Chapter 4: Example and Pseudo-example
The reason-formula states only the property of the position; the two remaining modes, positive concomitance and negative pervasion, are supplied by the example. Dignāga sets out the similar and dissimilar examples and the difference between them — the first negates while affirming something in its place, the second merely denies, which is why even a Sautrāntika who admits no eternal entities can still say “like space” — and then catalogues the pseudo-examples. The refutation runs against the Vādavidhi, the Naiyāyikas and the Vaiśeṣikas.
Condensed from Fazun’s Chinese translation-and-exposition of the Pramāṇasamuccaya and its auto-commentary, and rendered into English. Root verses are given as Fazun gives them; the prose beneath summarises his exposition. On the provenance and the limits of this rendering, see the series introduction.
Chapters: 1. Perception · 2. Inference for Oneself · 3. Inference for Others · 4. Example and Pseudo-example · 5. Apoha · 6. Sham Rebuttals
Volume Four
Chapter on the Examination of Example and Pseudo-example
Two main divisions: (1) stating one’s own position, and (2) refuting rival views. The first divides into the example and the pseudo-example.
1. Stating one’s own position
(一) The example
Though one states a reason known through its three modes (tshul gsum, trairūpya), [merely stating] its abiding well as the property of the position (pakṣadharma) [does not complete the proof]; the remaining two renowned modes — positive concomitance (rjes ‘gro, anvaya) and negative pervasion (ldog khyab, vyatireka) — the example fully displays.
The treatises on logic (因明) put only the pakṣadharma-nature into the reason-formula: “because this is produced, one knows sound is impermanent.” The other two characteristics never appear in that formula. The example is stated to supply them.
The reason’s following the probandum (sādhya), and its very absence where the probandum is absent — the similar property (chos mthun) together with its counterpart, the dissimilar — these two are to be fully shown in the [two] examples.
“Follows” means concomitance operating throughout, without exception. Take “sound is impermanent, because produced.” The similar example runs “whatever is produced by effort is seen to be impermanent, like a pot”; the dissimilar runs “whatever is eternal is seen to be not produced by effort, like space.” In both, the probandum is established by way of the probans, and both must be stated determinately: in the similar case the reason is followed by the probandum, in the dissimilar case the reason is absent wherever the probandum is absent.
In [the similar example] the negation is an implicative negation (ma yin dgag, paryudāsa); likewise [in the dissimilar example] it has the character of [bare] non-existence — a non-implicative negation (med dgag, prasajya).
The similar example negates while affirming something in its place; the dissimilar merely denies. (The Nyāyamukha puts the same distinction as “exclusion-affirmation” 遮詮 against “merely halting over-application” 止濫.) The consequence matters for cross-school debate: even an opponent who admits no eternal entities, such as the Sautrāntika who denies that space is a substance, can still use “like space” as a dissimilar example, since nothing is being asserted to exist.
Why does the text put “the reason is followed by the thesis” first, and “where the thesis is absent the reason does not exist” second, rather than “where the reason is absent the thesis does not exist”? Because stated in that order it displays the reason as determinately present in the similar class (同品定有) and wholly absent from the dissimilar class (異品遍無), with no inversion. [Objection] By parity with the pot, if one says “where the reason is present the thesis necessarily follows,” one should also say “where the thesis is present the reason necessarily follows.” Reply: that must not be stated. The pot is a particular instance of “being produced” that the thesis follows; the point of the wording is to display determination, not mutual entailment. (Fazun flags these phrases as still requiring study and as diverging greatly from Xin-hui’s translation.)
[Were one to argue] “permanent, because non-produced,” or to prove being-produced here as the effect by [the reason] disintegration — if this is stated in converse, it does not pervade and is unwished; so it goes equally [for the conjunction-form and the disjunction-form]. [Stated thus, lightning and the like would wrongly come out permanent, or person-made.]
(The Nyāyamukha verse: “One would have to prove its eternity by non-production, or establish being-produced by impermanence; if so, it would become not-what-was-to-be-stated — not pervading, not wished, equally for conjunction and disjunction.“) Invert the similar form into “all that is non-produced is eternal” and one is proving eternity by non-production, which is not the probandum. Invert the dissimilar form into “whatever is impermanent is produced” and one is proving being-produced by impermanence, again not the thesis. A reason like “produced by effort” also fails to pervade equally: lightning would come out both eternal and produced by effort.
[Objection] Granted the fault. But if the reason is understood without displaying equal pervasion, then since things other than impermanence may lie outside, one should not say “where the probandum is absent the reason is determinately non-existent”; and if the reason rests only on positive instances, the pot’s impermanence too would be “produced by effort, non-occurring-without it,” so it should not be said to be displayed by reference to the external matter. Rather, state a reason of equal pervasion: the sense alone then yields “where the probandum is absent the reason is determinately non-existent,” and no separate dissimilar example is needed. When both examples are stated together, one displays presence in the similar class; “equally pervading” already displays absence where the probandum is absent. (Fazun marks this whole stretch as requiring study, one phrase as apparently corrupt, and the closing sentence as translation-uncertain.)
Xin-hui’s much fuller rendering of the same passage turns on specification. If one specifies both “whatever is not produced by effort is determinately eternal” and “the impermanent too is only produced by effort,” the fault follows; specify only the first and there is none, since what it says is that eternity is absent among things produced by effort. Specification follows the speaker’s wish. Were the reverse route allowed, then just as one infers impermanence from “among the eternal there is no produced-by-effort,” one could infer “not produced by effort” from “among produced-by-effort there is no eternity,” thereby proving a non-probandum. Hence only the reason’s non-existence where the probandum is absent should be displayed. The parity case is “eternity is audible”: audibility being absent among impermanent things, this uncommon reason would likewise become a reason for eternity.
The reply to be stated here: the reason’s remaining characteristics, along with the pakṣadharma-nature, are displayed by the reason itself, so the sense must be “present only in the similar class.” From that restriction the negative sense follows by itself, since presence only in the similar class entails that the dissimilar class holds nothing but the absence of the similar. “Absent where that is absent” means only non-existence where the probandum is absent. It does not mean the non-existence of an entity in the dissimilar class, and it does not extend anywhere else.
Were the probandum comprehensible from the dissimilar class alone, many reasons would arise, both contradictory and non-contradictory, and the similar example still to be stated would leave the reason inconclusive. That “non-resistance” is absent among eternal things must likewise be displayed by reference to existent things, not by the rest or by contradictory non-things. So with “not produced by effort”: because produced-by-effort is bound to impermanence, the absence of impermanence displays the absence of the reason. The dissimilar example means the reason is absent in the eternal, and displaying it as non-existent there is no inversion. Even without saying “because non-produced, therefore eternal,” taking eternal things such as space as the thesis would establish “non-produced” — a thesis nobody set out to prove. Stating the counter-class in that way yields both contradiction and inconclusiveness, so it must not be stated. (Fazun: all the above seems to belong to the opponent’s side and requires further study.)
Khenchen Kunga Wangchuk’s annotation commentary (ཚད་མ་ཀུན་བཏུས་ཀྱི་མཆན་འགྲེལ་ཚིག་དོན་གསལ་བྱེད་) states the division this stretch is groping towards. A correct similar example displays the positive concomitance directly and the negative pervasion by implication (ལྡོག་ཁྱབ་ཤུགས་ལ་བསྟན་པ་); a correct dissimilar example displays the negative pervasion directly and the positive concomitance by implication. Both are therefore stated — “དེས་ན་དེ་གཉིས་ཀྱི་དཔེ་བརྗོད་གཉིས་སུ་བརྗོད་པར་ཤེས་པར་བྱའོ།” — except where stating one already produces understanding of the other. The reason both are needed is the one the next verse gives: on one example alone the common, the uncommon and the contradictory reasons would all pass as valid.
[Otherwise] the common (thun mong), the uncommon (khyad par), and the contradictory (‘gal ba) [reasons] would too [wrongly] become means of proof. Since each stated singly expresses neither [mode], one therefore states the example in two.
Neither example is valid stated singly, since each alone incurs a fault, and what results is a worldly inference-formula rather than the true reason-characteristic. State only the similar example and the reason extends even to where the probandum is absent, raising doubt: the common-inconclusive fallacy. State only the dissimilar and the reason is absent from the probandum’s own class: the uncommon-inconclusive. State neither and the reason fails to operate in the similar class while operating in the dissimilar: the contradiction fault. Hence both examples, and hence the counter-class example determinately, to remedy contradiction and inconclusiveness together.
Against an opponent for whom the matter is already largely established, stating one characteristic can serve as proof, just as a single word may carry two meanings; the second characteristic is then stated by force of the sense. The same principle governs inference for oneself: having determinately grasped the reason with respect to the thing to be inferred, one recalls that the property is present in the similar class and absent where the probandum is absent, and certainty arises.
Just as by one’s own certainty, so, wishing to generate certainty in another, [one states] the property of the position, the connection, and the probandum; the other [proof-members] to be stated — the application, the conclusion, and the rest — are set aside.
The reason-word displays the pakṣadharma-nature; the example-word displays the thing to be inferred together with the “non-occurring-without” nature; the thesis-word displays the thing to be inferred. There is no further inferential member, which sets aside the conjunction-member (合) and conclusion-member (結) that other schools hold necessary for cognition. [Objection] Then the example-word should not be a member either, since it only displays the reason’s meaning; just as the conjunction-member is not separately stated, neither should the example be. Reply: not so.
The example is not to be stated as something separate from the reason; since it displays the reason’s meaning, the comparison [it draws] (nyer ‘jal, upamāna) is not without purpose.
Within the three-characteristic reason the probandum-nature and the reason-word are already stated; the example-word, stated to display the reason’s remaining characteristics, therefore has its own work to do.
[Objection:] “Since [the example] establishes the reason’s connection, if there are the two example-statements, then make the example something separate.”
If the reason-word states only the pakṣadharma-nature, then the example must be a separate member in order to establish the reason’s other two characteristics.
[Reply:] If separate, it would become like the worldly [logicians’ account].
Made separate, it would be the worldly logicians’ example, with no connection to the reason at all.
When those [worldly logicians] state an example, they state only the similar property;
That is, “because produced,” and nothing more.
having dropped the probans that proves the probandum,
Their inference-formula contains no statement of the reason that proves the probandum.
it thus becomes a mere comparison.
Their formula does not state the example as the reason’s meaning. (A “bare inference-example” is an example with no relation to the reason; “inference” 比度 is an alternative name for the example.) It displays only the inference-probandum: “that is like a pot — because produced, impermanent.”
Nor does their dissimilar property state the probans proving the probandum. Why?
As for their dissimilar property too: serving only to make known that [space,] the reverse of the comparison, is not a fit example, that dissimilar-property formula displays merely the co-presence of the dissimilar class (mi mthun phyogs, vipakṣa). [translation uncertain]
They say no more than “in the eternal one sees the non-produced,” which displays non-production and eternity as established simultaneously. To say “where the probandum is absent the reason is determinately non-existent” would display the reason’s meaning; to say two things are found together states only their co-presence. (Fazun leaves the phrase “it is a non-affirming-negation inference” unresolved and marks it for study; he also notes that the gloss “negate the eternal example” reverses the verse-text.)
So an example stated apart from the reason’s meaning is merely a worldly example. [Objection] Even with the reason unstated, the example can still uphold the probandum. Reply: no.
When the converse is stated, [the example] is either a reason that proves the probandum or a particular of it; either way, showing it as merely similar leads to endless regress.
(Xin-hui: “both the reason and the mere-probandum, or its difference, as the example — the example would lead to infinite regress, and the dissimilar class would have no reversal.“) Three ways of positing an example, all unreasonable. First, a thesis-with-reason as example: “like a pot, which because produced is impermanent” amounts to asserting that the pot is impermanent, which itself now needs an example, and that example’s impermanence another, without end; and it still gives no certainty about sound. Second, a mere-probandum thesis as example: “impermanent like a pot, not eternal like space” merely displays being-produced as reversed from space. Third, a difference of the thesis: “impermanent like a pot” states being-produced as reversed from cloth and the like, which accords with no probandum generally and is unconnected with the thesis-meaning, since impermanence alone is the thesis. None of the three can dispel the reason’s presence in the dissimilar class. (Fazun judges Xin-hui’s rendering the better one here.)
[Objection:] “Only the property of the position is the reason [so the example must stand apart].” [Reply:] then even an erroneous [reason] would [wrongly] succeed.
Take the pakṣadharma alone as the reason and the fallacious reasons all become valid reasons.
[Objection:] “Stating both [examples], it would not [err].” [Reply:] among the nine [reason-types], in two [cases error] is still possible.
The example has [now] been explained.
[Objection] Only the similar example leaves room for error, since the reason may also occur in the dissimilar class; only the dissimilar example’s mere-absence makes the reason uncommon there; but state both and there is no fault. Reply: even both admits error. Among the nine pakṣadharma cases, the third of the first triad and the third of the last (cases 3 and 9) still err although both examples are stated.
The Tibetan commentary names the two: the last of the first triad is proving that conch-sound is produced by effort by the reason “impermanent” (མི་རྟག་པས་དུང་སྒྲ་རྩོལ་བྱུང་སྒྲུབ་པ་), and the last of the last triad is proving that sound is permanent by the reason “not tangible” (རེག་བྱ་ཅན་མ་ཡིན་པས་སྒྲ་རྟག་པར་སྒྲུབ་པ་). Both examples are stated in each, and error is still possible (འཁྲུལ་བ་སྲིད་དོ་).
(二) The pseudo-example
[The pseudo-examples (dpe ltar snang) are:] lacking the reason, lacking the probandum, lacking both; failing to establish [the reason’s] absence in the dissimilar class; the two kinds of inverted positive concomitance; and the pseudo-example that is not [genuinely] without [apparent] positive concomitance.
“No-reason” and the rest, and “no [correct] example,” the inverted positive concomitance and the rest — because they do not show the connection, by mere comparison they yield no [genuine probative] connection. These are interstitial verses.
Taking “sound is eternal, because non-tangible” as the frame: lacking the reason is “like an atom”; lacking the probandum is “like action (karma)”; lacking both is “like a pot.” Operating in the dissimilar class is “whatever is impermanent is non-tangible, as in directly-seen action, as in an atom, as in space.” A similar example, if it is to connect with the probandum-property, must state a property fully established for both parties; argued against someone who admits no eternal entities such as space, the dharma-bearer is unestablished and the case falls under “lacking both.”
Inverted concomitance comes in two forms: inverted conjunction (倒合) in the similar example, “whatever is impermanent is produced by effort, like a pot”; inverted disjunction (倒離) in the dissimilar, “whatever is not produced by effort is seen to be eternal, like space.” Stating no concomitance at all means giving reason and probandum together with their dissimilar counterparts as merely co-present: “because produced by effort, impermanent; because eternal, not produced by effort; like pot and space.” Some hold that only what possesses both conjunction and disjunction is a true example, and count the two inversions as pseudo-examples likewise. (Fazun notes 無喻 “no-example” should read 非喻 “non-example.”)
2. Refuting rival views
Three parts: against the Vādavidhi, against the Naiyāyikas, against the Vaiśeṣikas.
(一) Against the Vādavidhi
The Vādavidhi holds: “That which definitely displays the connection of those [reason and probandum] is the example — e.g. saying ‘like a pot.‘”
[Reply to the Vādavidhi:] One should not state [it as the connection] “of those [two],” since [reason and probandum] are not reciprocally probandum; nor does the mere similar property show their operating together.
[Objection (Vādavidhi):] “By [implication of] the sense, the other [the negative] is also stated.” [Reply:] then you should not [need to] state the dissimilar class [separately], for it is understood by implication.
When that invariable concomitance (med na mi ‘byung) is shown [merely] by the dissimilar class, there is nothing that fails to grasp the probandum; and since the other [connection] thereby establishes [it], even by its displaying the probans one arrives at the unwished “permanent” [— forests and lightning would then have to be permanent].
[Reason and probandum] being identical is not the [only] fault of the example — as has already been shown [in the section on “lacking both reason and probandum”].
What is it that fails to display the connection? “Whatever is produced by effort, that is impermanent.” Appending “like a pot” to it adds nothing, because what is thereby displayed is not an example, and that much alone cannot display the “non-occurring-without.”
“One should not state those,” because they are not reciprocally probans and probandum. Were the non-occurring-without mutual, then alongside “because produced by effort, impermanent” one would equally have to say “because impermanent, produced by effort.” The connection must therefore be stated one-directionally, from reason to thesis.
“Nor does the mere similar property display their co-operation.” Impermanence on its own cannot show that the reason is non-errant.
“If the sense also states the other.” Should the force of the sense in “whatever is produced by effort is impermanent” already supply the reason’s absence among eternal things, then the dissimilar class need not be stated at all; and if the probandum is known by implication already, stating the reason’s non-occurring-without with the dissimilar class serves no purpose.
If the dissimilar example’s display of the non-occurring-without does not grasp the probandum, it is useless entirely. Suppose that displaying “the eternal is not produced by effort” could by itself establish “the produced-by-effort is impermanent.” Then nothing is discarded, because what it establishes is something else; the intended proof accomplishes nothing for the actual thesis.
And the unwished “eternal” follows. If eternity and not-being-produced-by-effort displayed the non-occurring-without by the affirmative route, the dissimilar example would have to be admitted as eternal, and the two would be mutually non-occurring-without: alongside “because eternal, not produced by effort” one would have to say “because not produced by effort, eternal.” Nobody wants that, and there is no such connection between the reason and the dissimilar class in the first place.
The Vādavidhi’s own examples fare no better. It offers “like cognition-ness and pot-ness” as an example of a not-fully-established nature; but since it is merely an example stating the connection, it only fixes a sense already stated, and calling it a pseudo-example on that ground is unreasonable. It offers “whatever is eternal is non-tangible” as a pseudo-example, since it cannot display the non-occurring-without of non-tangibility with eternity. (Fazun’s gloss: in “sound is eternal, because non-tangible, like cognition, like a pot,” the cognition-example lacks the probandum-property eternity, and the pot-example lacks both eternity and non-tangibility. Where an example contradicts neither thesis nor reason, one cannot determine it to be a pseudo-example at all.)
“It is not only the single-example fault.” Unestablishedness is not the sole example-fault, as was shown earlier; inverted conjunction and disjunction are faults too.
(二) Against the Naiyāyikas
The Naiyāyikas hold: “The example that, through the probandum’s similar property, makes one comprehend that property — this is stating an example; by its contrary, it is an inverted example.”
[Reply to the Nyāya:] The example is not seen [to have the power] to make one comprehend that [property]; stating both [comprehension and example] would be purposeless, since [each] is but an alternative aspect (rnam grangs) and [the result] is established by something else. [The example] is not to be combined with the probandum-property.
[Some Naiyāyikas hold the example is that on which one relies to realise the probandum-property.] Relative to the stating of the example, to apply “so it is likewise” — the application (nyer sbyor, upanaya) — and then “so too for the probandum” — the conclusion (nigamana) — is unreasonable; for once the example is ascertained the probandum is already established, and many faults of verbal excess and deficiency attach.
The universal alone — “producedness” — is stated as the reason; the particular, depending on its own [property-bearer], follows [that basis, so the producedness of pot and of sound are not alike]. [Were “produced”] to possess the counter-class [permanence, to be negated thereby], through the particular gateway too the [same fault] would result. [translation uncertain]
Once again, if the example exists apart from the reason’s meaning, it cannot be said to be necessarily connected with that meaning, and “impermanent, because produced, like space” becomes an acceptable example; similar and dissimilar examples for the probandum could then be had from “existence-ness,” “audible-ness,” or anything else.
Take the qualifier as preceding the reason and the example takes in the probandum’s similar and dissimilar generally. “Comprehension does not see that”: an example that does not state the external matter cannot make one understand the probandum-property or penetrate it. Take the qualifier as following the reason and stating both is meaningless, since “comprehends that property” just means that by accomplishing the similar property the sense accomplishes the probandum and the property contrary to it, and there is no further comprehending besides. “Example” already means a property fully established for both parties, an example not identical with worldly conventional cognition (Xin-hui reads “the same as worldly cognition,” which Fazun prefers). And calling it “what comprehends that property” fails through difference of gateway (異門): every proof-member is only one alternative gateway for comprehending the probandum-property, so the example has no power of its own to obtain it (Xin-hui: “has no capacity to obtain that”).
One thing only should be stated: that by the probandum’s similar or dissimilar property one establishes it, for nothing else comprehends that property. Should “comprehending that property by the probandum’s similar property” be an epithet (增語) of the dharma-bearer, it becomes a dharma-bearer by those two and not by any comprehending sense. Should the example’s meaning be the accomplishing of both probandum and probans-property, then even so the binding of word-meaning cannot make one comprehend the probandum-meaning. Inferring by the probans alone would require the probandum-property to be stated without end. And the fruit of stating an example is not to establish what is already established: the fruit comes about by displaying the reason, and otherwise what the similar-property example accomplishes is no fruit at all.
Others say “comprehending that property” is the example of the example. That too fails to accord with the probandum-property, and the same holds for the dissimilar side, “because comprehension does not see that.” What establishes the pot as a similar property is comprehension of the pot’s probandum-property and of the producer–produced relation, not some other property’s being similar to the probandum. Those two phrases, taken as an example displaying the sense, ruin the verbal member of the proof.
Against the application-member (合支): “thus combining” must be either general or particular, and both fail. General fails because the universal of everything cannot be spoken of as “thus” — the universal here is “being-produced,” and stating it says only that sound has a connection; example and probandum are not one meaning, so “thus it becomes impermanent” does not combine with the external matter, on pain of an undifferentiated universal. Particular fails as well: “this produced thing, like a pot, not like cloth” makes the particular depend on its own concomitance, and sound’s production issues from something other than the pot’s. (Fazun’s analogy: the pot is produced from a clay-lump, sound from palate and teeth, each operating by its own cause; property and example therefore cannot be combined.)
Dissimilar combining, “not like space, which is not produced,” fares no better. Read the double negation (二遮) as negating only “being-produced” and it does not amount to the reason’s meaning. Read it as negating the eternity that co-operates with non-production and it still has not stated the combination, so it does not fully possess the counter-class; where the property “produced” obtains, there is no occasion for a counter-class at all, and by the similar route “being-produced” yields no comprehension of impermanence either. Then by what is the eternity possessing that counter-class to be negated?
Say that eternity is unestablished and only the dissimilar property can negate it, and the particular gateway should do the same work. Then “because audible, sound is impermanent”: pots and the like, being impermanent, co-operate with non-audibility and so negate impermanence, and the unwished eternity results. The form of establishing impermanence being the same, “because produced by effort, like lightning” goes the same way. This is unreasonable throughout (Fazun suspects “similar property” here should read “dissimilar property”). Non-production is not itself the cause of eternity, since its absence does not make a thing eternal; nor can the absence of the causes of impermanence, non-resistance and the rest, negate impermanence. As to verbal meaning, space can in principle be combined, since apart from the eternity that is not produced there is nothing else, and for impermanence one likewise says “like a pot.” Combining is unreasonable either way.
As for the conclusion-member (結), it is a restatement and not a separate member, so it should not be declared. (Fazun marks the Chinese here as corrupt and prefers Xin-hui: “The conclusion, being a restatement, is not a separate member; hence the conclusion is unreasonable.“)
(三) Against the Vaiśeṣikas
The Vaiśeṣikas hold: “That which is fully established for both [parties] is the example.”
[Reply to the Vaiśeṣikas:] Either [you must give] a formula stating [uncompounded] space as the example, or else [showing] merely one portion of the statement is not [a correct probans]. Not stating the positive concomitance, that is but the semblance of an example (dpe yi gzugs brnyan). Even taken as autonomous (rang rgyud), [the example] set up for the sake of the probans-application is purposeless; for if the example is held to be other than the reason it cannot prove the probandum, and what realises a hidden probandum is the three-mode reason alone.
If thesis and reason count as fully established with respect to space because they are its qualities, then everything whatever becomes an example with respect to space. If instead the example is “wherever the probandum-property is accomplished,” then its being only a portion of the verbal proof-member is unreasonable. Where the concomitance of reason and probandum is not displayed, what remains is the semblance of an example, a pseudo-example; the two examples are matched as set out above. And if the example is self-standing, independent in its meaning, then it must be stated as no portion of the reason-meaning, in which case there is neither probans nor combining sense.
(Fazun’s closing note: the exposition of this chapter is overly brief and diverges considerably from Xin-hui’s translation; it requires further study, and should a commentary by a Tibetan master be obtained, the text should be carefully collated.)
The collation Fazun asks for has been made against Khenchen Kunga Wangchuk’s annotation commentary ཚད་མ་ཀུན་བཏུས་ཀྱི་མཆན་འགྲེལ་ཚིག་དོན་གསལ་བྱེད་. Four results, given at their places above and gathered here:
- Each example displays one mode directly and the other by implication; both are stated unless one already produces the understanding.
- The two negations are as Fazun has them — མ་ཡིན་དགག་ in the similar example, མེད་པ་ཡི་དགག་པ་ in the dissimilar, the latter being the “non-affirming negation” he leaves unglossed.
- The two of the nine reasons that still err with both examples stated are named: “conch-sound is produced by effort, because impermanent” and “sound is permanent, because not tangible”.
- The pseudo-examples come out as three lacking (reason, probandum, both), three failing the negative pervasion against the converse thesis, two inverting the concomitance, and one lacking positive concomitance altogether — “whoever speaks is passionate, like the man on the highway” (ལམ་པོ་ཆེའི་སྐྱེས་བུ་). The annotation adds that eighteen faults of the example are taught elsewhere.
The Tibetan also confirms the reading marked [translation uncertain] above at the worldly logicians’ dissimilar formula: “like space, what is permanent is not produced” shows only that permanence and non-production are found together, not the relation of reason to probandum.
End of Chapter Four, the Chapter on the Examination of Example and Pseudo-example.
← 3. Inference for Others | All chapters | 5. Exclusion of the Other (Apoha) →