition (Guyer and Wood, 1998:
181) is closer to the original: “Our assertions accordingly teach the empirical reality of time, i.e., objective validity in regard to all objects that may ever be given to our senses.” This understanding is telling that time itself is regarded to be empirically real because it is a real valid condition to all possible objects of the senses, not that time itself is an empirically real existence. The empirical reality of (space and) time itself in no relation to any objects of cognition actually automatically implies its (their) transcendental reality.
59 “In what way and through whatever means a cognition may relate to objects, that through which it relates immediately to them, and at which all thought as a means is directed as an end, is intuition. This, however, is possible only if it affects in a certain way. The capacity (receptivity) to acquire representations through the way in which we are affected by objects is called sensibility” (KrV: A 19 / B 33).
60 Kant's own reply in P to the “Garve-Feder Review (Göttingen Review)” in 1782 (P: 4:372-80). Vaihinger's Commentar zu Kants Kritik der reinen Vernuft, vol. 2, (1881-92: 494 – 50) is a great resource for accessing the debates in the 1780s.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
assume a causal explanation which itself contains entities and relations which can never be representational. As the result of the treatment (in order to avoid the embarrassing scenario of space, time and the thing in itself with their non-realistic character implied in the idea of the ideality of space and time), the idea of the ideality of space and time in Kant has been abandoned (Strawson and Prichard) or placed in isolated suspension (Guyer and Langton), while his epistemology is still employed to reject skepticism. This means that, putting aside the issue of space and time being transcendentally ideal and embracing the two postulates (of the thing in itself and the causal relation between it and sensibility), the postulated causal relation between our sensitivity and the thing in itself has successfully distinguished the theory itself from the other two trends of idealism by answering to the issue of unrooted and uncertain sensation and preserving the trust in sensibility.
However, such a “treatment” is definitely different from Kant's own position, and people who learned Kant through this series of treatment should bear in mind that it disagrees with Kant in the kernel. In his defense, Allison (1983/2003) plausibly drew our attention to the appropriate target of defense: the transcendental ideality of space and time. Additionally, he plausibly indicated that this series of treatment on Kant's transcendental idealism were the results of these interpreters reading Kant's
epistemology “ontologically.” The meaning of the “ontological reading” that Allison tried to convey can be further discerned from his responses in the second edition of his Defense to two influential thinkers who did Kant justice but nonetheless disagreed with Kant's idealism as well. In response to Van Cleve (1992)'s phenomenalist
criticism, which understood Kant's transcendental distinction between the concept of representation and the concept of the thing in itself to be a distinction between the human world in the view of humans and the world in itself in the view of God's eyes, Allison proposed that we should just keep the epistemological distinction simply epistemological and not suppose an ontological distinction thereafter. There is no other world implied or suggested naturally in the epistemological distinction.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Ameriks (1992) refused the epistemic understanding of Kant's distinction between the appearance and the thing in itself by arguing that “there is still no reason to think the non-ideal (empirical real entities) has a greater ontological status than the ideal,61” when the empirical reality in Kant itself is the “mere appearance” of the thing in itself while it is supposed to represent what something really is. In response to that,
Allison proposed that we have to pay ample attention to the fact that what Kant was doing is epistemology, not metaphysics. This means that the reading of Ameriks is a reading of metaphysics, and that the ontological status by itself, which has to be beyond our cognitive capacity, is always deliberately but covertly assumed there. To be more specific, in Kant's position, we would not say beyond the appearance that there is some ontological value of the thing in itself, while the comparison between the empirical reality and the transcendental ideality with regard to the “ontological status” is just plain wrong. Moreover, such an assumption of the ontological account of the thing by itself is a character of metaphysics, and this metaphysical ground of transcendental realism is definitely inappropriate for reading Kant. Both responses by Allison can be summarized as follows: Kant's transcendental idealism should not be read from the angle of transcendental realism, which is exactly the position that transcendental idealism rejects. In other words, the “ontological reading” of Kant is a reading that attributes a transcendentally realistic ground to his theory and that
attributes a self-independent ontological reality by itself to space, time, the thing and the cognitive faculties in themselves.
More specifically, the problem is with “causality” and “the affection of the mind.”
One may understand Kant as Jacobi (1968) did, in that the problem with the affection of the mind and the production of representational cognition rests with a systematic dilemma: “without the presupposition [of the thing in itself] I cannot enter the [critical] system, and with that presupposition I cannot remain in it.62” Jacobi believed that there are only two possible candidates for the affecting object to our
61 Ameriks, 1992: 334.
62 Jacobi, 1968: 304. Allison's translation.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
mind: an appearance and the transcendental object. As such, Jacobi reasoned that since an appearance defined by Kant as a mere representation cannot really affect our mind, we hence need to presuppose the existence of the thing in itself. However, the thing in itself, as presupposed to be a true transcendental object, can never be
cognizable. Consequently, the causality, as one of the categories, can never be
effective to that object. This means that even the transcendental object cannot really affect the mind, if we go along with Kant. Thus, both candidates cannot hold in Kant's system. Allison directly refuted Jacobi and wrote: “[f]or affection, as Kant construes it, is clearly an epistemic rather than causal relation” (2004: 64; note 28, p.
460). And by “epistemic relation,” Allison meant that “[the affection relation] holds between a discursive intelligence and … a representation of an object, with this representation being determined in part by the mind's mode or manner of being affected” (ibid.). Allison continued to revert to his main line of defense. If one does not assume any metaphysical, especially transcendental, reality in this system (in this case, the assumption is the empirical affection relation between the thing in itself and the mind), one would not have to call for “a God's-eye account of what it is that really supplies the matter of cognition” (i.e., “some ultimate metaphysical story about affection”, Allison, 2004:73). We can also observe this affection problem with Dignāga, especially with his Ālamb. Here, Dignāga also came to a similar conclusion to Kant's that there is no possibility for the object that is affecting the cognitive subject to be externally real (transcendentally real).
To Allison, the critical work in KrV is to depict the “systematic unity” of cognition (2004: 445; KrV: A 647 / B 675) via following the self-critique of reason. The
“unity” is but the “command” of reason, and hence a projected unity from it, as some neither observed nor inferred unity from empirical data or in the world by itself. In such a scenario, the ideality of space and time is the only option to maintain that unity (for it excludes all possibilities for any transcendental reality), while the concept of the thing in itself is a necessary theoretical presumption in order to respond to and
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
sustain that unity. If one does not insist on reading that concept (the thing in itself) with some assumption of transcendental reality (which goes beyond the scope of reason), that concept simply answers to the command within reason, and the
systematic unity from reason and our understanding about our experience are already confirmed with each other (for the former and the latter derive from the same
determination of our epistemic condition). I think this was the essence of what Allison was trying to argue in the defense, and this implies the inseparability of the ideality of space and time (transcendental idealism) in Kant's theoretical philosophy.
As Allison concluded, “the rejection of transcendental idealism amounts to the rejection of Kant's theoretical philosophy as a whole” (2004: 448), and one cannot accept the latter without also accepting the former. And so far, we have yet to find many sufficiently strong answers to Allison's challenge from the camp of the
mainstream reading of Kant.
In this dissertation, I will attempt to elaborate Allison's systematic understanding by (1) specifying the significance of the causality of freedom (and then the
transcendental logic) in Kant's theoretical philosophy as well as in relation to his moral philosophy upon the ground of his theoretical treatises (this has to do with the argument for freedom in Grundlegung zur Metaphysik der Sitten(G)). Besides, the elaboration will also invites philosophical discussions about (2) the idea of self-awareness in Buddhist epistemology, especially Dignāga's account, together with its systematical significance to the soteriological project in Buddhism – especially its significance to the almost paradoxical relation problem between saṃsāra and nirvāṇa in contrast with the relation between the empirical reality of phenomena and the transcendental ideality of noumena in Kant. More precisely, the self-awareness in Dignāga is contained within the formal conformity between the determination of the object of cognition and the determination of the consciousness of that cognition. With this position, Dignāga rejected the realistic account of the causal relation between the cognized object and the cognitive measures (faculties). I will argue that such a
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
formal, coordination relation supports the strict determinism in the conventional (cognition-dependent) reality and the openness to be free in that strict determinism.
formal, coordination relation supports the strict determinism in the conventional (cognition-dependent) reality and the openness to be free in that strict determinism.