1. Introduction
4.5 The Holistic Argument for Mental Perception from Verse 6 on
4.5.4 Appendix. Explanation for the Twofold Appearance in the Demonstrative Fact of
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
between pramāṇa and prameya is not really causal but only formal and perceptual – owing to this fact, we are only able to perceive, not to infer, and we can express it because the perception itself allows it, as much as we express our direct sensible experience. Next came the conclusion:
PS(V) 1. 10. (Hattori, 1968: 29)
whatever the form in which it [viz., a cognition] appears, that [form] is [recognized as] the object of cognition (prameya). The means of cognition (pramāṇa) and [the cognition which is] its result (phala) are respectively the form of subject [in the cognition] and the cognition cognizing itself. Therefore, these three [factors of cognition] are not separate from one another.
4.5.4 Appendix. Explanation for the Twofold Appearance in the Demonstrative Fact of “Recollection”
In order to back up the twofold appearance theory in the self-cognition as the demonstrative argument for mental perception, Dignāga utilized another fact, viz., recollection, to demonstrate the necessity of the true nature that is easily
“overlooked” by the common people. In Verse 11ab, he wrote, “that cognition has two forms is [known] from the difference between the cognition of the object and the cognition of that cognition,” which are actually the same single cognition. Aside from the former characteristic, the latter is also required as a necessary fold of the same appearance in the sense that it is demanded perceptually (in perception alone) that the cognition (a) has to possess the form of the object (the cognition of that object must be cognized) and (b) itself must appear as the cognition-holding subject of the cognition itself, viz., that object must be known by itself (oneself).
“Otherwise,” as posited in the vṛtti, “if the cognition of the object had only the form of the object, or if it had only the form of itself,” i.e., discursively in real separation,
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
“then the cognition of cognition would be indistinguishable from the cognition of the object,” i.e., the real connexion of the cognition and its object would lose its
perceptual basis and would fall into the predicament, as the proceduralists do, whereby all cognitions about “the external” are but inferences and should not be recognized by us as real “empirical reality.” This problem will rear its ugly head when we put our finger on the issue of deciding whether Dignāga's logic is inductive, deductive or neither by nature.10 Since it is factual that in perception alone, we are allowed to “distinguish” the cognition of the object and the cognition of the cognition itself in one single cognition reflexively, as a matter of fact, we are having
recollections that must demand such a twofold characteristic in the appearance, and we have to admit that in the holistic perception alone, both (a) and (b) above are true. It follows that both the twofold appearance theory and the assertion that the three factors of cognition are related perceptually are true as well, and therefore mental perception is established. Many scholars have conducted in-depth studies on Dignāga's theory of self-awareness with the case of recollection, cf. Kellner (2010).
However, the emphasis we focus on here is that the fact of recollection is the
evidence for the necessary underlying fact of the cognition, the self-cognizing nature, and not on its demonstration of the probably problematic and complicated real causal structures among the different stages of cognition. From this point of view, the argumentative line does not get caught up in a vicious circle, as was the case in the paradigm established by Dharmakīrti.
In the vṛtti of Verse 11b, Dignāga said, in recollection, that “because that [object of the preceding cognition does not exist when the succeeding cognition arises and]
could not be the object of the latter. Hence it is proved that cognition has two forms.”
In the commentary tradition after Dharmakīrti, the argument here is understood, briefly speaking, as follows. There are two distinct folds of the appearance in the
10 Both the inductive and deductive logic assume strong ontology, either positive or negative assumption. Since Mahāyāna Buddhism hesitates to embrace both, it is not surprising that both candidates are not reasonably suited to its logic.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
causal relation (the real separation), viz., the appearance of the object known and the appearance of the subject knowing the object (having the form of the object); the causal relation in between ascertains the conformity of the two forms (the latter's having the form of the former cognition because of the causal chains). Thus, in the case of recollection, if there were only one fold of the appearance, either there would be only the cognition of the object alone or the cognition of itself alone, or
recollection would be impossible. This is because in the case of recollection whereby the cognized object no longer exists, neither of the options would result in the
cognition of recollection as such. On the contrary, only when there are two folds, namely that the succeeding cognition takes the preceding cognition (which takes the form of the object when it exists) as its object and the conformity between the two cognizing activities is ascertained by real causal relations, is recollection then
possible (see Dharmakīrti in PV III: 379-380; Kumārila in Śūnyavāda: 112cd-114ab;
Hattori 1968: 108-110; cf. Kellner 2010). However, this account in general is not only absurd from our experience but also in stark disagreement with the theoretical context of Dignāga's philosophy.
The real causal warrant among cognitive activities in this account implies that the succession of the cognition takes place in time and thus in causal reality. This means that when we recollect, we do cognize a previous cognition as much as we cognize a present object in certain real causal chains, and this is not really what happens when we recollect. It is absurd and runs counter to our experience to say that when we recollect, we cognize some previous cognition and its object; we simply remember that cognition. In recollection, we do no cognize the form of the previously cognized object. Besides, our memory may serve us incorrectly, and there is no such causal warrants in our mind alone (where the recollection is supposed to take place
according to the theory) to help us check the veracity of our memory unless we resort to the present world. Recollection is not re-cognition. Besides, this real separation not only ushers in the problem of the infinite regress of additional cognitions but was
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
also openly refuted by Dignāga himself; furthermore, this real separation also violates Dignāga's definition of pratyakṣa, viz., the aspect of the cognition that is free from conceptual constructions/differentiation. It would be prudent not to ignore the not so trivial implication in Dignāga's twofold appearance theory whereby, since that theory concerns perception alone, it does not imply any differentiation as two real folds but rather a feature of a perceptual fact which allows for and can be understood in a structure of two related forms/characteristics in thought alone, as much as the analysis of a dharmin's holding dharma-s does not imply that they are two ontologically distinct entities.
Since it is a fact that we have recollection and the recollection must come from experience, viz., from real previous cognitions, the twofold perceptual characteristics in perception alone must be admitted. Therefore, the overlooked characteristic of the cognition, viz., the self-cognition, is revealed. Thus wrote Dignāga (Hattori, 1968:
30):
Self-cognition is also [thus established]. Why?
k. 11d. because it [viz., recollection] is never of that which has not been [previously] experienced.
Furthermore, Dignāga, in the vṛtti of Verse 12b openly refuted the proposition to assume another cognition to cognize a given cognition, because “the latter
recollection of this separate cognition does not arises unless it is experienced”; it is not true because we do not have such an experience of the additional (self-)cognition (and our experience only allows us to think of itself with “twofold” and no more). As he continued, “so if it should be that this [separate] cognition is experienced by the third cognition [so that it may be recollected], there would be an infinite regression,”
a point that is also contrary to our experience.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Furthermore, the proposition to assume a third cognition to take the previous
cognition in recollection is also problematic in the sense that, since the recollection is regarded as internal cognitive activities via real causal chains in the mind alone, “in such a case,” as Dignāga wrote in Verse 12cd, “there could be no motion [of
cognition] from one object to another. But actually such [a movement of cognition]
is accepted.” Real causality is not arbitrary, and its alternation is not submitted to our will or to whether we experience it or not. If we admit the real, mental casual
account in recollection, the complete causal network will not only render our
cognition of the world comprehensive and complete (both with regard to the admitted external spatial manifold reality and the internal temporal/historical reality) but also halt the “motion of cognition” from one instance to another, for such
comprehensiveness and completeness is contrary to alternation. On the other side of the problem of the infinite demand (regression) of the additional cognition is exactly the problem whereby the cognition as such makes no sense since such a motionless comprehensive state is contrary to our real experience. Only the sharp distinction between the manifold perception and the limited differentiation in the conceptual constructions assures the discursive nature of the cognition; the causal perceptual process of differentiation in our mind (in manas; in epistemology) not only allows in the end no perception at all by ruling out the possibility of the discursiveness of the cognition but also destroys the theory of perception itself, and thus epistemology itself. On the contrary, Dharmakīrti's understanding of this problem was absurd in the context of Dignāga, too. He understood “no motion” in such a way that since
“inasmuch as we admit the movement of the cognition from one object to another, the series of cognitions must be broken at certain point. If this is the case, the last in the series of cognitions is not cognized by another cognition” (Hattori 1968: 113). It is obvious that such an understanding was based on his proceduralist causal account of the self-cognition, and he believed that once we stop recollecting, a motionless stage shall ensue. Hattori (ibid.) clearly pointed out the untenability of this account, viz.,
“since all cognitions are of the same nature, the last one must also be cognized, so
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
that it may be recollected.” Furthermore, it is also absurd that inasmuch as admitting the additional cognition in recollection, as Dignāga indicated, there must follow the problem of the infinite regression of additional cognitions. Why then, on the same basis that there will be a stop point of the regression as Dharmakīrti maintained, especially as he himself did not openly advocate, and as the real externalists such as Kumārila maintained, is the series of cognition finite (Hattori, ibid.)? In contrast, our holistic understanding of the motionlessness in this passage is more tenable. That is, Dignāga did mean to refute the causal account of mental objects – at the very least, the two sharply distinct kinds of causal relations do not causally connect with each other – and felt that this confused account would invite the problem of the infinite regression and the problem of motionlessness in the cognition (the destruction of the discursive nature of the cognition); in this sense, the theory of the sharp distinction between perception and inference, especially the criterion that perception by its nature is free from conceptual construction, is a sound theory that is tenable to our experience, when we interpret the theory in a holistic manner.
Since the reflexive nature, viz., the self-evidence of the cognition, is established in the demonstration of the twofold characteristics of perception alone, mental perception, be it the awareness of the form of the object, the awareness of feelings or the
awareness of conceptual constructions, is established; “in this way,” Dignāga concluded this chapter, “it is established that perception is free from conceptual construction.”
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
4.6 Appendix. Jinendrabuddhi – The Difficulty in the Sharp Distinction between