Why the Definitions of "Sapaksa" and "Vipaksa" in Hetuvidyā Do Not Include "Excluding the Dharmin"

Gang Xiao

Broadly speaking, the core canonical sources for Han-transmitted Buddhist logic (Hetuvidyā) are just two: the Nyāyapraveśa (因明入正理论) and the Nyāyamukha (因明正理门论). Neither of these works contains the explicit phrase "excluding the dharmin" (除宗有法). In the section of the Nyāyamukha that expounds the three aspects of the reason (因三相), there is the following passage:

If the mark is verified with respect to the subject of inference; it is certainly present in all other instances of the same class; and it is universally absent wherever that class is absent. This is what gives rise to a definite conclusion. [1]

I have three commentaries on the Nyāyamukha at hand. The first is the second volume of Mr. Shen Jianying’s Buddhist Logic (《佛家逻辑》), subtitled “Translation and Explanation of the Nyāyamukha” (《〈因明正理门论〉译解》). The second is Mr. Zheng Weihong’s Direct Explanation of the Nyāyamukha (《〈因明正理门论〉直解》), whose first section offers a line-by-line explanation and the second presents scholarly research. The third is my own Two Treatises of Han-Transmitted Hetuvidyā (《汉传因明二论》), whose second half is a commentary on the Nyāyamukha. These three commentaries interpret this Nyāyamukha passage in nearly identical ways, particularly the lines “among other things of the same class it is definitely present here; where that is absent, this is universally absent” (于余同类今此定有；于彼无处念此遍无). Mr. Shen Jianying states:

“Among other things of the same class” (于余同类): this refers to the sapaksa, with the dharmin excluded. [2]

Mr. Zheng Weihong writes:

“Among other things of the same class this is definitely present” (于余同类念此定有): “among other” (于余) means apart from the dharmin. “Of the same class” (同类) refers to the sapaksa, meaning objects whose nature matches the property to be proven (所立法, the predicate). The Great Commentary (《大疏》) glosses this line as: “This is the presence of the reason in the sapaksa” (同品定有性也) (vol. 4, p. 18r). This also confirms that the sapaksa referenced in the phrase “presence of the reason in the sapaksa” excludes the dharmin. “Where that is absent, this is universally absent” (于彼无处念此遍无): The Great Commentary says, “This is the universal absence of the reason in the vipaksa” (异品遍无性也) (vol. 4, p. 18r). “That” (彼) refers to the property to be proven. “Where that is absent” (彼无处) denotes the vipaksa: all objects other than the dharmin that lack the property to be proven. The statements of the second and third aspects of the reason both make clear that sapaksa and vipaksa exclude the dharmin. Works that criticize Dignāga for an inconsistency between his definitions of sapaksa and vipaksa (which omit the exclusion of the dharmin) and the Nyāyamukha’s system are therefore one-sided and unfounded. [3]

I note:

“Among other things of the same class this is definitely present” (于余同类念此定有) — this line describes the presence of the reason in the sapaksa. The character “other” (余) means “the remainder” or “apart from,” which makes clear that the sapaksa must exclude the dharmin. Anyone who claims that Buddhist Hetuvidyā never explicitly proposed excluding the dharmin is mistaken. [4]

We all take this as the basis for excluding the dharmin. Among Tang dynasty commentaries, the one that explicitly states “exclude the dharmin” is Venerable Master Wen Gui’s Zhuangyan Commentary (《庄严疏》), which says:

“Equal-meaning class” (均等义品) refers to all dharmins other than the thesis subject (宗有法), which are collectively called meaning-classes… [5] All dharmins other than the thesis subject are called loci… [6]

Venerable Master Huizhao’s Essentials of the Commentary on the Nyāyapraveśa (《因明入正理论义纂要》) records the following:

Question: What is meant by “the presence of the reason in the sapaksa” (同品定有性)? What dharmas count as sapaksa? Where does the reason apply? Preliminary answer to the first question: There are several early interpretations. One holds that the substance of a pot and similar things is called sapaksa, because a pot shares the class-character of impermanence with sound, and so is called sapaksa. Another holds that aside from sound, any class (品) that possesses impermanence counts as sapaksa; this aligns with the first interpretation. A third holds that if one is proving “sound is impermanent,” then all impermanent things apart from sound constitute the sapaksa. A fourth holds… [7]

Other works also explicitly state that the dharmin must be excluded. For example, the Ruiyuan Ji (《因明论疏瑞源记》) records:

Master Xuanying’s Nyāyamukha Commentary states that for the line “the locus of meaning equal to the property to be established is called sapaksa” (与所立法均等义品说名同品), there are four main schools of interpretation:

1. The view of Venerable Wen Gui of the Zhuangyan monastery: all dharmins other than the thesis subject are collectively called meaning-classes…
2. The view of Master Bi of Bianzhou: all distinctions other than the thesis subject are called meaning-classes…
3. One interpretation holds that, apart from the thesis subject, both the dharmin and the predicate (能别), together with the meaning-class equal to the property established by the thesis, count as sapaksa…
4. The view of Master Ji (Kuiji) and others: the inseparable relation between property and dharmin, apart from the thesis subject, is the sapaksa of the thesis. [8]

There is also the Rudimentary Survey of the Great Commentary on Hetuvidyā (《因明大疏蠡测》), whose table of contents explicitly includes the heading “Sapaksa and Vipaksa Exclude the Dharmin” (宗同异品除宗有法). The text says:

Although the treatise never explicitly states that the sapaksa excludes the dharmin, examining its full context makes the principle clearly evident. The treatise states: “The term ‘non-shared’ refers to cases like asserting that sound is permanent because it is audible. Both the class of permanent things and the class of impermanent things lack this reason.” (see Great Commentary vol. 6, p. 13 left). The non-shared non-determinative — that is, the fifth case among the nine cases of the reason — has no sapaksa and no vipaksa. Permanent things constitute the sapaksa… It can be seen that what is called sapaksa encompasses only permanent classes apart from sound, and does not include the thesis subject’s dharmin… Sapaksa excluding the thesis subject is therefore a universal principle of Hetuvidyā… I would argue that, in line with the principles of Hetuvidyā, excluding the thesis subject from the sapaksa has strong and well-founded justification… [9]

Mr. Zhou Shujia’s explanation matches the account in the Ruiyuan Ji exactly, stating:

In the Tang dynasty, there were four distinct schools of interpretation for this term “same” (同):

1. Master Wen Gui held that all dharmins apart from the thesis subject are collectively called meaning-classes; if a meaning-class is equal in import to the property established by the thesis, it is called sapaksa.
2. Master Bi of Bianzhou held that all distinctions apart from the thesis subject are called meaning-classes; if a meaning-class is equal and similar to what the thesis establishes, it is called sapaksa.
3. One interpretation holds that, apart from the thesis subject, both the dharmin and the predicate, together with the meaning-class equal to the thesis’s established property, count as sapaksa.
4. Master Kuiji and others held that the inseparable relation between property and dharmin, apart from the thesis subject, is the sapaksa of the thesis. The fourth interpretation is correct, as Dignāga takes the inseparability of property and dharmin as the defining nature of the thesis. [10]

Mr. Zheng Weihong states in Chapter 3 of the “Lower Treatise” section of his Direct Explanation of the Nyāyamukha:

While the Nyāyamukha does not explicitly state that sapaksa and vipaksa exclude the dharmin, this is self-evident. [11]

He returns to this topic in Section 3 of Chapter 3 of his General Treatise on Buddhist Logic (《佛家逻辑通论》), devoting two and a half pages to the argument that sapaksa and vipaksa exclude the dharmin. His discussion is split into two parts: first, the nine cases of the reason (九句因) require that sapaksa and vipaksa exclude the dharmin; second, this exclusion is an implicit principle of Dignāga’s new Hetuvidyā. I will not reproduce this lengthy section here to avoid unnecessary bulk [12]. In Section 3 of Chapter 4, “The Three Conditions of a Valid Reason,” the first condition is that “sapaksa and vipaksa exclude the dharmin,” a point he elaborates on for half a page. This topic appears in many other places in the book as well, all with substantial discussion. All these accounts show that scholars of Hetuvidyā, both past and present, have almost universally accepted that the dharmin is to be excluded.

For the past while, I have been studying the Pramāṇasamuccaya (《集量论》). I have three translations of the text at hand: the first is Master Fazun’s Brief Explanation of the Pramāṇasamuccaya (《集量论略解》), published by China Social Sciences Press in March 1982; the second is Mr. Lü Cheng’s Abridged Commentary on the Pramāṇasamuccaya (《集量论释略抄》), included in volume 9 of the Supplement to the Tripiṭaka (《大藏经补编》) and also collected in volume 1 of Selected Buddhist Studies of Lü Cheng (《吕澂佛学论著选集》), published by Qilu Press; the third is Mr. Han Jingqing’s edition, which has unfortunately never been published publicly and exists only as a handwritten manuscript. Each of these three versions has its own merits and limitations.

Let us now compare how these three core texts — the Nyāyapraveśa, the Nyāyamukha, and the Pramāṇasamuccaya — define sapaksa and vipaksa. First, the definition given in the Nyāyapraveśa:

What are sapaksa and vipaksa? The meaning-class equal to the property to be proven is called sapaksa. For example, if one proves impermanence, a pot and other impermanent things are sapaksa. Vipaksa is the class in which the property to be proven is absent. If there exists something that is permanent and unconditionally arisen, such as empty space, and so on. [13]

The definition of sapaksa and vipaksa in the Nyāyamukha:

If a class is adjacent and equal to the property to be established, it is called sapaksa. This is because all meanings are termed classes. If the property to be established is absent, it is called vipaksa. [14]

In Master Fazun’s translation of the Pramāṇasamuccaya:

By virtue of being equal in class to the general characteristic of the property to be established, it is called sapaksa. [15] If the sapaksa is absent, it is called vipaksa. [16]

In Mr. Lü Cheng’s translation of the Pramāṇasamuccaya:

Here, taking all meanings as classes, that which is similar by sharing the common characteristic of the property to be established is sapaksa. [17] Likewise, that in which no sapaksa is present is vipaksa. [18]

Dignāga’s successor, Dharmakīrti, was another leading figure in Buddhist Hetuvidyā. He defined sapaksa and vipaksa in his Nyāyabindu (《正理滴点论》) as:

Sapaksa: that which shares the same meaning as the general category of the property to be established. Vipaksa: that which is not sapaksa. [19]

None of these treatises explicitly includes a clause excluding the dharmin in their definitions of sapaksa and vipaksa. Mr. Wu Shoukang, noting this omission in the canonical sources, argued that it contradicts the theory of the nine cases of the reason (九句因) and the three aspects of the reason (因三相) in Hetuvidyā [20]. Mr. Zheng Weihong offered a detailed rebuttal to this claim in Section 5 of Chapter 6 of his General Treatise on Buddhist Logic.

The three aspects of the reason (因三相) are core concepts in Buddhist Hetuvidyā, and their theoretical foundation is the nine cases of the reason (九句因). The Nyāyamukha presents the nine cases as follows:

In cases where the reason is present in all sapaksa: the reason may be present in the vipaksa, absent from it, or present in some vipaksa and absent in others. In cases where the reason is absent from the sapaksa, or absent from both sapaksa and vipaksa, there are similarly three variations. If the thesis is that something is impermanent, there is no vipaksa at all. For someone who does not accept the existence of empty space and similar entities, how can one claim the reason is absent there? If the reason is absent there, it does not apply there, and there is no doubt whatsoever, so no fault arises. This yields nine distinct combinations of thesis and reason. Their characteristics are outlined in order below:

1. To assert sound is permanent because it is measurable.
2. To assert sound is impermanent because it is produced.
3. To assert effort-produced phenomena are impermanent because they arise from continuous exertion.
4. To assert something is permanent because it is produced.
5. To assert something is permanent because it is audible.
6. To assert something is permanent because it arises from continuous exertion.
7. To assert non-effort-produced phenomena are impermanent because they do not arise from continuous exertion.
8. To assert sound is impermanent because it arises from continuous exertion.
9. To assert something is permanent because it is intangible. [21]

The fifth of these nine cases is one where the reason is absent from both the sapaksa and the vipaksa. Dignāga’s example for this case is:

Thesis: Sound is permanent. Reason: It is audible.

Under this formulation, following the standard definitions of sapaksa and vipaksa, things like empty space would count as sapaksa, while things like pots and basins would count as vipaksa. Neither the sapaksa (e.g., empty space) nor the vipaksa (e.g., pots and basins) possesses the reason “audibility.” From Dignāga’s exposition here, it is clear that he has excluded the dharmin “sound” from both sapaksa and vipaksa.

The Nyāyamukha also states:

The dharma of the thesis (宗法) is taken only to be what both the proponent and opponent are mutually and definitely agreed upon. The same standard applies to whether the reason is present or absent in the sapaksa. [22]

This line makes clear that the sapaksa must consist only of mutually accepted, established dharmas (共许极成) [23]. Take the thesis “sound is impermanent” as an example: the proponent and opponent are actively debating whether “sound” is in fact impermanent, so this property cannot be mutually accepted. “Sound” is therefore naturally excluded from both the sapaksa and the vipaksa.

Earlier, I cited the Nyāyamukha passage — “If the mark is verified with respect to the subject of inference; it is certainly present in all other instances of the same class; and it is universally absent wherever that class is absent. This is what gives rise to a definite conclusion” — as evidence for “excluding the dharmin.” Strictly speaking, however, this passage only supports excluding the dharmin from the sapaksa. As for excluding the dharmin from the vipaksa, Mr. Shen Jianying writes in his Answers to Questions on the Three Aspects of the Reason (《因三相答疑》):

The dharmin and the vipaksa of a proposition belong to two contradictory classes of things, and are not part of the same set to begin with — so what question is there of excluding or not excluding it? To argue that the vipaksa must also “exclude the dharmin,” and to claim this is mandated by “Dignāga’s Hetuvidyā system,” is wishful thinking. Dignāga never stated that the vipaksa must “exclude the dharmin.” What is more, claiming both sapaksa and vipaksa must “exclude the dharmin” leads to a paradox. I recall that Yao Nanqiang of East China Normal University put forward a sharp insight on this at a discussion: if the dharmin is to be excluded from the logical extensions of both sapaksa and vipaksa at the same time, where does the dharmin fit? In fact, if it is “excluded” from the extension of the sapaksa, it must fall into the extension of the vipaksa, and vice versa. Excluding the dharmin from both sapaksa and vipaksa is a paradox of logical extension, and cannot be resolved within the same linguistic framework. Since this is just a version of the Barber Paradox, the inference, based on the law of identity, that the vipaksa in the nine cases of the reason and the three aspects of the reason must also “exclude the dharmin” does not hold. I find this analysis compelling. [24]

I have reservations about the view of Mr. Shen Jianying and Mr. Yao Nanqiang, grounded in the principle of mutual acceptance (共许极成). Since all dharmas apart from the thesis subject (宗体) must be mutually accepted, the vipaksa must also be mutually accepted by definition. If the dharmin were not excluded, it would not be mutually accepted. This also touches on the question of how to characterize Hetuvidyā. In the second issue of 2003 of World Religious Studies (《世界宗教研究》), Mr. Zeng Xiangyun published an article titled “Hetuvidyā: A Theory of Buddhist Dialogue” (《因明：佛家对话理论》), arguing that Hetuvidyā is not fundamentally a strict logical system, and that it may be inappropriate to judge it by the standards of formal logic. Mr. Zeng Xiangyun writes in the article that “to conduct research in Hetuvidyā, one must… move beyond the narrow stance of logicalism.” Of course, Mr. Zeng Xiangyun’s article still has points open to debate, which I will not go into here.

Returning to the definitions of sapaksa and vipaksa in the core canonical texts — the Nyāyapraveśa, Nyāyamukha, and Pramāṇasamuccaya — we can see that there is no need to explicitly write “exclude the dharmin” into the definitions, as sapaksa and related categories are already required to be mutually accepted. Including the exclusion explicitly would be redundant, not helpful. As for why Dignāga excluded the dharmin in his explanatory prose but did not state this exclusion explicitly in his definitions, Mr. Zheng Weihong offers the following explanation:

For Dignāga's contemporaries, excluding the thesis subject from similar and dissimilar instances was probably as obvious as daylight—there was no need to spell it out, and doing so would have been redundant. So why do the four Tang dynasty commentators all explicitly state "excluding the thesis subject" in their definitions? It comes down to different cultural traditions. Hetuvidya is Indian logic. In earlier Hetuvidya, the positive and negative examples could not be the thesis subject, otherwise you'd run into circular reasoning. When Hetuvidya arrived in China during the Tang, it was an unfamiliar, brand-new field of study. When transmitting a new discipline, it's perfectly natural to spell out principles that were self-evident to Indian logicians but hidden from newcomers [gx25]. [25]

Now, when Mr. Zheng Weihong uses the word "probably," the element of conjecture becomes rather pronounced (Mr. Shen Jianying's "Answers to Questions on the Three Characteristics of the Reason," cited earlier, also uses "probably"). But be that as it may, similar and dissimilar instances do indeed exclude the thesis subject. There's no question about that. In Mr. Lü Cheng's Excerpted Annotations on the Commentary on the Pramāṇasamuccaya, where he discusses the three characteristics of the reason, we find this passage:

Furthermore, among things of the same class as what is to be inferred (the similar instance), one sees the property in question either wholly or partially through a generic character. Why is this not definite? Because it is said that the property is present only in the same class, not that the same class is the only thing possessing it. To seek further certainty, one says it is absent wherever the property is absent. "Where the property is absent" (the dissimilar instance) means everything other than what is to be inferred, and is not the contradictory of what is to be inferred. [26]

Fa Zun's translation is similar in content, but the wording is genuinely awkward, so I won't cite it here. In this passage, "that property" (彼法) refers to the reason property. The meaning is: within similar instances, the nature of the reason property is either fully present or partially present. Why not state definitively whether it's complete or partial? Because Hetuvidya requires "presence only in the same class" (唯同类有), not "the same class as the only possessor" (同类唯有)...

Let's look at the difference between "唯同类有" and "同类唯有." If it were "the same class as the only possessor," that would be equivalent to "partial presence in similar instances." But "presence only in the same class" is different. As we know, the domain of discourse in Hetuvidya exhaustively covers similar instances and dissimilar instances—no third possibility can arise. So "presence only in the same class" actually becomes equivalent to "total absence in dissimilar instances." In Fa Zun's translation of the Brief Explanation of the Pramāṇasamuccaya, volume three, when discussing the three characteristics of the reason, it says: "To say that the dissimilar instance is merely without [the reason] is to say that the similar instance is not without it" [27]—making the latter two characteristics equivalent. In other words, the latter two of the three characteristics of the reason are really one and the same.

But we should note: even though the latter two characteristics amount to the same thing, in practice Master Dignāga requires both to appear. In his Brief Explanation of the Pramāṇasamuccaya, explaining the two verses "It is not asserted that there is [the reason] only in the similar instance; from one [reason] one sees many meanings," he says: "'It is not asserted that there is [the reason] only in the similar instance.' If it were specified that the reason occurs only in the similar instance, then it would apply only to the similar instance. Though this avoids the first two faults, it would mean that the reason arises only when the established similar instance exists, does not arise when it doesn't, and is absent everywhere else. 'From one [reason] one sees many meanings.' Yet from a single reason one also sees many things that could not arise without it. For instance, 'being produced' is what prevents impermanence, non-self, and so on from failing to arise. Therefore, all three kinds of specification have faults" [28]. Read alongside the surrounding context of the Pramāṇasamuccaya, these three specifications become clear: "the dissimilar instance is merely without [the reason]," "only the dissimilar instance is without [it]," and "only the similar instance has [it]."

With Master Dharmakīrti, both don't need to appear—he uses either the positive method or the negative method, and either one suffices. This is the difference between Master Dignāga and Master Dharmakīrti. The equivalence of the latter two characteristics in Dignāga's system means the example proposition absolutely cannot be an exclusive proposition, yet similar and dissimilar instances do in fact exclude the thesis subject. This is determined by the religious sensibility underlying Hetuvidya—

to avoid arguing over utterly meaningless questions... Therefore Hetuvidya is always constrained by facts. It requires that everything invoked in a syllogism must not depart from what is factual—not only must the inference not be "wrong," it must also be "true" [29].

When Master Dharmakīrti catalogued the faults of reasons, he didn't include the "uncommon non-determinative" (不共不定) that Master Dignāga had established as a fault. This omission suggests that the fifth of the nine modes of the reason—"absent from the similar instance, absent from the dissimilar instance"—escaped Master Dharmakīrti's notice. Put in modern terms, this means that in Dharmakīrti's system, similar and dissimilar instances don't exclude the thesis subject. For instance, Mr. Zheng Weihong, on page 279 of his Direct Explanation of the Nyāyapraveśa, writes:

The Nyāyabindu nowhere specifically states that similar and dissimilar instances exclude the thesis subject. But from the fact that among faults of reasons it lists only the unestablished (不成), the non-determinative (不定), and the contradictory (相违)—omitting the uncommon non-determinative reason—and that after explaining the fault of the unestablished reason from failure to satisfy the first characteristic, it proceeds directly to the non-determinative fault from failure to satisfy the third characteristic, without discussing the fallacious reason that fails the second characteristic, namely the uncommon non-determinative—one can see that for Dharmakīrti, similar and dissimilar instances do not exclude the thesis subject.

In section five of chapter ten of his Comprehensive Treatise on Buddhist Logic, "Dharmakīrti's Contributions to Buddhist Logic," he writes:

After the Nyāyabindu finishes discussing the unestablished reason that violates the first characteristic, it goes straight to the non-determinative fault that violates the third characteristic. It makes no mention at all of the uncommon non-determinative, which is the fifth of Dignāga's nine modes of the reason. This shows that Dharmakīrti eliminated the fifth case... [30] By eliminating the fifth case and downgrading the role of the positive example, Dharmakīrti effectively shows that similar and dissimilar instances do not exclude the thesis subject... [31]

There are other ways to read Dharmakīrti's omission of the "uncommon non-determinative." Let's consider what Mr. Ju Zonglin has to say:

Dharmakīrti emphasized the consistency of language and thought, and the consistency of inference for others with inference for oneself. He held that "uncommon non-determinative" situations cannot occur in thought, because inferential thinking always works through comparison of similarities and differences. If you leave aside such comparison and consider only whether a single property belongs to the subject, you'll never reach a determinate result—there's no complete inferential process. Since "uncommon non-determinative" doesn't arise in normal thought, there's no need to affirm it linguistically. That's why Dharmakīrti didn't list it among his non-determinative reasons. The difference between Dharmakīrti and Dignāga on this point doesn't stem from the same standpoint. Dharmakīrti emphasized practical application and discarded what doesn't occur in normal use. Dignāga emphasized systematic completeness: the "uncommon non-determinative" does exist among non-determinative reasons, and when this special case arises, it should be recognized as a type of non-determinative, one entirely distinct from the others. [32]

Admittedly, Mr. Ju Zonglin's Historical Outline of Tibetan Buddhist Hetuvidya reads somewhat like a personal meditation—it's more of an introductory work, and its format isn't as rigorous as a typical academic monograph. But given Mr. Ju's seventeen years in Tibet, I cite his views here as representative of how Tibetan Hetuvidya scholars generally understand Dharmakīrti's omission of the "uncommon non-determinative" fault.

What Mr. Ju's passage tells us is that Dharmakīrti's omission shouldn't be read as abolishing the exclusion of the thesis subject, but rather as a matter of practicality. This is the gap between theory and fact: cases that are theoretically correct but never occur in practice are countless. Dignāga pursued completeness of the theoretical system; Dharmakīrti paid closer attention than Dignāga to whether things actually exist in fact.

In section four of chapter ten of his Comprehensive Treatise on Buddhist Logic, "Inference for Others," Mr. Zheng Weihong writes:

The existence of this fault [the author's note: referring to the uncommon non-determinative] marks the independence of "definite presence in similar instances" from the third characteristic, "total absence in dissimilar instances." But Dharmakīrti didn't mention this fault, which shows that he eliminated the uncommon non-determinative reason, and further indicates that his second characteristic differs from Dignāga's—Dharmakīrti's latter two characteristics can be derived from each other. [33]

In Master Dharmakīrti's Nyāyabindu, there's a passage stating that the latter two characteristics are in fact one and the same:

Inference for others: a sign for a person possessing the three characteristics. Because the cause is provisionally established as the effect. By virtue of differences in construction, there are two types: first, possessing the nature of agreement in presence; second, possessing the nature of agreement in absence. Apart from the differences in their construction, these two are in reality without distinction. [34]

This is Dharmakīrti's own text. As we noted earlier, Dignāga states in the Pramāṇasamuccaya "presence only in the same class" (in Fa Zun's translation, "presence only in the similar instance"), which is completely equivalent to "total absence in dissimilar instances." So we can say that the mutual derivability of the latter two characteristics in Dharmakīrti comes directly from Dignāga's "presence only in the same class."

Regarding the claim that in later Hetuvidya the latter two characteristics can be derived from each other, I asked Teacher Han about the question of commonly accepted premises (共许极成) in Tibetan Hetuvidya, and he told me plainly that yes, this is affirmed. Dharmakīrti's Nyāyabindu is divided into chapters on perception, inference for oneself, and inference for others; his Pramāṇavārttika into chapters on inference for one's own purpose, establishing valid cognition, perception, and the purpose of others. Since there are chapters on inference for others, there must of course be commonly accepted premises—which include: shared acceptance of the components of the thesis; shared acceptance of the concept of the reason; shared acceptance of the universal concomitance of the reason with the thesis subject; shared acceptance of similar and dissimilar instances; shared acceptance of presence or absence in similar instances and presence or absence in dissimilar instances; shared acceptance of the example proposition; and shared acceptance of the example instance [35][gx35]. Since shared acceptance is required, whatever is under dispute is by definition not commonly accepted—so neither Dharmakīrti's nor Dignāga's propositions can contain the thesis subject.

As for using logical methods to study Hetuvidya, that's hardly my strength. And from the standpoint of religious sensibility (which I discuss on pages 5–11 of Two Treatises on Chinese Hetuvidya), it's simply inadmissible: in a Buddhist Hetuvidya syllogism, the thesis stated first is directly realized and immovable, whereas the major premise that comes first in a logical syllo