Is Mental Perception Necessary in a Pramāṇa Theory?

8/3/2019 Is Mental Perception Necessary in a Prama Theory?

1/21

Is Mental Perception Necessary in a Pram a Theory? A Re-reading of Digngas

Theory of Svasa vedana and Mnasa-pratyak a in his Pram asamuccaya

Ching Keng, Aug. 26, 2004

Introduction

In his book on self-cognition (svasavedana), Paul Williams suggests that in the history of

Indo-Tibetan Buddhism, there existed two different types of self-cognition. The first is patterned

on a cognition of others and can be put as: self-cognition means one cognition is cognized by

another cognition. The second is not patterned on a cognition of others and can be regarded as

resorting to the self-luminosity or self-reflexivity of cognition. It means that whenever one

cognizes, one is at the same time conscious or aware of her cognizing.1

In the following, I will designate these two types as model (1) and model (2) of

self-cognition respectively. Namely,

Model (1): self-cognition means that one cognition is cognized by another cognition or x is

cognized by y. Whether this second one is simultaneous with or succeeding the first one, whether

this second one may be regarded as separate from the first one or not, is another issue.2

Model (2): self-cognition means that a cognition is always cognizant of its cognition while it

cognizes. Self-cognition refers to cognitions self-reflexivity alone, never to its relation to another

cognition.

In the following, I will base my discussions on the distinction between these two models of

self-cognition. I will first reexamine Digngas view of perception (pratyaka) and argue that

1 Cf. Williams, chapter one. Rueggs contrast between other-illumination vs.self-illumination may also be inspiring

to characterize the distinction between these two models. See note 194 of Rueggs English translation of the first

chapter of thePrasannapad in Ruegg, p. 109.

2 Cf. Williams, chapter one, note 9 and my following discussion in section four.

1


2/21


3/21

Now the question is: how could such self-cognition be possible? A possible clue may be

drawn from the above quote. Note that When a cognition possessing [the form of] an object

(saviaya jnam) is itself the object to be cognized (Italic by this author) is described by

Dignga as in accordance with the nature of the self-cognition. With this clue, I would argue

that self-cognition is understood by Dignaga through the above-mentioned model (1), i.e.,

cognition cognized by another cognition.

If we adopt Hattoris explanatory symbols3, then the above process of self-cognition can be

formulated as follows:

At moment t1

, we have cognition C1

=(S1

O1

), then at moment t2

, we have cognition C2

=

(S2O2). Since C2 cognizes, or, takes C1 for its object, we have O2=(S1O1). Therefore, the

self-cognition C2= (S2(S1O1)). At the moment of t2, the cognition also cognizes O1 as

something either desirable or undesirable.

That Dignga understands self-cognition by means of the model of a cognition cognizing a

cognition can be further corroborated by looking at his argument for the existence of twofold

appearance in a cognition. As Dignga says:

The cognition which cognizes the object, a thing of color, etc., has [a twofold appearance,namely,] the appearance of the object and the appearance of itself [as subject.] But the

cognition which cognizes this cognition of the object has [on the one hand] the appearance

of that cognition which is in conformity with the object and [on the other hand] theappearance of itself. Otherwise, if the cognition of the object had only the form of the object,

or if it had only the form of itself, then the cognition of cognition would be indistinguishable

from the cognition of the object. (PSV Ha, Hattoris translation, p. 29-30; Italics by thisauthor)

According to Hattoris reconstruction, this argument means:

3 Cf. Hattoris notes 1.70 and 1.71. (Hattori pp. 108-109)

3


4/21

Now in case the cognition had only arthkra (=bhsa), then C1=O1, and C2=O2. SinceC2 takes C1 for its object, O2=O1. Therefore, C2=C1. Thus, viaya-jna-jna would be

indistinguishable from viaya-jna. If, on the other hand, the cognition had only svkra,

then C1=S1, and C2=S2. However, since the cognition which does not possess the form of

an object within itself remains the same at all times, S2=S1. Therefore, C2=C1. (Hattoris

note 1.70, p. 108)

If Hattoris reconstruction of the argument is correct, then the only way that renders this

argument intelligible is to read the cognition of cognition as resorting to model (1) of

self-cognition. The above argument denies that a cognition could be formulated as C 1=S1 or

C1=O1. The implication is that cognition and self-cognition should be formulated as C1=(S1O1)

and C2= (S2(S1O1)) respectively. In other words, Dignga is clearly adopting model (1) for

construing self-cognition. If he did not, then the above argument would just be unintelligible.4

However, if the above reading of Digngas idea of self-cognition is correct, then we would

have to face the following difficulties:

First of all, in the same chapter of the PS, Dignga also obviously opposes the idea that a

cognition is cognized by means of a separate cognition.5

As he says:

Hc. 2. Some may hold that cognition also, like a thing of color, etc., is cognized by means of

a separate cognition. This is not true because:

K. 12a-b1. if a cognition were cognized by a separate cognition, there would be an infiniteregression.

Now if Dignga really holds the above model of self-cognition, then it would be very

difficult to argue how C1 and C2 as formulated by Hattori could not have been separate

4 Another passage that would suggest the resorting to model (1) for self-cognition in the PS is the following:

Or [it can be maintained that] the self-cognition or the cognition cognizing itself(svasavitti) is here the result

[of the act of cognizing]. (PS k. 9a; Hattoris translation, p. 28; Italics by this author)

5 This is the view hold by the Naiyyikas. See Hattori, note 1.76, p. 111.

4


5/21

cognitions. In other words, it would be difficult to argue that C2 or self-cognition at t2 could be

somehow identified with C1 or cognition at t1.

Furthermore, we would also encounter the difficulty of what I will call the reproduction of

a previous cognition. To elaborate, lets go back to another argument Dignga provides here:

Further, [if the cognition had only one form, either that of the object or of itself,] then theobject which was cognized by a preceding cognition could not appear in a succeeding

cognition. Why? 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 thatcognition has two forms. (PSV. Hb. Hattoris translation, p. 30)

Again, adopting Hattoris formula, the argument goes like this:

At moment t1, we have cognition C1=(S1O1), then at moment t2, we have cognition C2=

(S2O2). Since C2 cognizes, or, takes C1 for its object, we have O2=(S1O1). Therefore, C2=

(S2(S1O1)). This, according to Hattoris interpretation of Dignga, is the only possible way to

account for how an object O1 in the previous moment t1 could ever appear in a subsequent

cognition C2 at t2.6

But the problem here is: Hattori forgot to explain how cognition C1 couldsubsistat t2 in the

first place in order to be cognized by C2 at t2. Given the Buddhist view of universal flux or

momentariness (anityat), it seems far from self-evident that C1 could ever exist at t2. In short, it

seems to me that, in order for the model (1) for self-cognition to be possible, it is necessary either

to argue for the possibility of reproduction of the previous cognition or for the possibility of

the simultaneity of the two cognitions. (I will come back to this difficulty later in the last

section on mnsa-pratyaka)

To conclude this section, I have tried to pinpoint the possible difficulties in understanding

6 Cf. Hattori note 1.71, p. 109.

5


6/21

Digngas text. It is my belief that people after Dignga such as Dharmapla, Dharmakrti, and

ntarakita, all tried to fix these problems implicit in Dignga. Or, to put it another way, only

when we claim that there exists such inner tension in Dignga, would later re-formulations of

self-cognition become intelligible.

Section Two: The Suggestions by Dharmapla

For most if not all scholars of East Asian Buddhism, Dharmaplas theory of fourfold

appearance has been deemed strange if not unintelligible. In this section, I would try to argue

that his theory of fourfold appearance would become most intelligible in light of the difficulties

in Digngas PS.

In his commentary to the Cheng Weishi Lun (*Vijaptimtrat-siddhi), Kuiji pointed out that

up till Dharmapla, there had been four different theories of cognition7, namely, Sthiramati, who

held that cognition consists of one appearance (bhga orbhsa) alone; Nanda, who argued for

twofold appearance; Dignga, who argued for threefold appearance; and Dharmapla, who

argued for fourfold appearance.8 It is noteworthy here that, according to what is reported by the

Cheng Weishi Lun, what Dignga really held was not a theory of twofold appearance in cognition,

7 It seems not perfectly clear to me whether the following four theories are of cognition or of consciousness. I am

also not sure about whether cognition can be identified with consciousness in this context. According to the Cheng

Weishi Lun, what is said to be consisting of fourfold appearance is consciousness rather than cognition.

Here I am following Hattoris translation for the moment and ascribe those appearances to cognition rather than

consciousness. But note also that in note 1.68, Hattori cites from Prajnakaraguptas Pramavrttikabhya or

Vrttiklakra: katha punar jyate [text: jyate] dvi-rpavijnam iti. (emphasis by this author). Cf. Hattori

note 1.68, p. 108.

It seems clear to me that at least Dignga in his PS directly talks about the twofold appearances of cognition ().

8 in Chapter Three of the Commentary to the ChengWeishi Lun (T43, no. 1830, p. 320, c20-22).

6


7/21

as would be inferred from the text of PS. Instead, what Dignga held was a theory of threefold

appearance, namely, in every cognition by a consciousness, there are three appearances, viz. the

appearance of object (viaybhsa), that of subject (svbhsa), and that of self-cognition

(svasavedana).9

As stated in the previous section, one of the difficulties left for Dignga in the PS is about

how the self-cognitions cognition of the previous cognition would not fall into the Naiyyikan

model of a cognition cognized by a separate cognition. Now we can tell that one merit of such

a theory of threefold appearance by Dignga is that it may avoid the problem of falling into the

Naiyyikan model of separate cognitions by arguing that all these three appearances are

simultaneous and pertain to one and the same cognition.10

In this way, Digngas theory of

self-cognition can be successfully differentiated from the Naiyyikan model in that the latter

adopted a model of separate cognitions in consecutive moments.

On the other hand, in order to argue for the theory of fourfold appearance, the Cheng Weishi

Lun also offers the following criticism of Digngas theory:

Moreover, if we distinguish the citta and caitta in more detail, there should exist four

9 Due to the limitation of scope of this paper, it is neither possible to trace the origin of such theories back to earlier

Yogcra theory, nor to relate the theories of the appearance of cognition to later debates between the so-called

Skravdins and the so-called Nirkravdins. It suffices here to focus on the possible development between

Dignga and Dharmapla.

10 I am aware of the conflicts between the simultaneity among these three appearances and Hattoris diachronic

analysis employed in the above. I am not able to solve these conflicts in this paper. For the moment, I can only argue

that Hattoris diachronic analysis can still be useful to the extent that at least the cognition of object C1 is logically

preceding the self-cognition. The reason for this is that Dignga implies at least the logical precedence of cognition

over self-cognition by saying: When a cognition possessing [the form of] an object (saviaya jnam) is itself the

object to be cognized, then, in accordance with the nature of the self-cognition, one conceives that [secondary] object

(artha) as something either desirable or undesirable. (PSV G. Hattoris translation, p. 29)

7


8/21

appearances. Three of them are as above, and the fourth is included as the cognition of

self-cognition (svasavitti-savitti-bhga or savitti-svasavitti). If this [the fourth] did

not exist, what would cognize the third [appearance]? Since all these [four] are appearancesof mind, all of them should be cognized [with no exception]. [On the other hand, if the

cognition of self-cognition did not exist,] there should be no result (phala) of the

self-cognition. But there must be results for each and every prama. One should not takethe appearance of subject as the result of the third appearance since sometimes the

appearance of subject is not subsumed under pramas. Hence, the appearance of subjectcould not cognize the third because whatever cognizes the third must be a direct perception

(pratyaka). (T31, no. 1585, p. 10, b17-22)11

According to this passage, the main issue the Cheng Weishi Lun takes with Dignga is the

question about what would further cognize the self-cognition. The criticism can be reformulated

as follows: if the appearances of object and of subject need to be cognized by the third

self-cognition in order for the cognition to become cognizant of them (the first two appearances),

then it seems self-contradictory why Dignga thinks that the third needs not be cognized by a

fourth appearance in order for the cognition to be cognizant of the third. It is find that Dign ga

thinks that, in order for the cognition to be aware of the first two appearances, self-cognition as

the third appearance is necessary. But Dignga ignores that the self-cognition itself needs be

cognized as well!

As for the second part of the criticism, Dharmapla opposes the possibility that the second

appearance could play the role of cognizing the self-cognition and hence render the cognition

cognizant of the latter. Based on his critiques of Dignga, Dharmapla suggests his theory of

fourfold appearance.

One obvious problem in Dharmaplas theory is this: how could Dharmaplas theory not

11Cheng Weishi Lun: T31, no. 1585, p. 10, b17-22

8


9/21

fall into infinite regression, as the third needs the fourth, and the fourth needs the fifth, ad

infinitum? Dharmapla was fully aware of this difficulty and did find a way out of this impasse

by giving the following explanation:

[Among these four appearances:]the second cognizes the first alone, either through a

prama or a non-prama, either through a pratyaka or an anumna. The third can cognizeboth the second and the fourth. The cognition of self-cognition cognizes the third alone. It(the fourth) does not cognize the second, since such is useless. Therefore, both the third and

the fourth are subsumed under pratyaka. (T31, no. 1585, p. 10, b23-26) 12

Whether Dharmaplas critiques of Dignga and his own solution are convincing or not is

not my main concern here. The more important point here is that, according to Dharmaplas

further development of Digngas theory of threefold appearance, it seems clear that

self-cognition was understood by Dharmapla by resorting to model (1), with the four

appearances pertaining to one and the same cognition. As will be shown in the following section,

such a model about self-cognition was later superceded by a completely different model of

self-luminosity.

Section Three: The Suggestion by ntarak ita (and Dharmakrti?)

That self-cognition was later understood not through model (1) but through model (2), i.e.,

the model of self-luminosity of cognition, is evident in the Tarkabh, as it is said:

As a lamp illuminates itself, so is also knowledge considered to know itself, since it is, quite

differently from insentient things (jaapadrtha), produced by its own cause with the nature

of self-luminosity. (Kajiyama, p. 48)

12Cheng Weishi Lun: [6](T31, no. 1585, p. 10, b23-26)

9


10/21

This idea of self-luminosity, according to the Tarkabh, comes from ntarakitas

Tattvasagraha, as the latter is quoted in the former as follows:

Knowledge is by nature opposed to insentient matter; this immateriality is nothing but the

self-consciousness of knowledge. (Kajiyama, p. 48; quoted from Tattvasagraha verse

1999)

The self-consciousness of knowledge is not to be analyzed into action and its agent, since

knowledge, being a single unity without compartments, cannot be divided into the three

parts [viz. the knower, the known and the knowing.] (Kajiyama, p. 48; quoted from

Tattvasagraha verse 2000. Cf. also Masamichi Ichigos translation in Ichigo, p. 176)

According to Masamichi Ichigo, one of the natures of self-cognition for ntarakita is

illumination (praka). It is considered to be non-dual because it lacks the dichotomy of the

grasping and the grasped. (Ichigo, p. 173) Ichigo also draws our attention to the following two

passages from the Tattvasagraha:

(P) Therefore, that which is the subject of the dispute (svasavedan) is considered to benon-dual, since it is devoid of object and subject (vedyakarttvaviyogt),

(H) Because it is of the nature of knowledge (vijnatva),

(D) Like a reflection. (Tattvasagraha verse 2078, Ichigo, p. 173)

TS 2081 The nature of knowledge is its capacity to illuminate, so that it does not have aplace in the grasped (object). Since (knowledge endowed) with no image, etc., is not

reasonable, the pervasion (vypti) (of knowledge by non-duality) is established.

(Tattvasagraha, Ichigo, p. 173)

Based on these passages, Ichigo thus concludes, the process of knowledge cognizing itself

can never be understood by means of the relation between either the agent and its action or the

knower and the known. The connection cannot be established between two things that are

different in quality, despite their apparent relationship. (Ichigo, p. 176, emphasis by this author)

Paul Williams nicely summarizes this new idea of svasavedana as follows:

10


11/21

Thus the character of self-awareness here has nothing to do with taking itself as an object in

a way which might lead to an infinite regress. Rather, self-awareness means reflexivity,where there is no sense of referring to an actual subject/object relationship and, therefore, no

stage of validation beyond the consciousness itself. Kamalala implies that it would not be

correct to think of self-awareness on the model of x is aware of y where y=x. In other

words, the epistemological model based on act and agent where an agent acts on itself isinappropriate, and, therefore, the common criticism of self-awareness found in other

Buddhist sources, grounded on the impossibility of an action directed towards itself, simply

does not apply in this case. (Williams, pp. 28-9)

It should be clear by now that model (1) of self-cognition was completely abandoned by

ntarakita and several later Buddhist philosophers. Matilal characterizes ntarakita as who

redefines self-awareness from the Buddhist point of view. (Matilal, p. 156) Willaims points out

that such formulation of self-cognition may have been a common interpretation in his time (8th

century). (Williams, p. 28)13

If such a new formulation of self-awareness originated around the time ofntarakita, then

the question of how Dharmakrti would think of self-awareness becomes a very interesting issue.

Many scholars assume that model (1) of self-cognition was abandoned by Dharmakrti as

well.14

Indeed, Dharmakrtis famous verse All cognitions (citta) and feelings (caitta) are

13 The reasons underlying such a transition between two models for self-cognition may be too complicated to be

explored here. To name but a few of them, I suggest that the following aspects must be taken into account:

(1) the critique of the infinite regress: namely, the final raya must be self-justifying instead of being justified

by others.

(2) the critique that the action cannot be acted onto the agent itself. Note that both Bhviveka and Candrakrti,

following Ngrjunas Vigrahavyvartan, denied the existence of self-cognition based on this critique. See

BhvivekasMadhyamakahdayakrik/Tarkajvl in Hoornaert, p. 160; and CandrakrtisPrasannapad in Ruegg,

p. 108-111.

(3) the incorporation of the Yogcra view into the Mdhyamaka system as the conventional truth. Namely, on

the level of the highest truth, such Yogcra terms as grhya, grhaka do not exist.

(4) the debate between the Skravda and the Nirkaravda

14 This includes: Mokkarakupta, Stcherbatsky, Matilal, Nagatomi, Dreyfus, etc.

11


12/21

self-cognizant; this is called self-consciousness (svasavedana) in the Nyyabindu has been

extensively quoted in order to support the model of self-luminosity of self-cognition. I am quite

sympathetic to this position since Dharmakrti thus characterized knowledge/cognition in the

Pramavinicaya:

[Excepting knowledge itself], there is nothing to be experienced by knowledge, and[likewise] it has no experience other [than self-experience]; since knowledge is deprived of

cognitum and cognizer, it is illuminated by itself. (Ichigo, p. 175, quoted from the

Pramavinicaya I: 38)15

That being said, I remain hesitant about such a conclusion about Dharmakrtis view of

self-cognition. One of the reasons for my hesitation is that Dharmakrti could have made the

above assertion not in the context of the relation between cognition and self-cognition, but

between external objects and the internal self. By deprived of cognitum and cognizer, He could

have aimed to deny the Sautrntika idea that there exist some external objects (dravya) as the

cognitum and the internal self as the cognizer. His goal here could have been to deny the

independence of objects from consciousness. By illuminated by itself, Dharmakrti could

have meant to argue that all knowledge/cognition is consciousnesss self-knowledge of itself but

not that all knowledge is self-luminous and hence self-justifying, as later philosophers assumed.16

To conclude, the goal of this section is to show that the understanding of self-cognition

underwent a dramatic change at least afterntarakita. The model of self-luminosity became

the dominant model of self-cognition later. For this reason, I would argue against the idea that the

model of self-cognition remained the same throughout all Buddhist philosophers in India.

15 nnyo nubhvyo buddhysti tasy nnubhavo para |

grhyagrhakavaidhuryt svaya saiva prakate || (Ichigo, p. 235)

16 Another reason for my hesitation will be discussed in the final section on mental perception.

12


13/21

Section Four: Mnasa-pratyak a in the two different models for self-cognition

This section aims to contrast the roles played by mental perception (Mnasa-pratyaka)

under the two different models of self-cognition. It is my claim that mental perception is

necessary in the first model but not in the second model.

To begin with, it is noteworthy that the necessity of the mental perception has been cast

more and more doubt after Dharmakrti.17

For example, as early as Dharmottara, in his

Nyyabinduk, says about the mental perception:

(11.1) This internal sensation is a postulate of our system. There are no facts to prove it(directly). But there is no contradiction in admitting it, if it were of the described kind. In

this sense its definition has been given. (Stcherbatsky, Vol. II, p. 28)18

This disregard of mental perception was most sharply expressed in the Tarkabha. After

admitting that it is not possible to find a justification for it, Mokkarakupta says:

If mental perception fits in with such a definition as made above, there is no [logic] fault

found. Thus the testimony of the sacred text is shown to be impeccable. This is the purpose

[of admitting mental perception] as a species of indeterminate knowledge. (Kajiyama, 47)

As Mookerjee points out, mental perception was actually omitted by ntarakita.19 We have

also seen in section two that it was also around ntarakitas time that model (2) for

17 For a more detailed description, see Hattoris notes 46 to the Pratyaka chapter (Hattori, p. 93) and Nagatomis

article (pp. 243-244).18 As Mookerjee comments: Dharmottara seems to offer the key to the solution of this tangle by observing that

mental perception is an accepted doctrine for which there is no logical warrantIt is clear, therefore, that

mano-vijna has no epistemological importance and can be jettisoned without harm. The inclusion of it in the

scheme of perception is made only in deference to scriptural authority and not for any logical or epistemological

necessity. (Mookerjee, p. 315)

19 Mookerjee, p. 311

13


14/21

self-cognition became dominant. My question then would be: are these two shifts of theory

connected to each other?

Now the question falls back to Dignaga: is mental perception of the same insignificance in

Digngas prama theory as it is in later works? Is the justification for the testimony of the

sacred text the only reason why Dignga postulated mental perception as belonging to one of the

indeterminate pramas? In what follows, I will argue that mental perception plays an

indispensable role in so far as self-cognition was understood by Dignga through model (1).

To reexamine the necessity of mental perception, let us return to one of the difficulties in

Digngas theory of indeterminate perception that I designated as the reproduction of a previous

cognition in the first section. To put it simple, the difficulty there is about the possibility of the

retention of the previous cognition. If, according to Dharmaplas model of self-cognition,

self-cognition means one cognition cognized by a (at least logically) succeeding cognition, then it

is necessary that the previous cognition must still be present in the following moment. But how

could this be possible given the Buddhist presupposition of universal flux or momentariness?

Adopting again Hattoris model here, we then have:

(External?) Object20 Sense Perception Self-cognition

t1 O1 S1O1

t2 O2 (S1O1)S2

20 Here I follow the conventional wisdom that it is Digngas view to try to make his theory acceptable to both the

Sautrntikas, who admit the independence of external objects from consciousness, and the Yogcrins, who deny

such an independence. This view has been described by Dreyfus as the ascending scales of analysis. Cf. Dreyfus,

pp. 49 ff.

14


15/21

In order for self-cognition C2= ((S1O1)S2) to be possible, the subsisting of C1=(S1O1)

at t2 must be possible. But if the cognition C1=(S1O1) would subsist at t2, then it is necessary

that a thirdterm must be able to re-produce C1 at moment t2. Such, I would argue, is exactly the

specific role played by mental perception. Namely, if the role of the mental perception could be

admitted, then we have:

(External?) Object Sense Perception Mental Perception self-cognition

t1 O1 S1O1 (S1O1)a

t2 O2 (S1O1)b (S1O1)S2

In order for self-cognition C2=((S1O1)S2) to be possible, it is necessary to admit that

what is re-produced by the mental perception must be exactly the same as what is grasped by the

sense-perception. Namely, we have to admit that (S1O1) must still have the same content as it

was cognized by S2 at t2. It has been reported that Kumrila set forth a critique that mental

perception could not be a prama (tools for knowledge) since the content is not new at all.21

Only if Dignga and Dharmapla admit that the content of sense perception remains exactly the

same can Kumrilas critique be meaningful.

A possible opposition to this diagram is that why not we just grant sense perception the

function of re-producing (S1O1)? My answer to this would be: since at t2, what we have is no

more object O1 but O2, and since it is the function of the sense perception to grasp (external)

object alone, it is not possible to argue that the sense perception is able to re-produce (S1O1) on

21 This is the so-called principle of anadhigatrtha-gant. See Hattori note 1.24. For Kumrilas critiques of the

redundency, see Mookerjee, pp. 311-312.

15


16/21

its own at moment t2.

Another reason for postulating such mental perception is due to Digngas assertion that the

self-cognition is not dependent on any sense-organ (PSV Db, Hattori, p. 27). Given that sense

perception C1=(S1O1) must be dependent on sense-organ and the necessary retention of its

content (S1O1), it is necessary that the content (S1O1) must be re-produced somehow by the

faculty other than sense-organs. The only candidate to achieve this is the mental consciousness

(mano-vijna) given Digngas theory of six consciousnesses. This, I suspect, is why the

indeterminate perception (pratyaka) that plays this role is named after mano-vijna.22 In other

words, mental perception serves to impress the content of sense perception inwardly and then

re-produce such content. Moreover, by means of the memory argument, we can prove that mental

perception is able to re-produce a pervious sense perception inwardly, as long as both of the

mental and sense perceptions belong to the same consciousness continuum (satna).

However, it is not clear to me whether such re-productive function of mental perception

occurs at t1 or t2 if we follow Hattoris diachronic analysis. More specifically, whether the

function of mental perception mainly refers to (S1O1)a or (S1O1)b is not clear in the above

diagram. In fact, with the recognition of the function of mental perception, we do have two

different theories. On the one hand, we have Dharmaplas theory (as reported by Kuiji) that

mental perception functions at the same moment as sense perception, the so-called

22 I agree with Nagatomi that mental perception is closely related to mano-vijna. But I cannot agree with his

assertion that PS k. 6ab refers not to, as Hattori understands it, two different kinds of m nasa-pratyaka, but to

object-cognizing and self-cognizing aspects of it. (cf. Nagatomi, p. 254) I dont see how, according to Nagatomi,

the object-cognizing aspect of mental perception could be differentiated from sense-perception, which is also

object-cognizing. Moreover, to identifying the self-cognizing aspect of mental perception as having the

appearance of passion, etc. (Cf. Nagatomi, p. 255) seems to me to be just collapsing the distinction between mental

perception and self-cognition.

16


17/21

five-accompanying mental perception.23

On the other hand, we have Dharmakrtis theory that

mental perception functions at the following momentof sense perception, with the latter being the

immediately preceding homogeneous cause (samanantara-pratyaya) of the latter.24

Notwithstanding this complexity, my main concern here is to argue that mental perception is

indispensable only in a prama system that construes self-cognition through model (1). Hence,

once self-cognition is notunderstood by means of model (1), the function of mental perception

could be put into serious doubt. In other words, it is my claim that as long as self-cognition was

understood through model (1), it is never possible to deny its function in the pram a system.

That is to say, only under the situation that self-cognition has not been understood through could

mental perception be put into doubt.25

23(T43224a) As for the function of this five-accompanying mental perception, Kuiji points out that there are twofunctions for this mental perception: (1) to understand (the objects of the five sense perceptions) ((T43, no. 1830, p. 485, c10)); (2) to assist the five sense perceptions in their rising; the five do not arisewithout the leading/intriguing of mental perception. ((T43, no. 1830,

p. 485, c5-6)).

I must admit that I am not sure about what exactly five-accompanying mental perception means here. The

details remain to be explored.

24 Cf. for example,Nyyabindu, Stcherbatskys translation,Buddhist Logic Vol. II. p. 26

25 This paper does not try to prove that the only function of mental perception is to reproduce the content of sense

perception so that self-cognition through model (1) could be possible. Stcherbatsky has reported a different argument

for the existence of mental perception by Jnagarbha, which may establish one more function of mental perception.

Since sensation and mental construction are, in this system, two quite heterogeneous sources of knowledge,something intermediate must be found which would be sensous on one side and mental on the other, in order to

account for a knowledge which combines sense-data with mental construction. Thus the existence of an internal

sense is proved by the existence of a subsequent mental construction (nla-mano-vijnt

samna-jtya-nla-vikalpa-udayt). (Stcherbatsky, pp. 28-29, note 3.)

Further, Stcherbatsky comments that Dharmottaras view of mental perception is evidently not what was meant

by DigngaAfter having established a radical distinction between the parts of the senses and of the intellect in

17


18/21

Recently, scholars have been debating about whether Dignga admitted three or four

pramas.26 If we examine this question from a philosophical point of view, then I would argue

against Hattoris subsuming of self-cognition under mental perception. The reason for Hattori to

subsume self-cognition under mental perception is that he mistakenly took mental perception to

be a cognition of external objects, and took self-cognition to be a cognition of internal objects.27

If such should be the case, then it seems difficult to me how the function of mental perception is

to be differentiated from sense perception. Moreover, Hattori does not see the dramatically

different role played by mental perception from that by self-cognition. The difference cannot be

put as cognizing external object vs. cognizing internal object, although it is true that

self-cognition always cognizes internal objects. The idiosyncratic function of mental perception

is just to impress, to transform the external object cognized by sense perception into internal

object that is to be cognized by self-cognition. For this reason, I would side with Yao and argue

that there are four pramas for Dignga, each with its specific function.

Finally, let me go back to the question of whether self-cognition was understood by

Dharmakrti through model (1) or model (2). One more reason why I suspect that Dharmakrti

cognition, Dignga was evidently in want of something which would be partly sensuous and partly mental. (ibid.)

Here Stcherbatsky tries to show another function of mental perception, i.e., its necessary mediation between the

sensuous and the mental in Digngas system. Since the dramatic heterogeneity between the sensuous and the mental

was still admitted in later theories, I suspect that Stcherbatsky may have difficulties explaining why mental

perception became less and less important in later prama theories.

For this mediating function of mental perception, see also Hattori note 1. 46, p. 93.

26 For a detailed reexamination of the debate, see Yao, pp. 214 ff.

27 Compare Hattoris translation of the PS k 6ab:

There is also mental [perception, which is of two kinds:] awareness of an [external] object and self-awareness of

[such subordinate mental activities as] desire and the like, [both of which are] free from conceptual construction.

(Hattori, p. 27)

18


19/21

adopted model (1) is because, unlike ntarakita, he still did notabandon the necessity of mental

perception. According to the above discussion, if Dharmakrti understood self-cognition in terms

of model (2), then he could just leave mental perception in the periphery and did not have to

strain his brain to modify the definition of mental perception in order to meet Kum rilas harsh

challenges.

Conclusion:

It may seem to the readers that this paper aims at a Dharmaplian reading of Digngas

theory of self-cognition. It is not the main concern of this author to prove that this reading is the

only coherent reading of Digngas theory of self-cognition.28

Instead, the main concern is to ask

whether we will betteranswer the following questions by adopting a Dharmaplian reading of

Digngas theory of self-cognition. These questions include:

1, Why Dharmapla ended up with such a theory of fourfold appearance that sounds odd

both from the angle of the Yogcra theory of consciousness and from the angle of Indian

prama theory?

2, Why mental perception was not regarded as an indispensable part of the prama theory

afterntarakita?

3, How could mental perception have been necessary in Digngas theory of prama?

The answers to these questions have been suggested by establishing an interrelationship

between the model (1) for self-cognition and the necessity of mental perception. To recapitulate:

28 However, this author does believe that self-cognition understood through model (1) is preferable if we want to

avoid the difficulty in claiming that mental perception was first postulated by Dignga, although he did not deem it

as indispensable.

19


20/21

Answer to 1: the reason why Dharmapla proposed the theory of fourfold appearance was

because he read Digngas theory of self-cognition by adopting model (1) and tried to avoid the

difficulties that might arise out of it.

Answer to 2: The reason why mental perception was not regarded as an indispensable part of

the prama theories after ntarakita was because it became dominant to understand

self-cognition through model (2), under which mental perception is no more necessary for

self-cognition to be possible.

Answer to 3: The way how mental perception could have been necessary in Digngas

prama theory is that mental perception makes possible self-cognition understood through

model (1). In other words, if we understand self-cognition in Dignga through model (1), then

mental perception becomes indispensable.

References

Dreyfus, Georges, B.J. 1997.Recognizing Reality: Dharmakrtis Philosophy and its Tibetan

Interpretations. Albany, N.Y.: SUNY Press.

Hattori, Masaaki. 1968.Dignga, On Perception. Cambridge: Harvard University Press.

Hoornaert, Paul. 2001. An Anotated Translation of Madhyamakahdayakrik: Tarkajvl

V.85-114, translated by Paul Hoornaert. In Kanazawa Daigaku Bungakubu Ronshy

23: 139-170.

Ichigo, Masamichi. 1989. ntarakitas Madhyamaklakra, in Luis O. Gmez and

Jonathan A. Silk ed. Studies in the Literature of the Great Vehicle. Ann Arbor: Collegiate Institute

20


21/21

for the Study of Buddhist Literature and Center for South and Southeast Asian Studies, the

University of Michigan: 141-240.

Kajiyama, Yuichi. 1998. An introduction to Buddhist philosophy: an annotated translation

of the Tarkabh of Mokkaragupta: reprint with corrections in the author's hand.

Wien: Arbeitskreis fr Tibetische und Buddhistische Studien, Universitt Wien.

Matilal, Bimal Krishna. 1986. Perception: an essay on classical Indian theories of

knowledge. Oxford: Clarendon Press; New York: Oxford University Press.

Mookerjee, Satkari. 1975. The Buddhist philosophy of universal flux: an exposition of the

philosophy of critical realism as expounded by the school of Dign

ga. Delhi: Motilal

Banarsidass.

Nagatomi, Masatoshi. 1980. Mnasa-Pratyaka: A Conundrum in the Buddhist Prama

System, in M. Nagatomi et al. ed. Sanskrit and Indian studies: essays in honour of Daniel H. H.

Ingalls. Dordrecht, Holland; Boston: D. Reidel.

Ruegg, David Seyfort. 2000. Studies in Indian and Tibetan Madhyamaka Thought. Vol. I.

Wien : Arbeitskreis fr Tibetische und Buddhistische Studien, Universitt Wien.

Stcherbatsky, F. I. 1958.Buddhist Logic. Vol. II. 's-Gravenhage, Mouton.

Williams, Paul. 1998. The reflexive nature of awareness: a Tibetan Madhyamaka defense.

Richmond, Surrey: Curzon.

Yao, Zhihua. 2003. Knowing that One Knows: the Buddhist Doctrine of Self-cognition,

unpublished doctoral dissertation, Boston University.

21

Is Mental Perception Necessary in a Pramāṇa Theory?

Documents