[Candrakirti's Commentary on the "Middle Way"] states:
I say they exist for the effect, even though they do not.
Taking the perspective of the world, I speak of a self.


You cannot use this as a source to prove that all positions are taken only in terms of the perspective of others. Why? One posits phenomena as lacking essential or intrinsic nature only in the purview of reasoned knowledge that properly analyzes whether such a nature exists. One does not posit this in the purview of an ordinary conventional consciousness.


Why? Were that ordinary consciousness able to establish the lack of intrinsic nature, then reasoned knowledge would be pointless, which is absurd. Also, since the same text says, "I say they exist... taking the perspective of the world,"


it means that the existence of forms and such is posited in terms of the perspective of the world.


When that text, in the lines prior to those just cited, states that we do not accept the conventional, it means that we do not accept dependent entities as Cittamatra accepts them; it does not mean that we do not in our own system accept the conventional, for it says "real dependent entities such as you accept."


This stanza responds to a Cittamatrin argument given in a transitional passage of the commentary on that text: "If you use what is valid or reasonable to refute dependent entities, I will use your arguments to refute what you consider the conventional."


Hence, the meaning is this: You Cittamātrins believe dependent entities to be things that can withstand rational analysis. I do not accept such conventional phenomena. Hence, we disagree as to whether one can use reason to refute them.


We do not take the expression "taking the perspective of the world" to refer to the perspective of others who do not belong to our own system; rather it refers to unimpaired conventional consciousnesses. This is because the existence of conventional objects must always be posited within the purview of such a consciousness and because the valid cognitions that posit conventionalities exist even in the Mādhyamika's own mind-stream.


Therefore, we take the words "even though they do not exist" to refer to their lack of existence by way of intrinsic character. So the passage should be glossed as, "I tell them they exist, even though they do not exist by way of their intrinsic character." It is inappropriate to gloss it as, "I say they exist even though they do not exist."


This is because this passage represents our way of positing conventional objects in which existence by way of intrinsic character is not possible even conventionally.


Also, in the portion of the commentary to his 《Commentary on the "Middle Way"》 that explains that passage, Candrakirti states that "I accept whatever exists or does not exist for the world just as the world does." Thus, that passage cannot be taken to mean that things do not exist.


Therefore, since there are many such instances in which the text states that even though things do not ultimately exist, they nonetheless exist conventionally, there is no the slightest fault in taking the existence that is affirmed and the exisence that is denied as having different meanings in the pass "I say they exist even though they do not."


Objection: You still have to explain what Nāgārjuna's 《Refutation of Objections》 means when it says that Mādhyamikas have no positions and theses.


If you adopt the thesis, "the seedling has no essential or intrinsic nature," then you also have to accept the reason, "because it is a dependently arisen thing," and the example, "for example, like a reflected image."

- p.479 -

Accordingly, you accept that a syllogistic statement creates for the opponent (1) areason that has bearing on the subject and has the two forms of pervasion, (2) a probandum that the reason proves, and (3) an inference that understands that probandum.


This being the case, then—aside from simply hating the name "autonomous syllogism"—why do you go to the trouble of refuting autonomous syllogisms?


Reply: The Refutation of Objections does explain that we have no theses and positions, just as you have cited it, but since there are also many passages that state that it is necessary to posit assertions,


how can simply citing that one passage prove that Mādhyamikas have no positions?


It is quite true, however, that there has been concern that anyone who holds the absence of intrinsic existence as a thesis is a Svatantrika. This is because this is an exceedingly subtle and difficult point.


Our response to that will be explained when we set forth our own position.


As for Nāgārjuna's statement in the 《Refutation of Objections》 that he has nothesis, Mādhyamikassay that things lack intrinsic nature.


The essentialists argue that if the words of such a thesis have intrinsic nature, then it is incorrect to say that all phenomena lack intrinsic nature; if those words lack intrinsic nature, then they cannot repudiate the existence of intrinsic nature.


However, something without intrinsic nature can function to prove or to disprove; according to previously cited passages from the 《Refutation of Objections》 and its commentary, this is admissible.


Therefore, the issue as to having or not having theses is not an argument about whether Nāgārjuna has them in general. It is instead an argument as to whether the words of the thesis "all things lack intrinsic nature" have intrinsic nature.


Hence, the meaning of the lines from the 《Refutation of Objections》 is this: If I accepted that the words of such a thesis had an intrinsic nature, then I could be faulted for contradicting the thesis that all things lack intrinsic nature,


but because I do not accept that, I cannot be faulted.


That is why these lines cannot be taken as proving that Mādhyamikas have no theses, for there is a very great difference between the absence of intrinsic existence and nonexistence.


The verse of the 《Refutation of Objections》—"If sensory perception and so forth..."—does state that sensory perception and such have no perception at all. However, according to the previously cited passage from the 《Clear Words》, this means that valid cognition and the phenomena it perceives are not essentially existent perceived objects and perceiving agents. It does not mean that there are no dependently arisen valid cognitions or phenomena that they perceive.


In [the essentialists'] opinion, even if the intrinsic character of things as established by sensory perception could be refuted, the Madhyamikas' claim that all things are empty of intrinsic nature would still require that sensory perception and the objects perceived by it be empty of intrinsic nature inasmuch as they are included among "things." In that case, sensory perception and the objects perceived by it would in their view be nonexistent; hence one cannot refute the intrinsic character of things.


That passage from the 《Refutation of Objections》 which states that sensory perception lacks perception is given in answer to this essentialist view, a view which it states in these lines:
If the things perceived by sensory perception
Were then repudiated,
The sensory perception by which things are perceived
Would itself be nonexistent.


The Commentary to those lines states:
Nor is it possible to claim that perception observes all things and then to repudiate them by stating, "All things are empty."

- p.480 -

Why? Valid sensory perception is included among "all things," and is therefore empty.


That which perceives things would also be empty. Therefore, there would be no perception by valid cognition. What cannot perceive also cannot refute, and hence the claim that all things are empty is not valid.


As for the lines from Aryadeva's Four Hundred Stanzas, "You can never rebut...," Candrakirti's commentary teaches us that one cannot refute the advocates of emptiness, no matter how long one tries.


Since you claim that you do not even assert emptiness, how can you cite this passage in support of having no assertion at all?


Candrakirti's 《Explanation of the "Middle Way" Commentary》 quotes those four lines in the following passage:
It is incorrect for those who advocate the position that things exist only as designations to advocate dualism.Therefore those who advance critiques and responses grounded in dualism against the Madhyamaka never find the slightest foothold. As Aryadeva's Four Hundred Stanzas says,

Hence, Candrakirti uses this passage to explain that neither the essentialists, who hold that things exist essentially, nor the nihilists, who believe in the total repudiation of the functioning of things such as form, can refute those who accept imputed existence while repudiating essential or substantial existence. Consequently, that passage cannot be used as a source for the claim that Mādhyamikas have no position of their own.


It is quite clear that the positions of existence, nonexistence, and so forth refer to instances of dualistically advocated positions. Hence, they are to be explained just as they were previously in our treatment of the refutation of the tetralemma and the way to refute the advocates of existence and nonexistence.


Candrakirti's commentary on Nāgārjuna's "Sixty Stanzas of Reasoning" passage, "Mahātmas have no positions...," states:


This commentary gives the nonexistence of real things as the reason for having no position. Moreover, "real thing" here must be taken to refer to intrinsic character or intrinsic nature, for to construe it as referring to functionality would contradict the statement that seeing its nonexistence stops the afflictions.


Therefore, insofar as Mādhyamikas have no positions in which they accept intrinsically existing things, this text says that they have no positions.


This is because Candrakirti's 《Commentary on the "Sixty Stanzas of Reasoning"》, in the passage preceding this verse, identifies the assertion of real things as the reification of intrinsic character in things. Candrakirti's Commentary and its root text say:
As for those who have not fathomed this reality—dependent-arising—and who believe completely in the intrinsic character of things:


Therefore, these scriptural passages do not teach that the Mādhyamikas have no system of their own.


Hence, this is also how you should understand the meaning of the 《Clear Words》 passage that says, "...because we do not accept others' positions...," citing the 《Refutation of Objections》 and the 《Four Hundred Stanzas》.


"To be refuted" in the Refutation of Objections passage that states, "Since there is nothing to be refuted, I refute nothing," might be interpreted in two ways. If we take "what is to be refuted" as referring to the object of the reification of things as intrinsically existent, it makes no sense for Nāgārjuna to say that because intrinsic existence does not exist, he refutes nothing.


Hence, we must take it as referring to the reifying consciousness that is refuted.

- p.481 -

Candrakirti's commentary on that passage says that even the refuting agent does not exist; hence the nonexistence of both what is refuted and the refuting agent refers to the nonexistence of a refuted object and a refuting agent that exist by way of intrinsic character.


[The 《Commentary on the "Refutation of Objections"》] states that the essentialist, taking the refuting agent and object of refutation to existin that way, that is, by way of their intrinsic character, insults us by saying that we do that kind of refutation. Nāgārjuna does assert that both the refuted object and the refuting agent are like illusions.


Why? His 《Refutation of Objections》 states:
An emanation can send out an emanation,
And an illusory being can use illusion
To stop illusion.
This refutation works in the same way.


It also states that if the apprehension of a mirage as water intrinsically existed it would not occur in dependence upon its causes and conditions, and no one could overcome that apprehension:
If that apprehension intrinsically existed,
It could not arise dependently.
Can any apprehension that arises dependently
Help but be empty?
If an apprehension intrinsically existed,
Who could overcome it?
The same procedure works for everything else.
That is why there is no reply to it.


The 《Clear Words》 passage that states, "The opposite of the absurd consequence in a reductio pertains to the opponent, but not to ourselves, for we have no theses," can also not be taken as a source to prove that Madhyamikas have no system of their own, for that passage means that they have no autonomous theses.


What about the 《Commentary on the "Middle Way"》 passage in which Candrakirti says that he has no positions? In his own system Candrakirti holds that neither his refutation nor the position he is refuting exist intrinsically. Therefore, when he uses reasoned analysis to refute the other party's belief that causes intrinsically give rise to their effects, asking whether causes give rise to effects through contact with them or without contact with them, this refutation does not apply to Candrakirti himself because he does not have to assert things that can withstand rational analysis. This is what that passage means.


It does not at all imply that we have no system of our own, for Candrakirti's Explanation of the "Middle Way" Commentary comments on that passage:
We do not fall into a similar absurdity in our own system because, from our perspective, the refutation does not refute what it refutes by contacting it, nor does the refutation refute what it refutes without contacting it. This is because neither the refutation nor what it refutes intrinsically exist. Therefore, we do not consider whether they make contact.


Why is the rational analysis proposed by the essentialists not applicable to the Madhyamaka refutation? He says it is because the refutation and what it refutes do not intrinsically exist; he does not say that it is because Madhyamikas have no assertions.


In support he cites the 《Perfection of Wisdom Sūtra in Twenty-five Thousand Lines》 where Sariputra investigates and asks Subhuti about what kind of phenomenon, produced or unproduced, brings about the attainment of an unproduced phenomenon. After Sariputra has refuted that either kind brings it about,


he then asks, "Is it that there is no attainment and no clear knowledge?"


and Subhuti says, as in the earlier citation," that these two do exist, but not dualistically.

- p.482 -

Using this as an example, Candrakirti's 《Explanation of the "Middle Way" Commentary》 gives this explanation: Because it would lead to such dualistic absurdities, the sūtra denies that a produced or unproduced phenomenon brings about attainment. Moreover, as dualistic analysis is inappropriate in the absence of real things, attainment is accepted without analysis, according to the conventions of the world. Likewise, it is neither that a refutation and what it refutes come into contact nor is it that they do not come into contact. Nonetheless, you should realize that conventionally a refutation refutes what is to be refuted.


This clearly states that when analyzed in terms of contact or noncontact, refutation does not exist in either way. But since this analysis does not deny the existence of refutation, Mādhyamikas must accept that conventionally there are refutations of the positions taken by others.


What is more, Candrakirti also accepts the fact that a syllogistic reason establishes the probandum. How so? In another passage that immediately follows the one cited above, his 《Commentary on the "Middle Way"》 and 《Explanation of the "Middle Way" Commentary》 state:


This is how it should be understood: Since a reflected image does not exist in the least, it is totally impossible to consider whether it is produced through contact or without contact with the orb of the sun. Nonetheless, when you see in a reflection a nearby form that is a causal condition for that reflection, you ascertain the object that you are seeking to understand. Likewise, a refutation that is empty of intrinsic nature refutes what is to be refuted, and a reasoning that is "improper" and is empty of intrinsic nature proves the point that is to be proven. Since this does not lead to absurdities of a dichotomizing analysis, it is incorrect for you to claim that our own words suffer from absurdities similar to that which we find in your position. This is how you should understand it.


Thus, this is how Candrakirti explains his answer that the arguments we used against others cannot be turned around and used against us. He does not say that we have no system of our own.


Moreover, while analysis of whether a cause produces an effect with or without contact refutes those who accept intrinsic existence, Candrakirti gives our assertion of the absence of intrinsic existence as the reason why this fault does not apply to us. This is how he avoids that fallacy. He does not avoid it by saying that we have no system of our own.


For, the 《Explanation of the "Middle Way" Commentary》 states: How do we explain production? It is because both cause and effect are like illusions that this is not a fault for us; the things of the world exist. That analysis is applicable to those who maintain that cause and effect respectively are intrinsically characterized as the producer and the produced thing. But what about someone who maintains that since things are generated by false imputation, they are like illusions in that it is not their intrinsic nature to be produced? What about someone who holds that things lack intrinsic nature but are the objects of mental construction—like the falling hair that may appear to a person with eye disease? That person has no conception of intrinsic production. Therefore, there is no opportunity to fault us in the way just explained. Also, since the things of the World do exist when left unanalyzed, everything is established.


He gives the acceptance of intrinsic character as the reason why the fault applies to the other position, and the belief that things are like illusions as the reason why we do not have this fault.


Understanding this, you should realize how to present the Madhyamaka system as being free from fault.

- p.483 -

In general, there are innumerable instances of definitive scriptures and Madhyamaka treatises that make statements such as, "This is that way and not this way," or "Such things exist and such things do not." Hence, why should it be necessary to cite special scriptural passages to prove that these statements represent the assertions of their authors?


If it were necessary, then when explaining the meaning of passages where expressions like, "I accept such and such," and, "I believe such and such," are absent, it would be impossible to distinguish whether or not something represents the system and belief of a certain author.


But if you insist upon actual instances of expressions like, "I believe," "I accept," and 'I posit," these are abundant.


Nāgārjuna's 《Refutation of Objections》 states:
Apart from asserting conventions, We offer no explanations.


And also, Nāgārjuna's 《Sixty Stanzas of Reasoning》 states:
A thing that arose and was destroyed
Is labeled as having ceased; likewise,
Excellent beings accept cessation
That is like an illusion.


And: Everything that arises dependently,
Like a reflection of the moon in water,
Is neither real nor unreal.
Those who accept this are not seduced by dogmatic views.


Also, Nāgārjuna's 《Praise of the Transcendent One》 states:
What arises from causes
Cannot exist without them—
And so is like a mere reflection.
How can you not accept something so obvious?


And again from that same text:
Since it does not exist without what is felt,
Feeling itself is selfless.
You [Buddha] also accept that feeling
Does not exist intrinsically.


And also: You have taught action and agent
In a conventional way.
You accept that they exist
In mutual dependence.


And also:
An effect cannot arise
From a cause that has been destroyed,
Nor from a cause that has not been destroyed.
You assert that effects are produced as in a dream.


And also:
You hold that whatever arises Dependently is empty.


Candrakirti's 《Explanation of the "Middle Way" Commentary》 also states:
Experts are of the opinion that this position has no flaw and is highly beneficial. You should accept it without hesitation.


And also: Consequently, because we accept that things are dependently imputed, just as we accept dependent-arising—the merely conditional nature of things—our position does not face the absurdity of annihilating all conventions. Others should also accept the same.


These passages state that it is definitely necessary to accept a variety of positions, and there are many more like them.


Again, the 《Explanation of the "Middle Way" Commentary》 explains four theses: After stating those four theses, we explain how to establish them through reasoning: "This does not arise from itself, so how could it from something else? Nor does it arise from both itself and something else, and how could it arise without a cause?"


As the 《Clear Words》 also makes similar statements, the system of Nāgārjuna the Protector and Candrakirti does have its own beliefs, assertions, and theses.


(4) Refuting the fourth misinterpretation
This fourth system apparently accepts that intrinsic character exists conventionally, but refutes the conventional existence of intrinsic character that withstands rational analysis. We have already explained that this is wrong.

- p.484 -

It also maintains that in the system of the master Candrakirti, when Madhyamikas use other-centered arguments to prove something to essentialist opponents, they use reasons that meet the three criteria and that are established for both systems. This is incorrect because (1) the 《Clear Words》 specifically refutes this view,


and (2) if you hold such a view, even if you do not call that "a reason based on real fact," it is inevitably an autonomous reason. I have yet to explain these points, so that is all I will say at this point.

第二安立自宗。述應成派破自續宗而立自宗,二宗俱解,當如是說。明顯句論多說此事,然恐文繁,今於此中略顯宗要。 Setting forth our own position
Explaining how Prasangikas set forth their own system via a refutation of Svatantrika brings about an understanding of both systems, so this is how I will proceed. Candrakirti's 《Clear Words》 has a great deal to say on this matter, but I am wary of excessive verbiage, so I will teach only the main points here.

此中分二,一 正破自續,二 自不同破之理。

There are two divisions: The actual refutation of autonomous argument Why the faults we find in autonomous arguments do not apply to us

初又分二,一 顯示所依有法不極成之宗過,二 由此過故顯示因亦不成。

The actual refutation of autonomous argument has two divisions: Demonstrating a fault that pertains to the position— namely, that the basis or subject of an autonomous syllogism is not established Demonstrating that, because of that fault, the reason is also not established

初又分二,一 出計,二 破執。  

The subject of "an autonomous syllogism is not established" has two parts: stating what Bhavaviveka believes, and refuting it.

今初 What Bhavaviveka believes


The passages from Candrakirti's 《Clear Words》 on these points appear to be exceedingly difficult to understand, so let me explain by citing and explaining a passage from the 《Clear Words》.


It says: How? In the thesis "sound is impermanent," both the subject and the predicate are construed only in a general sense, not with specific qualifications. If they were taken under specific qualifications, the conventions "inference" and "what is inferred" would cease to exist. For example, if the subject were construed as "sound that evolves from the four great elements," it would not be established for the other party in the debate, the Vaiśesikas. And if it were taken as "a quality of space," the subject would not be established for us, the Buddhists. Likewise, when Vaiśesikas advance the thesis "sound is impermanent," to Samkhya opponents, if sound is construed as "produced sound," then it would not be established for the Sámkhyas. On the other hand, if sound is construed as "something that becomes manifest," then it would not be established for the Vaiśesikas themselves. Likewise, however it is placed in an argument, destruction, when qualified as something that requires some other cause than production itself, is not established for us Buddhists; whereas qualified as something that is causeless, it is not established for the other party, the Vaiśeşikas. Therefore, just as the subject and predicate are taken only in a general sense in the above cases, likewise, in the present case, it is the mere subject, without particular qualification, that is to be understood. This is what Bhāvaviveka says.


This means that when the Buddhists advance the thesis "sound is impermanent" to the Vaiśesikas, if "sound that evolves from the elements" is taken as the subject, it will not be established for the Vaiśeṣikas; if "sound as a quality of space" is taken as the subject, it will not be established for us.


Likewise, when the Vaiśesikas themselves advance the thesis "sound is impermanent" to those Samkhyas who are manifestationists; if "sound as something produced" were taken as the subject, it would not be established for the manifestationists;


if "sound as something that previously exists and comes to be manifested through certain conditions" were taken as the subject, it would not be established for Vaisesika proponents.

- p.485 -

Therefore, it is impossible to use as a subject something that is incompatible with one's own individual belief system.


For, given that the subject is the basis that both parties analyze in order to see whether a specific quality is present, it has to be something established as appearing in common to both.


Just as they must establish a commonly appearing subject, so too both sides must establish the predicate, "impermanent," only in a general sense, without particular qualifications. Also, they must commonly establish any example that they cite, and this must take place before they prove the probandum.


The situation is similar when we Mādhyamikas prove to non-Buddhists the nonexistence of something that is produced from itself—whether it is an internal sensory source, such as the visual faculty, or an external sensory source, such as a form—and when we prove to Buddhist essentialists that there is no production from other. If we were to use "a real eye," for instance, as the subject, it would not be established for us, but if we used "an unreal eye" as the subject, it would not be established for the other party. Hence, giving up such specificity, we must use the mere eye or mere form as the subject. Why? Because it must be established as commonly appearing to both parties, inasmuch as it is the basis that both Madhyamikas and essentialists analyze in order to see whether there is a specific quality, such as "being produced from itself."


This is what Bhavaviveka thinks. "To establish as appearing in common" means that the proponent and the opponent use the same kind of valid cognition to establish it.

第二破執分二,一 義不應理,二 喻不相同。 refuting what Bhavaviveka believes
This has two parts: showing that the meaning is incorrect showing that the example cited is not appropriate.

今初 showing that the meaning is incorrect


This is not like that, and is instead as follows. Insofar as one accepts the refutation of production as the probandum in this [demonstration that eye, etc. are not ultimately produced], the subject — the basis of that probandum and something found to have its own existence by an inaccurate consciousness — breaks down in reality, and therefore, it will be argued, there would be no commonly appearing subject. He [Bhavaviveka] himself must accept that this is so. Inaccurate and accurate consciousnesses are different. Therefore, when an inaccurate consciousness takes what is nonexistent as existent, as in the case of someone with eye disease seeing falling hair, it does not perceive even to the slightest degree an object that exists. When an accurate consciousness does not reify what is unreal, as is the case of someone without eye disease looking for the imaginary falling hair, it does not perceive even to the slightest degree objects that are nonexistentinsofar as they are merely conventional.


That is why the master Nāgārjuna himself states in his Refutation of Objections:
If sensory perception and so forth
Could actually perceive something
Then there would be something to prove or to refute.
But as they do not, I cannot be faulted.


Since inaccurate and accurate consciousnesses are different in this way, inaccurate consciousness cannot exist when accurate consciousness is present. So how could the conventional eye, as the subject of a syllogism, exist for an accurate consciousness? Therefore, since there is for Bhāvavivekano avoiding the fallacies of a nonexistent position and a nonexistent reason, the response he has given is no answer at all.

- p.486 -

I will explain this passage based on the following syllogism, since using this example makes it easier to understand:
Subject: Visible form
Predicate: Is not produced from itself
Reason: Because it exists
Example: Like the pot that is right in front of you


The passages of the 《Clear Words》 that reply to Bhāvaviveka show that the subject is not established as appearing in common to the two parties in this debate.


How do they show this? Here, in this section, the 《Clear Words》 states that those to whom it demonstrates how one cannot establish a subject in common with an opponent are those who refute production from self—that is, the Svātantrikas. But in general, the opponents of the 《Clear Words》 are both (a) to essentialists, who accept that things ultimately have intrinsic nature, and (b) the Svātantrikas, who refute that, but accept that things conventionally have intrinsic character or intrinsic nature. Svātantrika-Madhyamikas are called "non-essentialists". However, so as to simplify the terminology in this discussion, "opponents of intrinsic existence" will refer to the Prāsangikas, and "advocates of intrinsic existence" will refer to both the essentialists and the Svātantrikas.


According to the advocates of intrinsic existence, visible form, the stated subject of the syllogism, must be established by the valid perceptual cognition of the visual consciousnesses that perceive it. Moreover, if those consciousnesses did not establish visible form in a non-mistaken way, then they could not be perceptions that establish their objects.


Hence, they must be non-mistaken. Do non-conceptual perceptions establish their objects in a non-mistaken Way? In non-Prasañgika systems, an object's existence by way of its intrinsic character appears to any consciousness that is non-mistaken with respect to that object; moreover, the object must exist just as it appears to that consciousness. As this is the case, we Prāsangikas say that the kind of valid cognition that establishes the subject for the essentialist opponent will not work for the Madhyamika proponent. Why? Since no phenomenon can, even conventionally, have a nature that is established by way of its intrinsic character, there is no valid cognition that establishes such a thing. It is with this in mind that the master Candrakirti refutes the notion of autonomous syllogism.


This also explains how to refute the need for an autonomous syllogism as part of the process of initially instilling in others the view that knows that things lack intrinsic nature. I leave aside for the time being the analysis of whether Prasangikas need to use autonomous syllogisms among themselves as part of the process of developing inferential knowledge of certain objects amongst the diverse conventional objects.


Now let me explain this by tying my analysis to Candrakirti's text. The meaning of the passage, "Insofar as one accepts the refutation of production... himself must accept that this is so," is as follows. "The basis of the probandum"—a subject such as the eye or form—"breaks down," that is, is not established, "in reality." This is something that Bhāvaviveka himself accepts.

- p.487 -

What are those subjects [eye and form] like? "Something found to have its own existence by inaccurate consciousnesses" affected by ignorance; this means that conventional consciousnesses, such as the visual consciousness, establish these objects.


How is it that "[Bhāvaviveka] himself must accept this?" He must accept it "insofar as," i.e., because the refutation of ultimate production, as the predicate of the probandum, is based upon those subjects;


if they did exist in reality it would contradict that relationship between subject and predicate granted that he accepts things in this way,


what does that entail? Those subjects—form, etc.—that neither exist in reality nor are reality itself cannot be considered objects found by non-mistaken consciousnesses. Hence, they are found by conventional consciousnesses, or subjects, that apprehend false objects. And thus those consciousnesses are mistaken; that is, affected by ignorance.


Therefore, objects found by non-mistaken consciousnesses do not appear to mistaken consciousnesses, and objects that appear to mistaken consciousnesses are not found by non-mistaken consciousnesses.


This is because "the inaccurate" mistaken consciousnesses "and the accurate" non-mistaken consciousnesses "are different," which is to say that each engages its object by excluding the object of the other. This is the meaning of Candrakirti's statement that "inaccurate and accurate consciousnesses are different."


The explanation of that occurs in the passage, "Therefore, when an inaccurate... are nonexistent insofar as they are merely conventional."


There, "inaccurate" refers to a conventional consciousness that is affected by ignorance, such as the visual consciousness. That such consciousness "takes what is nonexistent as existent" refers to the fact that while form, sound, etc. have no essential or intrinsic character, the sensory consciousnesses apprehend them as having such a character.


The way that non-conceptual consciousnesses apprehend their objects is in terms of mere appearance, and that is why form and such appear to sensory consciousnesses to exist by way of their intrinsic character.


The words, "it does not perceive, even to the slightest degree, an object that exists," mean that because intrinsic character appears, despite its nonexistence,


there is no way for those consciousnesses to establish even the slightest object that exists by way of its intrinsic character.


The falling hair is an example of an object that does not exist by way of its intrinsic character, yet appears as though it did.


These sentences mean that the sensory consciousnesses to whichthingssuch as formand sound appear are mistaken, and are therefore not suited to attest that an object exists by virtue of its intrinsic character.


The passage that begins, "When an accurate consciousness." indicates that non-mistaken consciousness does not at all apprehend form, sound, and such.


The word "accurate" refers to non-mistake consciousness. Noble beings who perceive reality possess such consciousness, and no one else.


That non-mistaken consciousness "does not reify what is unreal." This means that it "does not reify," or take as existing, such things as form and sound—which cannot be final reality.


For example, it is like the fact that the visual consciousness of someone without eye disease does not see an image of falling hair.


In that same sentence, the phrase, "insofar as they are merely conventional," refers to false objects, like form and sound.


"Nonexistent" means not existing by way of intrinsic character. Such conventional objects are not established even in part by non-mistaken consciousnesses, that is, by consciousnesses that have final reality as their object. This is because non-mistaken consciousnesses do not see such conventional objects.

- p.488 -

On these points, Candrakirti cites a proof-text by Nāgārjuna the Protector: "If sensory perception and so forth...." This supporting citation states that the four [direct, inferential, scriptural, and analogical] valid cognitions—sensory perception and so forth—do not at all establish an object that exists by way of its intrinsic character.


The sentence that begins, "Since inaccurate and accurate...," summarizes the point Candrakirti has already explained.


The sentence, "So how could the conventional eye, as the subject of a syllogism, exist," is not claiming that subjects such as the conventional eye are nonexistent.


Instead, as explained above, it means that a form that exists by way of its intrinsic character, or is established by non-mistaken perception, cannot be the subject of the syllogism even conventionally.


The meaning of "Therefore, since..." is that when both the opponents of essential or intrinsic existence and the essentialists posit visible form as the subject of a syllogism, non-mistaken perception does not establish it as appearing in common to both parties in the debate. Therefore, since there is no valid cognition attesting to a subject that is proven to appear in common for both systems, there will inevitably be a fault in any position that you try to prove to an opponent using an autonomous reason.


Objection: What you say is true in regard to a position that has no essential or intrinsic nature even conventionally. However, since this is not what we [who follow Bhavaviveka] assert at the conventional level, the subjects and such in autonomous syllogisms do exist. Therefore, the position is free from fallacy.


Reply: The existence of such an intrinsic nature is inadmissible conventionally. Since we have already explained this above, and will explain it again below, your answer is unreasonable.

◎第二喻不相同。 showing that the example cited is not appropriate.


Candrakirti's 《Clear Words》 states:
The example [i.e., the syllogism proving the impermanence of sound] is also inappropriate. Whereas in this example, neither party wants to qualify the general sense of "sound" or the general sense of "impermanence," here in the case of the syllogism proving that the eye is not ultimately produced, the advocates of emptiness and the advocates of non-emptiness do not both accept that eye, as a generality, exists only conventionally, nor do they both accept that it exists ultimately. That is why the example is inappropriate.


Do not misread this passage to say that the example fails because an eye that is neither true nor false does not exist, but there does exist a sound that is neither evolved from the elements nor a quality of space, as well as a sound that is neither produced nor a causal manifestation of something that already exists; and that there does exist something that is impermanent in a general sense, yet neither relies on causes nor fails to rely on causes.


For (1) those [i.e., sound that is neither evolved nor a quality of space, etc.) are things that neither of those two parties accept, and (2) if they were to accept such things, no one could ever demonstrate that the analogy fails.


Well then, what does this passage mean? In the systems of both of the parties in the example, it is possible to ascertain the existence of sound, unspecified as to whether it is "sound that is evolved from the elements" or "sound that is a quality of space."


But in the systems of the advocates of emptiness of intrinsic existence and the opponents of emptiness of intrinsic existence, there is no such thing as an eye or a form as a generality that is established by a valid cognition that is neither a non-mistaken consciousness nor a mistaken consciousness.

- p.489 -

Its being established by a mistaken consciousness is not established for the opponent,


and the Prasañgika-Mādhyamika proponent's valid cognition does not establish its being found by a non-mistaken consciousness. Hence, the analogy fails. This is the meaning of that passage.


The word "non-mistaken" generally refers to the equipoise that directly perceives the ultimate truth.


But here, it must refer both to a perceptual valid cognition that is non-mistaken with respect to an intrinsically characterized appearing object and to an inferential valid cognition that is non-mistaken with respect to its intrinsically characterized conceived object. Since there are no such valid cognitions that establish the three criteria, the subject cannot be an object that is found by a non-mistaken consciousness.


Here the term "intrinsic character" is not used, as the logicians use it, simply to mean something that performs a function.


Instead, as previously explained, it refers to something's own intrinsic nature, which any functioning thing or non-functioning thing is believed to have. That is why the advocates of intrinsic nature claim that even an inference that comprehends a non-thing is not mistaken regarding a conceived object that has such an intrinsic nature.


Every consciousness that is non-mistaken with respect to such an intrinsic nature must also be non-mistaken with respect to its appearing objects and conceived objects; and since this makes such a consciousness non-mistaken with respect to ultimate reality itself, our own system does not hold that such a valid cognition establishes the subject, etc.


However, we do not deny that there are, in the mind-streams of both parties, conventional valid cognitions that perceive things like eyes and forms.


In fact, even in the mindstream of the opponent, the forms, etc. that are elicited by sensory consciousnesses that are unimpaired—in the sense previously explained—are ascertained simply as existing, and there is no rational fault in regard to the object of such knowledge.


To explain this in greater detail, we can say that there are three ways of apprehending the existence of, for instance, a seedling: (1) apprehending a seedling as truly existing, which means apprehending it as having an essential or intrinsic nature; (2) apprehending it as existing in a false way, which is the apprehension that the seedling lacks essential existence, but exists like an illusion; and (3) apprehending it as merely existing in general, without specifying whether it is true or false.


You might also apprehend the seedling as permanent or impermanent, etc., but since there is no apprehension that does not involve one of these three ways of apprehending, there is no need to explain those other ways here.


Living beings who have not developed within their mindstreams the view that knows the absence of intrinsic nature posess the third and first modes of apprehension, that is, the apprehension of mere existence and the apprehension of true existence, but they lack the apprehension of things as like essenceless illusions.


It is completely wrong to claim that before living beings find the view that phenomena are like illusions, any conception they have of something as existing is a conception of true existence.


This is something that I have already explained above, in the section that discusses conventional valid cognition and in the section in which I differentiate the four—intrinsic existence, lack of intrinsic existence, existence, and nonexistence.


Suppose that this were not the case—that is, suppose that those who have not yet understood the view that there is no intrinsic nature did apprehend everything as truly existent whenever they thought of any conventional thing.

- p.490 -

There would ensue a complete logical breakdown of the need for Mādhyamikas to accept, conventionally, the objects that are posited by the world's ordinary conventional consciousnesses, insofar as those consciousnesses are not affected by the previously explained circumstances that cause error. Therefore, since there would be no way to distinguish the ontological status of conventional objects from the ontological status of a putative divine creator,


this erroneous view would be a great impediment to understanding the meaning of the Madhyamaka.


There are many who show indications of having misunderstood emptiness in this way. They initially engage in many virtuous activities that require conceptual thought. But later, when they systematize the philosophical view that they have found, they see all of their previous activities as grasping at signs and thus as binding them to cyclic existence.


They reflect, "Those virtuous activities were taught for those who have not found this definitive view."


Developing such an understanding, they repudiate the teaching in many ways with this misconception that regards all conceptual thought as faulty. In this sense they resemble the Chinese abbot Ha-shang.


Before they find the view that things lack intrinsic nature, it is impossible for them to distinguish between mere existence and existence by way of intrinsic character.


This is because—as indicated in the passage from Candrakirti's 《Commentary on the "Four Hundred Stanzas"》 cited above—they think that anything that exists must exist essentially.


As result of this, they take everything that lacks intrinsic nature to be nonexistent, making it impossible for them to posit cause and effect for that which is empty of intrinsic nature. There are many who argue in this way.


Those who have developed in their mind-streams the view that knows the absence of intrinsic nature may apprehend things as existing in all three ways.


When that view has been developed, and while its influence has not diminished, the conception of true existence that believes that things essentially exist is temporarily absent. This lasts as long as they are rationally analyzing something so as to determine whether it exists essentially. However, this does not mean that they do not have an innate conception of true existence.


Therefore, even those who have developed the view that knows the absence of essential or intrinsic nature, and who have not let that view degrade, do not always apprehend a seedling as existing like an illusion whenever they apprehend a seedling as existing.


Why? If they did, it would lead to the absurdity that they would never again develop a manifest form of the conception of true existence with regard to those seedlings and such.


There are Madhyamikas—such as the master Bhāvaviveka—who accept that, conventionally, phenomena have essential or intrinsic character. The conventional existence of essential or intrinsic character is their reason for accepting autonomous reasons in their own system. Whether one posits autonomous reasons in one's own system finally depends upon what one posits as the extremely subtle object of refutation.


Therefore, in their Svätantrika system, unimpaired sensory perceptions to which essential or intrinsic nature appears are, conventionally, non-mistaken with respect to their appearing objects.


The conceptual consciousness that conceives of a subject such as a seedling as having that kind of nature is also non-mistaken with respect to its conceived object.


Otherwise, if they accepted that those consciousnesses are mistaken, then what valid cognition would establish the elements of a syllogism as appearing in common for both their system and that of the essentialists?

- p.491 -

If the sensory perceptions to which essential nature appears establish the elements of a syllogism for the essentialist even though, as Candrakirti holds, there is no intrinsic nature such as the essential character that appears, then how can one use an autonomous reason? For, one would have already proven the absence of intrinsic nature to the essentialist while establishing the subject.


Opponent: Let the essentialist opponents establish the subject, etc. as they may; it is not necessary for the subject, etc. to be established as appearing in common to both the Madhyamikas and to them.


Reply: But that is something that Bhavaviveka himself does not accept, nor is it correct for him to do so,


for if he did it would entail that all probative reasons and syllogisms are offered simply in terms of what the other party accepts, and that would make him a follower of the Prāsangikas.


There are [Yogācāra-Svātantrika] masters such as Śāntaraksita who assert that external objects do not exist conventionally.


However, like those Cittamātrins who are Satyākāravādin, they assert that blue and such conventionally exist in the substance of consciousness. Therefore, since the sensory consciousnesses to which blue, etc. appear do have a view of them, apprehending them as existing by way of their intrinsic character, Sãntaraksita and other do not consider the sensory consciousnesses mistaken in their view of blue itself.


When hidden objects, such as the eye, are posited as the subjects of a syllogism, perception cannot explicitly establish them. Still, if we work back to the fundamental establishing agent, we must arrive at a perception.


This is a belief of all Buddhist philosophical schools. Why? Because inference is like a blind person who is guided by perception. So Śāntaraksita and his followers accept that, even in the case of hidden things, the fundamental establishing agent is, in the end, perception.


At that point, they believe that the fundamental perception is either a non-mistaken cognition of something else or a non-mistaken auto-cognition. Also, as I have already explained, they believe that an object that exists by way of its intrinsic character does appear and must objectively exist as it appears.


This being the case, there can be no non-mistaken perception that establishes anything as appearing in common both to them and to Mādhyamikas who maintain that there is no essential or intrinsic nature.


Even in the case of objects that cannot be traced back to perception, it is still possible to reply. The proponents of intrinsic nature claim that valid cognitions establish all objects, compounded and non-compounded. What do they mean by this? Is it necessary for those valid cognitions to establish objects whose ontological nature is to exist objectively? If so, then since reason can refute them, they cannot be valid cognitions that establish their objects.

◎第二由此過顯因亦不成。 Demonstrating the reason is also not established.


Candrakirti's 《Clear Words》 indicates this in the passage that states: The same method that was used to show that the position is defective insofar as its basis is not established should be used to show the defect that the reason, "because it exists," is also not established.


Previously, Candrakirti explained that because there is no valid cognition that establishes a commonly appearing subject for the systems of both parties—the proponents and the opponents of emptiness of essential or intrinsic existence—the thesis, or probandum, which combines the autonomous argument's subject, "visible form," with its predicate, "is not produced from itself," also does not exist.

- p.492 -

On this account, the reason, "because it exists." also is not established, for there is no valid cognition that can attest to its being established as appearing in common to both parties.


You should understand this on the basis of what has been explained above.


At this point, the 《Clear Words》 states:
This is so, for this logician [Bhāvaviveka] himself accepts implicitly the points we have just made.


How so? Another party offers him this proof: "The causes that serve to bring about the internal sensory faculties and such are existent, without qualification; this is the case because the Tathâgata said they are;


for whatever the Tathâgata has said is accurate, as in the case of his teaching that nirvâna is peace." [Bhāvaviveka replies,] "What do you believe the import of the reason to be? Is it that such causes exist because the Tathâgata has said so in terms of the conventional? Or is it because he said so in terms of the ultimate?


If it is true because he said this in terms of the conventional, the import of the reason would not be established for you.


If you take it that the Buddha made this statement in terms of the ultimate, since neither the probandum nor what proves it are established in terms of the ultimate, the reason would not be established and would in fact contradict the thesis."


This is how [Bhāvaviveka] states the fault in that proof. Since it is through such considerations that he asserts that the reason is not established, the reason and so forth are not established for him in any argument that posits a reified thing as the reason, and hence all probative arguments would fall apart.


Some Tibetans who consider themselves followers of Candrakirti interpret this passage in the following way: Bhāvaviveka's Blaze of Reasons and other Svätantrika texts put forward the following syllogism:
Subject: Earth
Predicate: Is not ultimately of the nature of solidity
Reason: Because it is an element
Example: Like wind


They say that Candrakirti refutes this approach as follows: If you posit "because it is an element ultimately" as the reason, then it is not established for us.


If you posit "because it is an element conventionally" as the reason, then it is not established for the opponent, the essentialist.


If this argument does not induce you to accept that your own reason is not established, then you contradict your belief that a reason that is not established from either of those two points of view [i.e., conventionally and ultimately] must be a reason that is not established.


And there are those who say that [Bhāvaviveka] is refuted because when he states just being an element as the reason, without any specification, reasoned knowledge does not establish it.


But [Candrakirti's] refutation [of Bhāvaviveka] does not proceed in this way, This is not at all the purport of the 《Clear Words》, nor dos, [Bhāvaviveka] accept any such thing. Hence, these individuals misrepresent both systems.


Well then, how is the passage to be interpreted? In the passage that states, "for this logician himself accepts implicitly the points we have just made," the phrase "the points we have just made" refers to the previously explained method for showing that the subject is not established and also to the application of that method to the reason, for the text states this in the immediately preceding passage.


This being the case, it follows that the valid cognitions such as perceptions that establish the subject and reason are inevitably either mistaken or non-mistaken.


If you posit an object found by a mistaken consciousness as the reason or as another part of the syllogism, then it will not be established for the essentialist;


if you posit an object found by a non-mistaken consciousness in that role, our own valid cognitions will not establish it.


Therefore, autonomous reasons and subjects are not established. This is what we explained above, and it is the meaning of Candrakirti's phrase, "the points we have just made." Bhāvaviveka himself asserts that this kind of analysis leads to positing that a syllogism's reason is not established.

- p.493 -

To show how Bhāvaviveka asserts this, Candrakirti cites Bhavaviveka's analysis in terms of the two truths [conventional and ultimate] of his opponent's reason, "because the Tathâgata said so."


Contrary to the interpretations proposed above, that analysis is not at all meant as an analysis of whether the stated reason is "because the Tathâgata has said so conventionally" or "because the Tathâgata has said so ultimately."


Why? As explained above, it is Bhāvaviveka's position that you must posit the subject without qualifying it either as real or unreal. Just as Bhāvaviveka accepts that subjects with such qualification would not be established for one or the other of the parties, he likewise accepts that such is the case for the reason, the example, and so forth.


Therefore, if Bhāvaviveka had faulted the essentialist's reason by applying qualifications such as "conventionally" and "ultimately," his argument would be a gross self-contradiction. How could this consummate scholar make such an error?


Therefore, the correct interpretation of Bhavaviveka's argument is as follows: Which of the two truths is the referent of the reason "because the Tathâgata said so"?


If it is the conventional one, it is not established for you, the essentialist, since you do not accept that the reason refers to a conventional object;


and if it is the ultimate, it is not established for me, since I refute the ultimate production of an effect from a cause that is existent, nonexistent, or both, as well as production that is causeless.


Since neither party accepts that there is an object that is neither of the two truths, it is not necessary for Bhavaviveka to clear that up.


You may interpret Bhāvaviveka's question to his opponent as, "When you say 'because it is an element,' which of the two truths is the element that is posited as the reason?" This interpretation is correct in that it is just like what we have explained above.


However, if you claim that Bhāvaviveka is asking, "In terms of which of the two truths [i.e., ultimately or conventionally] is the element stated as the reason?" then you completely misunderstand the position of the opponent [of the 《Clear Words》, that is, Bhavaviveka]."


If that is what Bhāvaviveka meant, then how could he say to his opponent, "Of the two truths, which is it? If the ultimate, then it is not established for us, and yet, if conventional, it is not established for the other party?"


If it were possible to claim that Bhāvaviveka's analysis refers to things existing ultimately and conventionally, then since the internal sensory sources that he posits as the subject in his syllogism refuting ultimate production exist conventionally, that subject would not be established for those essentialist opponents.


Well then, how can Candrakirti claim that Bhavaviveka accepts "the points we have just made" through his use of the two truths to analyze the reason given by the essentialist?


I shall explain. Here Candrakirti is of the opinion that what is found by a non-mistaken consciousness is the ultimate


and what is found by a mistaken consciousness is the conventional.


This being the case, the question, "Which of the two truths is it?" is conceived of as a question regarding which of the two consciousnesses it is that finds the reason; it has to be one of these two options.


For, if the object stated as a reason is neither the conventional nor the ultimate, then that reason cannot be established; also, if the object stated as a reason is not an object found by either a non-mistaken or a mistaken consciousness, then the object stated as the reason cannot be established. In that the reasons are parallel, Candrakirti states that Bhavaviveka himself accepts that his reason, "because of existing," is not established, but he does not say that Bhavaviveka accepts this explicitly.


That is why Candrakirti, in his 《Clear Words》, specifies "reified thing" when he states, "in any argument that posits a reified thing as the reason." The master Bhavaviveka believes that, of the reasons he posits, some are directly established by non-mistaken perception, while others are not, but are proofs that finally derive from non-mistaken perception. This master [Candrakirti] refutes this.

- p.494 -

To prove that it is incorrect to accept objects that exist by way of their intrinsic character, the earlier citation from the 《Clear Words》, "Mādhyamikas do not accept others' positions," —quotes passages such as the Refutation of Objections stanza that begins, "If perception and so forth...." By drawing from such citations the conclusion that there are no valid cognitions that perceive intrinsic character, Candrakirti aims to prove this point to the partisans of master Bhavaviveka's system.


> Why the faults we find in autonomous arguments do not apply to us.


Do our arguments not have the same faults that we find in others' arguments, such as the subjects and reasons not being established?


And if they do, should we not refrain from finding contradictions in others' arguments?


The reason others have those faults lies in their acceptance of autonomous arguments. [The 《Clear Words》] states that since we do not accept autonomous arguments, we do not have those faults.


Here, the term "arguments" refers to syllogisms.


If you accept autonomous syllogisms, then you must accept that both parties agree that there are valid cognitions that are valid in regard to intrinsic character. It is therefore necessary to prove the probandum by having both parties establish the three criteria with those non-mistaken valid cognitions.


But since such valid cognitions do not exist, the subject and the other parts of the autonomous syllogism cannot be established.


If you do not accept autonomous syllogisms, then you may allow the essentialist to use that kind of valid cognition to establish the subject, etc., but you yourself do not need to establish the subject, etc. with those valid cognitions.


Therefore, the arguments found in texts such as the 《Clear Words》 are "arguments based on what the other party accepts"; their sole purpose is to refute the other party's thesis. They are not autonomous arguments.


For example, the third chapter of Nāgārjuna's 《Fundamental Treatise》 says:
Sight does not see
Its own self.
How can what does not see itself
See something else?


This is like the argument that uses an eye's not seeing itself as a reason to prove that an eye does not see other things. In this case, the reason is accepted by the other party while Mādhyamikas also accept the thesis—that the seeing of other things lacks essential existence. Such a syllogism is called "an inference based on what others accept."


Candrakirti's 《Clear Words》 states:
We do not use autonomous arguments because the refutation of others' theses is the only effect of our arguments.


This means that the syllogisms he uses are not autonomous syllogisms. He does not say that he does not use syllogisms, for he does accept those that have as their sole purpose the refutation of others' theses.


The very next passage of the Clear Words states how he uses syllogisms to refute their theses: Those who think that the eye sees other things also hold that the eye does not see itself. They also believe that if the eye did not have the quality of seeing other things, it would not see.


So we argue: "Anything that does not see itself does not see other things either, as in the case of a pot. The eye also does not see itself; therefore, it too cannot see other things. Therefore, its seeing other things, such as blue, contradicts its not seeing itself. Your position is contradicted by this argument based on what you accept." This is how an argument that is established for them refutes their position.


When "what you accept" is addressed to the opponent, it is the same as saying "what others accept" from the point of view of the Madhyamika proponent.

- p.495 -

This is our procedure for refuting misconceptions by stating a syllogism based on what others accept. As this is very important, I will explain it in detail.


The phrase, "that is established for them," at the end of that passage does not mean that we do not, in our system, accept (1) the subject-the eye, (2) the example-the pot, (3) the reason—that it does not see itself, and (4) the predicated quality—that it does not see blue, etc. Nor does it mean that the reason, pervasion, and so forth are established only for the opponent since they are simply what others assert.


What does it mean? Even though we do accept these four in our own system, our system does not have, even conventionally, a valid cognition that establishes them and that understands its objects as existing essentially.


For proponents of intrinsic existence, the establishment of the subject, etc. definitely depends upon those being established by this kind of valid cognition. Therefore, there is no valid cognition which understands its object as existing essentially that can establish a common appearance of the subject, etc. for both parties.


Hence, they are said not to be established for both parties; they are said to be based on what is accepted by, or established for, others.


Query: If such valid cognitions do not exist even conventionally, then reason contradicts the acceptance of what they establish for the opponent, just as it contradicts the superimposition of intrinsic nature. How, then, could you find the Madhyamaka view in reliance upon their testimony?


For, if it were possible to find an accurate philosophical view using arguments that valid cognition contradicts, then you could use all sorts of inaccurate tenets to find such a view.


Reply: In this case, even our own system accepts as conventionally existent the objects that the opponent apprehends as existing— namely, (1) the subject—the eye; (2) the reason—that it does not see itself; (3) the example—the pot; and (4) the predicate—that it does not perceive blue and such.


So reason does not contradict these. Nonetheless, since the opponent fails to distinguish between their existing and their existing essentially, reason does contradict the conceit that valid cognition establishes objects that it understands as essentially existent.


Still, reason cannot refute what is established by the unimpaired conventional consciousnesses in the opponent's mind-stream.


Therefore, since that essentialist system and our system do not agree as to whether there is valid knowledge which understands objects to be essentially existent, we do not prove things to them with autonomous arguments; rather, we simply expose the contradictions within their own assertions.


How is this done? Let us take the example mentioned above of a syllogism based on what others accept. The reason is that the subject, the eye, does not see itself—does exist conventionally,


but essentially seeing things like blue does not exist even conventionally. Therefore, the former can be used to refute the latter.


In the case of the eye, (1) the reason, not seeing itself, and (2) the predicate of the refutation, not seeing other things, are either the same in that both exist or the same in that both do not exist. So how could either contradict the existence of the other?


Therefore, in a syllogism based on others' acceptance, the subject, predicate, and reason must be things that exist conventionally; the mere fact that an opponent claims that something exists is not adequate. The opponents themselves accept the existence of those subjects, such as the eye, as well as the reasons and examples. Therefore, why should the Mādhyamika have to prove them?


If they dishonestly disavow what they actually accept, and say, "Since they are not established for us, please establish them," that would mean that there was nothing that they were not prepared to disavow, and it would be pointless to argue with them. Who could ever help them?
- p.496 -


Objection: You claim to expose a contradiction between the opponents' assertion that the eye does not see itself and the opponents' assertion that the eye's seeing blue and such has an essential or intrinsic nature. But what kind of cognition understands the contradiction?


If it is a valid cognition that establishes the contradiction, it would have to establish it for both parties, so it would not be "based on what others accept."


If you posit the contradiction simply by its being asserted, since the opponents accept the eye's not seeing itself and its seeing something else as non-contradictory, it is untenable to use their assertion to posit a contradiction.


If you posit them as contradictory on the basis of your own assertion, this leads to utter absurdity.


How can you say to the opponents that it is incorrect for them to accept these two positions as non-contradictory because you assert that they are contradictory?


Reply: Our position does not suffer from this fault. That the eye does not see itself contradicts its having an essential or intrinsic nature; valid cognition establishes this contradiction, so it is not posited through simply being asserted.


Query: If it is possible to teach that valid cognition to the opponents and for them to then ascertain the contradiction, why is it necessary to rely on what they accept?


Reply. As the essentialist sees it, for something to be a valid cognition that proves the contradiction, it must understand an essentially existent object.


But how can we prove the contradiction by accepting for our own part something that does not exist?


Once we have shown them that there is no contradiction in positing something as a valid cognition even though its object lacks essential or intrinsic nature, this type of valid cognition will prove it to them. They will then have found the view that knows that phenomena lack intrinsic nature. So at that point there would be no need to prove to them that if something does not see itself, this contradicts its seeing essentially.


Therefore, if you want to understand Candrakirti's system, you should analyze these points in detail and penetrate their meaning.


Query: But how do you use what they accept to demonstrate the workings of the pervasion—that what does not see itself does not essentially see other things?


Reply: This should be explained as in Buddhapālita's 《Commentary on the "Fundamental Treatise"》:
For example, through association with water, earth is perceived to be wet; through association with fire, water is perceived as hot; and through association with jasmine flowers, a clothis perceived as fragrant. We see that those perceptions of earth as wet, etc. are contingent upon perceptions of wetness, heat, and fragrance in water, fire, and jasmine flowers respectively. This is something that you yourself accept. Likewise, if things had some essential or intrinsic nature, you would have to perceive that intrinsic nature in regard to the thing itself, and only then could you perceive it in some other associated thing. If you did not first perceive that in the thing itself, then, when it is associated with something else, how could you see it there? For example, if you do not perceive a bad smell in jasmine flowers, then you will not perceive a stench in cloth that has been associated with them.


Using examples that are acceptable and familiar to his opponents, he leads them to certainty about the pervasion and counter-pervasion and then applies them to the case at hand.

- p.497 -

If the eye had some seeing-essence, then its sight of itself should be noticed first, and only then would it be possible to notice that it sees such things as forms, and to notice that it sees the component of form within composite things.


But since the eye does not see itself, it does not see other things. This is how we demonstrate the pervasion using what others accept.


Also, Aryadeva's 《Four Hundred Stanzas》 states:
If the natures of all things
First appear within the things themselves,
Then why should the eye
Not also apprehend itself?


Objection: Fire burns other things, even though it does not burn itself. Likewise, it is not contradictory for the eye to see other things, even though it does not see itself.


Reply: We do not broadly refute that fire burns fuel or that the eye merely sees forms; instead we refute that the eye essentially sees other things.


As this is so, your example will have to be that fire essentially burns fuel. Consequently, your example is wrong, just as what you are trying to prove is wrong.


If fire and fuel both had essential or intrinsic nature, then they would have to be either of one nature or of different natures;


so which is it? If they are of one nature, then fire would burn itself; also, how could fire be the burning agent and the fuel the object that is burned?


Suppose that you insist that they could be. If someone were to argue that fire is the burnt object and fuel is the burning agent, what rebuttal could you give?


If they have different natures, fire could be present even without fuel, just as oxen can be present without horses.


As Aryadeva's 《Four Hundred Stanzas》 states:
Fire burns something that is hot.
If something is not hot, how can it be burning?
Therefore, there is nothing called "fuel" apart from fire,
And except for that fuel, there can be no fire.


So if you claim that burning has essential or intrinsic nature, it follows that what does not burn itself cannot burn other things. Likewise, if you claim that the eye has seeing as its intrinsic nature, you have to accept that if it does not see itself, it cannot see other things. Hence, the above faults do not budge. When one sees such critiques of the belief in intrinsic nature, one gives up tenets that conceive of the existence of essential or intrinsic nature.


One can then understand that objects and agents are tenable in the absence of intrinsic nature; hence, one distinguishes the absence of intrinsic existence from nonexistence. Consequently, one also distinguishes intrinsic existence from existence, so one knows that valid cognitions that have no intrinsic nature comprehend objects that have no intrinsic nature, and so forth.


Query: The valid cognitions that know that fire and fuel lack intrinsic nature cannot be perceptions, and thus you have to believe that they are inferences. If that is so, what reasoning are they based upon?


Reply: If things like fire and fuel have intrinsic nature, they must be either the same or different. After you have seen this, you see that a refutation of both intrinsic natures that are the same and intrinsic natures that are different must entail the absence of intrinsic nature. This fulfills the first two criteria of a correct reason.


Then, the ascertainment that there is no intrinsic nature that is the same or different fulfills the third criterion, the reason's presence in the subject. Therefore, it is a reason that fulfills the three criteria for being a correct reason capable of inducing inferential knowledge.


Based on that reason, one ascertains that fire and fuel lack intrinsic nature; this ascertainment is an inference.


In the case of the previously posited syllogism based on what others accept, you should understood that we use this same method to develop the three criteria and the corresponding inference.

- p.498 -

In reductio form, the argument works as follows: "If fire and fuel had intrinsic nature, they would have to be either the same or different," and "If they are the same, then fire would burn itself," and so forth.


This form uses something that the other party accepts as a reason, drawing implications contrary to the other party's beliefs.


This example should also allow you to understand how to construct other reductio arguments.


So as long as those opponents do not give up their essentialist tenets, they continue to believe that valid cognitions establish the referent objects of the various parts of a syllogism in reliance upon knowledge of things that essentially exist.


The moment that they know with valid cognition the essencelessness of any thing, they give up their essentialist tenets.


The Clear Words states:
Query: Still, does the refutation of the essentialists use an inference in which the subject, etc. are inferentially established for at least one of the two parties, even if not for both?


Reply: Yes. It uses areasoning that is established for the essentialists themselves, and not one that is established for the other party [i.e., the Madhyamikas], for that is what is seen in the world. In the world, sometimes who wins and who loses a dispute is decided through the testimony of a witness whom both parties accept as reliable. Sometimes, however, this may be decided using only their own testimony, without the need for someone else's testimony to determine who wins and who loses. This is true as much in the realm of philosophical reasoning as it is in the world, for it is only worldly convention that is at stake in philosophical treatises.


This gives an example and an explanation to show that reasons based on what others accept are appropriate.


The logicians claim that the three criteria and the subject, etc. must be established both for the proponent and the opponent, for they say that any valid cognition used to establish the three criteria, etc. for the opponent must also establish these for the proponent.


Candrakirti refutes that. That same text states:
Objection: One can prove or disprove a statement of something that is definite for both parties. But you cannot do this, as you claim that there is doubt as to whether the subject, etc. are established for either or both parties.


Reply: Even you who think this way should accept the method that we advocate, namely, that inference is based on the world's view of things. This means that when you use scripture to refute a certain point, you need not use only scriptures that are acceptable to both parties.


Why not? Because you can use those that only they accept. What we infer on the opponents' own terms will always be established for the opponents. This method is trustworthy, whereas attempting to establish something for both parties is not.


That is why the definitions of the logicians are superfluous, for the buddhas help disciples who do not understand reality by using what those individuals hold or accept.


So, when the reason that is used to prove the probandum is established for both parties with the kind of valid cognition explained previously, this is an autonomous [or "Svatantra"] reason.


When the reason is not established in that way and the probandum is proven using the three criteria that the other party, the opponent, accepts as being present, this constitutes the Prasañgika method. It is quite clear that this is what the master Candrakirti intended.

◎第二身生正見當隨誰行。 Which system to follow so as to develop the right philosophical view in your mind-stream.

- p.499 -

The great Madhyamikas who follow the noble father Nāgārjuna and his spiritual son Aryadeva split into two different systems: Prasañgika and Svatantrika. Which do we follow?


Here, we are followers of the Prasañgika system.


Moreover, as explained previously, we refute essential or intrinsic nature even conventionally; yet all that has been taught about cyclic existence and nirvana must be fully compatible with that refutation. Therefore, you should find certain knowledge both of how essential existence is refuted and of how cyclic existence and nirvana are still possible.


The texts of those two masters often say that you should conduct rational analysis that scrutinizes what things would be like if they were accepted as essentially or intrinsically existent. Seeing that the texts of the noble father and his spiritual son are in complete agreement on this, I accept that system. Accordingly, it is apparent that you should accept the Prāsangika position as explained above. The texts of those two masters often say that you should conduct rational analysis that scrutinizes what things would be like if they were accepted as essentially or intrinsically existent. Seeing that the texts of the noble father and his spiritual son are in complete agreement on this, I accept that system. Accordingly, it is apparent that you should accept the Prāsangika position as explained above.

- p.500 -