The Berzin Archives

The Buddhist Archives of Dr. Alexander Berzin

Switch to the Text Version of this page. Jump to main navigation.

Home > Fundamentals of Tibetan Buddhism > Level 5: Analysis of the Mind and Reality > The Basic Differences between Non-Prasangika and Prasangika Logic

The Basic Differences between Non-Prasangika and Prasangika Logic

Alexander Berzin, March 2004

Valid Ways of Correctly Ascertaining That a Certain Property-Possessor Is a Member of the Set of Phenomena Possessing a Certain Property

Correctly ascertaining that a certain property-possessor (chos-can) is a member of the set of phenomena possessing a certain property (chos) is a conceptual process requiring valid inferential cognition (rjes-dpag tshad-ma). Valid inferential cognition is a correct and decisive conceptual cognition of an obscure phenomenon (lkog-‘gyur) through reliance on a line of reasoning (rtags) as its basis.

There are three types of inferential cognition: inferential cognition based on

  • being well-known (grags-pa’i rjes-dpag),

  • the power of phenomena (dngos-stobs rjes-dpag, the power of evidence),

  • conviction (yid-ches rjes-dpag).

Through valid inferential cognition based on being well-known, we observe, through valid nonconceptual cognition, a defining characteristic (mtshan-nyid) in a property-possessor and infer that the property-possessor is a member of the set of phenomena defined and characterized by that defining characteristic. This is how we validly know that a particular piece of fruit we see in the market is an apple and apply to it the correct name.

Through valid inferential cognition based on the power of phenomena, we correctly and decisively know that a property-possessor is a member of the set of phenomena possessing that property, when that fact is slightly obscure (lkog-‘gyur), through proving it based on logical pervasions. This is how we know that the house with smoke coming out of it on the mountain over there belongs to the set of phenomena possessing a fire.

Through valid inferential cognition based on conviction, we correctly and decisively know the membership in a set when that fact is extremely obscure (shin-tu lkog-gyur). For example, how can we learn the name of a woman unknown to us whom we see in our friend’s photo album? We do not see any name written beneath the photo and we cannot figure it out by logic. The only way correctly and decisively to know her name is to ask our friend. Our friend answers “Mary.” We now validly know her name by inference based on conviction. We know that our friend has correct information and that he would not lie to us or pretend that he knows her name when he does not. Based on conviction in the truth of these two points, we infer that Mary is actually her name.

Valid Inferential Cognition of “Not-Static”

Let us look at the use of non-Prasangika Indian logic for gaining valid inferential cognition based on the power of phenomenon. For example, we can validly infer that “sound is not-static” by relying on the reason “because it is affected by causes” and the example “as is a vase.” In technical terms, what is to be proven or established (sgrub-bya) is the entire thesis (sgrub-pa) “sound is not-static, because it is affected by causes, as is a vase.”

  • Although the Tibetan word for “thesis” and for “affirmation phenomenon” is the same, they translate two different Sanskrit words. A thesis is “sadhana” in Sanskrit, while an affirmation phenomenon is “vidhi.”

  • Therefore, a thesis to be proven by logic may be either an affirmation phenomenon or a negation phenomenon (dgag-pa). In this case, as we shall see, the thesis, which contains “not-static,” is a negation phenomenon.

[See: Gelug Definitions of Affirmation and negation Phenomena.]

To prove a thesis requires reliance on three characteristics of a valid reason (tshul-gsum):

  • Applicability to the topic (phyogs-chos) – the reason (rtags) must apply to the topic (phyogs) of the thesis; it must be a property of the thesis. Here it does apply. Sound, as the topic of the thesis, is affected by causes.

  • Pervasion (rjes-khyab) – the reason must apply to the analogous set (mthun-phyogs, rang-phyogs) of all phenomena sharing the property to be proven. All phenomena that are not-static are products of causes. The analogous example of a not-static phenomenon that is affected by causes is a vase.

  • Non-pervasion (ldog-khyab) – the reason must not apply to the counter-set (mi-mthun phyogs, gzhan-phyogs) of all phenomena lacking the property to be proven. All phenomena that are not not-static (all static phenomena) are not products of causes. The counter example of a static phenomenon that is not affected by causes is space.

Based on these three characteristics of a valid reason, we validly infer and cognize the conclusion, “sound is not-static.”

The “not-staticness of sound” is a negation phenomenon. Valid conceptual cognition of it through inference required previously apprehending the counter-set “static phenomena” and precluding (cutting off, dismissing, rejecting) it. The preclusion was the third step of the line of reasoning: the reason must not apply to the counter-set.

Refutations in Non-Prasangika Indian Logic

We must be careful not to confuse a negation phenomenon with the technical term refutation (sun-‘byung-ba, Skt. dushana). Although in English, “a refutation” can mean several different things, in non-Prasangika Indian logic, “refutation” refers to the refutation of a proof.

The refutation of a proof may be based on a fault in the thesis, the reason, or the examples.

  • A fault in the thesis would be that it contradicts valid bare cognition. Any line of reasoning used to try to prove that sound is static would be refuted, because the staticness of a sound contradicts the evidence of the valid hearing of a sound.

  • Faults in the reason may be that its connection with the analogous set is unestablished, uncertain, or contradictory to the fact. Sound is not-static cannot be proven by the reason “ because it is loud” or “because it is not produced by causes.” Not all not-static phenomena are loud, and no not-static phenomenon is not produced by causes.

  • Faults in the examples may be that their connections with the analogous or counter-sets may be unestablished, uncertain, or contradictory. Sound is not-static because it is man-made cannot be established by the example, like space.

Thus, in non-Prasangika Indian logic, a refutation is not a refutation of an object to be negated (dgag-bya).

The Difference between Non-Prasangika and Prasangika Logic

Prasangika logic is different from the non-Prasangika form.

(1) Non-Prasangika logic aims to prove or establish a thesis regarding whether or not a certain property (chos) applies to a property-holder (chos-can). The property may be either an affirmation phenomenon or a negation one. By relying on a line of reasoning in which the reason has the three characteristics mentioned above, we may prove that sound is not-static, or that space is static.

Implicit in this logic is the belief that what establishes the existence (the valid knowability) of phenomena, such as the relation between a property-holder and a property is truly existent logical pervasions on the side of objects. For example, the pervasion that all nonstatic phenomena are the product of causes is established from the side of the phenomena themselves, like some findably existent fact.

(2) Prasangika does not accept that the existence of phenomena, such as the relations between property-holders and properties is established by findable, truly existent logical pervasions on the side of phenomena. It does not even accept that the existence of property-holders is established by some truly findable individual characteristics on their own sides (like solid lines around them) that make them individual validly knowable items. The issue of what does and does not exist depends solely on the validity of the mind that cognizes it.

The aim of Prasangika logic, then, is to refute ways of thinking that are false because they relate certain properties to property-holders to which they do not validly apply. Those properties may be either affirmation phenomena or negation ones. Thus, Prasangika logic aims to disprove a thesis, rather than to prove one. It does not rely on lines of reasoning containing the three factors, but relies instead on absurd conclusions (thal-‘gyur, Skt. prasanga) that would follow if the inapplicable property explicitly did apply to the inappropriate property-holder. By examining the absurd conclusions that would follow if, for instance, sound were static, we refute the thesis that sound is static. Thus, we validly cognize, inferentially, that sound is not-static.

We validly know that a conclusion is absurd if one or more of the following criteria, regarding the mind that cognizes and believes it, are satisfied. The cognition is inaccurate because it is contradicted by:

  • valid minds that share the same conventions (tha-snyad) of the terms we are using,

  • valid minds that cognize the superficial truths (kun-rdzob bden-pa, conventional truth, relative truth) of things – what they appear to be,

  • valid minds that cognize the deepest truth (don-dam bden-pa, ultimate truth) of things – what does and does not establish the existence (the valid knowability) of things.

The “not-static” derived from the conceptual preclusion of the object to be negated (“static”) through non-Prasangika logic, then, is not the same as the “not-static” derived from conceptual preclusion by Prasangika logic. This is because the methods for precluding the object to be negated are different.

  • Non-Prasangika precludes an object to be negated through asserting three logical pervasions that truly and findably exist on the sides of objects.

  • Prasangika precludes it through thorough examination of the absurd conclusions that follow if the object (the property) to be negated applied to the property-holder.

Thus,

  • Non-Prasangika preclusion of an object to be negated does not derive from a refutation based on exposing faults in logic.

  • Prasangika preclusion of an object to be negated derives from a refutation of the faults that would follow if the conclusion were true. Non-Prasangika preclusion does not derive from this type of refutation either.