為開示他故，有說非自見，他見亦能立。為斷彼執故，

而說自見言。觀察而取故，於所比度境，非許言為量。

彼比量不轉，為違害教故，說成立餘故，爾時彼非量。

以彼所成立，非成立無餘。若具彼教成，是誰人之教？

正量所妨害，彼教如何成？若受彼相違，如何以彼成？

除彼若受餘，彼捨支，非量。自解由何生，何不說能立？

何理持彼教，此亦何不持？先，由何能立，凡常能了解，

生根諸能立？知能修彼義，如知復能行。若斷絕繫屬，

總亦非行境，彼義無少許，觀能立所立。士意樂增上，

立真非真者，應無因、似因，彼非依事故。有義觀待知，

無知非能立。縱有，不繫事，解無解相同。自性及果法，

或現無見因。彼繫若事成，何故自不成？餘人由餘解，

亦不應道理。以他所妄計，成立其過失，二者繫屬故，

一無破餘故。言義彼諸聲，分別增益體，為成其非因，

由義成義故。分別及諸教，唯屬作者欲，除事，有他故，

假立是錯亂。由義解義故，說宗及說因，於義無功能。

故彼等自己，非住於能立。宗言若能示，說者之意樂，

是量彼生疑，故非親能立。由宣說所立，傳亦無功能。

說因自無能，而能示功能。即由因義境，而說彼無能。

若彼亦有能，因言轉故者。彼由疑，欲知，依止彼分位。

受許異品者，同故，應無窮。內支分功能，正住於三相。

於彼生憶念，正住於彼言。若謂未示境，因無轉趨故，

由顯示境故，許功能成者。先遍雖無彼，然說聲所作，

如此皆無常，義生彼壞覺。雖未顯示宗，成立無障故。

三中隨一相，未說即不全。說所立為宗，彼說亦非過。

由為說能立，似因非過故。雖無簡別說，一類則生疑，

餘凡說所立，皆應成為宗。說成，能立故，餘亦無過者。

彼顯示所立，豈是能立支？列擧及似等，餘無繫堪笑。

謂彼非正理，此如前已說。雖由說能立，是所了解義，

然為無愚故，顯示所立相。體、介、自、樂言，說彼相有四。

未成，非能立，義說立者許。雖未說樂遍，所立，如我用。

雖他欲皆遮，疏解說自我，為遮所疑處。此為彼義作。

此觀待而說，法、有法差別。雖未說所樂，由時而了解。

喻無隨行過。如說我是他，彼於喻不成。彼妨害所樂。

若為諍何事，列擧其能立，彼若非所立，所立為何等？

若是其他者，不樂，或無果。此於有二式，顯示無隨行，

及有諸相違。假若說以總，許為所立者。彼無餘義故，

身不成，不成。諸說所詮空，是無知所惑。計聲瓶差別，

壞滅亦同者。非，由成壞滅，聲如是成故。如是有餘義，

瓶亦具足彼。若由聲差別，無常不隨行。遮不繫故，非。

不具及具餘，其遮有二種，僅遮不具故，非無隨他轉。

彼總是所立，故非立已成。彼有法差別，無不隨行過。

此顯諸破除，有法差別法，法差別有法，是除遣。有法，

不如是立故。於有法除遣，諸聚，非相違。非樂如是立，

唯法是所立。雖於一有法，論立多種法，自樂謂所立，

故說隨自聲。若謂許論故，則受許一切，若妨害一義，

因宗皆有過。成立聲無常，遮香地德故，應成相違因。

非時故，非者。彼餘亦相同。此有法是時。彼亦害論義。

若說由能立，有法之法等，立者欲樂者。有時成所樂，

由彼亦了解。強謂汝樂此，顯是自在行。說非有果因，

由錯亂妨害。非是無不生，雖妨害彼義，於餘何妨害？

比量自行境，說不待於教。彼成彼善成，爾時不待論。

爾時捨所諍。無受故非捨。此是受方便，雖有亦非支。

如是二境淨，彼是持論時。爾時論能害。彼成義與彼，

不可思違故。轉於第三處，受論則應理。彼與所立法，

違害無不生，應捨。非捨餘，犯無窮過故。謂一切觀察，

須取諸論者，此派由何作？諸無宗派者，由烟不解火。

其人單獨生，未見諸德失，此宗派惡鬼，嗚呼誰所差？

若謂於同喻，一切論顯示，遍一能立性，是為能立者。

此規超世間。無繫法不成，如何不成立？彼成，說能立，

如何因有過？有法無餘法，喻唯無全取。由語及烟等，

比有心、火等。不錯亂，能立，自性及因義，由害有所說，

非遮於自性。彼樂所立義，了解無不生，餘人由彼因，

如何非能解？隨說不說因，若相違，立者，於此何所作？

非由彼說過，彼生妨害論。說害故過失。豈若不宣說，

彼便無害耶？非作不應理，云何有過失？似因於餘人，

能立自樂故，說無義成過。成彼故，此非。若謂由有論，

某事非理故，破餘說理者，是剎帝利規。諸義皆同己，

非能說能立，故諸能所立，建立皆失壞。於一有法上，

相違不可故。於相違決定，妨害，此非有，如是相違過。

若謂彼雖非，能害與所害，由害論義故，無繫亦害者。

一切害一切。若謂有繫屬，故當害彼者。不爾，一切因，

皆於自所立，觀察德失故。無所立不生。彼繫，此不解。

此唯妨害論，是過。餘亦同。若謂許論故，論見盡所立。

說因喻不成，亦應成為宗。說能立故非，說所樂故者。

所立以理得，非由言而遮。不樂非所立。立者所不樂，

餘法非所立。爾時非所立，妨害何相違？是宗相外義，

自聲亦無義。為欲趣論者，此疑由何起？彼，量無破者，

受持何能遮？若量能破者，由言何能入？彼等先於宗，

由自樂而持，彼後何不得，隨意而持餘？樂無害，非立，

顯示無窮故。除邪分別故，此作隨自聲。餘說為捨棄，

教所作差別，所共許有法，故說隨自聲。由觀察時故，

共許依已成。自欲立差別，諸事無諍故。若由說已成，

有法非所立。由說自體性，此了解彼義。說已成，能立。

若是有疑者，說顯所立性，此則具果利。由說正比量，

是以總為境。此中縱不說，宗無少相違。有法為所立，

彼有何過失？彼無能。何故？因無隨行故。豈非彼等過？

觀待於後支，不許為宗過。若爾，因等失，亦應成宗過。

皆妨害宗故。故唯彼繫屬，許宗過。非餘。如違現量等。

不觀待遣除，因等相所遣。斷不遍及返，故宣說宗相。

隨自、唯、性聲，及不被遣等，能除其相返。樂聲除不遍。

許所立宗相。彼等無宗相，遣則有妨害，餘相不轉故。

說隨自、所樂，雖解指定義，作邊相屬故，遮餘時故說。

樂緣此非支，成所樂言等，是支。即由彼，破不成因等。

非詮故非理，故「隨自」自樂，不待一切論，彼樂即所立。

故不樂相屬，所樂亦能害。如立無害故，宗因皆無過。

彼所許亦是，無諸量能破。因對疑說故，遣非因所依。

由比量差別，說四種能害。所許是果因，眾稱是性因。

欲說正決定，自他信同故，將自語與論，合說為一種。

如自若非量，言語則不轉。如是不依論，彼稱義觀察，

亦說相同故。彼論時所依，則有所妨害，為顯此義故，

此自語同說。故於此諸喻，亦顯示相同。若無有諸量，

論語不可故。自語違明喻，於教示方隅。彼說法未來，

不感樂是喻。不依彼說事，縱違論、非害。二者諸功能，

相等故，能碍。喻如自語違。彼亦自語性。彼中何有量，

彼於餘能害？非爾，離理宗，功能相等宗，如何能妨害，

如實義比量？諸教為量者，如前已廣遮。是故受許已，

觀察許此過。妨害故。受許，非離比量外，顯境差別故，

而作為別異。非爾，應太過，或別說無義。自語亦別異，

唯語即能碍。若由彼受許，論是定量故，應害一切事。

若不許彼者，豈是能害耶？若自語違者。則自語及論，

相違應無異。若士夫欲作，此即圓滿量。故於共許義，

捨論無過失。不現不許論，觀察則不轉。於此善觀察，

多說相違義。貪等非法本，浴能淨非法。於見不見事，

理成及自語，無害所有論，應取，觀察轉。士夫欲相繫，

欲聲所詮義，無破故無害，眾許所信解。名言所生法，

名聲之所詮。現量等所量，增能量聲說。依彼所出生，

欲轉故，無遮。已作未作等，種種性皆可。唯義相繫屬，

已生或當生，聲可。即由彼，能害諸異品。雖無名，於事，

可性。由彼力，彼繫一切聲，故於彼彼成。餘則於此中，

害因，非不共。破彼應比量，聲義根不轉。若有不隨行，

諸因名不共。如是說因喻，許具足果利。聲依止於名，

彼復唯依欲。聲成，非不成，如是說聲名。顯比量所成，

相違，無不謬。如是者即是，共稱比量故。又世間所說，

說無有比量。以是離比量，名稱異有境。故諸事有無，

隨轉比量異，彼不自在轉，顯彼從欲生。有兔不許月，

彼稱何所許？故彼無同喻，許彼是不共。非唯由此喻，

故作如是說。如是說答帝，非樹，亦害故。此龍腦汞等，

世間亦見故。由名所轉者，豈亦是不共？若有彼成者，

彼應事力成。若許由名成，則於非有兔，亦無能遮故。

若於事成彼，有兔亦無遮。若無彼事者，於有兔雖遮，

然非有過失。故事無決定，有名從聲生，諸法可說義，

於欲無碍故。若有彼成者，彼應事力成。若許由名成，

有兔無遮故。破彼可說性，從無破名生，謂名稱比量，

可有境能害。此說隨名轉，諸聲義無定。彼等若於彼，

破則成相違。故有因聲義，或是諸聲義，破勝義性者，

顯示無妨害。顯境差別故，名稱雖非是，離比量之外，

然作異，如前。成就說異者，所為故。此二，及原因廣說，

非現比妨害。

To reveal this to others, some say that it is not established by one's own seeing, but can also be established by what others see. To refute that clinging, we speak of "one's own seeing." Because it is grasped through observation, in regard to an object of inference, mere words are not accepted as a valid means of knowledge. That inference does not operate because it contradicts the teaching; for the sake of establishing something else, at that time it is not a valid means of knowledge. What is established by that is not established without residue. If it is established by that teaching, whose teaching is it? How can that teaching be established when it is obstructed by a valid means of knowledge? If one accepts its contradiction, how can it be established by that? If, excluding that, one accepts another, then it abandons the branch and is not a valid means of knowledge. By what is one's own understanding produced? Why do you not state the proof? By what reasoning do you uphold that teaching, and why do you not uphold this as well? Previously, by what proof could ordinary people understand? What proofs give rise to understanding? One who knows can cultivate that meaning, and just as one knows, one can again act. If the connection is severed, even the general is not within the scope of activity; that meaning has not even a little of the proof and the proven. When a person's intention is predominant, one who establishes the true and the false should have no reason and no pseudo-reason, because that is not based on the matter. A meaningful thing is observed in dependence on knowledge; without knowledge it is not a proof. Even if there is one, when the matter is not connected, understanding and non-understanding are the same. Intrinsic nature and resultant phenomena, or a reason whose presence is not seen—if the matter connected with that is established, why is it not established by itself? That others understand through something else is also not reasonable. Because of what others falsely imagine, one establishes their fault; because the two are connected, when one is absent the other is not refuted. The meanings of words—those sounds—are superimposed constructions; to establish that they are not reasons, because meaning is established by meaning. Constructions and teachings belong only to the speaker's desire; apart from the matter, because there is something else, the imputation is confused. Because meaning is understood by meaning, the thesis and the reason are said to have no function regarding the meaning. Therefore, they themselves do not abide in proof. If the words of the thesis can show the speaker's intention, that is a valid means of knowledge; but because it gives rise to doubt, it is not directly a proof. By stating the thesis, transmission also has no function. The reason itself has no function, yet it can show a function. That is, by the reason's object, one says it has no function. If that too has a function because the reason's statement operates, then because of doubt, desire to know, and reliance on that stage. When one accepts a dissimilar instance, because it is the same, there should be no end. The inner branches' function rightly abides in the three characteristics. In regard to that, memory arises and rightly abides in those words. If one says that because the object is not shown, the reason does not operate, then because the object is shown, one accepts that the function is accomplished. Although that was not present previously everywhere, still one says that sound is produced; in this way all is impermanent, and the meaning gives rise to the awareness of its destruction. Although the thesis is not shown, because there is no obstruction to establishment. Among the three, if one characteristic is not stated, it is incomplete. To state the thesis as the thesis—that statement is also not a fault. Because it is for stating the proof, a pseudo-reason is not a fault. Although it is stated without qualification, for one type doubt arises; all other statements of the thesis should become theses. If one says "establishment," because it is a proof, others also have no fault. Does showing the thesis make it a branch of proof? Listing and pseudo-forms and so on—the rest being unconnected is laughable. One says that is not correct reasoning; this has already been explained above. Although by stating the proof the meaning is understood, still, for the sake of no delusion, one shows the characteristic of the thesis. Substance, connector, own, and desired word—the characteristic of that is said to have four. Unestablished is not a proof; the meaning is stated as accepted by the establisher. Although the desired pervasion is not stated, the thesis, like "I use," is understood. Although all others' desires are excluded, the commentary explains the self, in order to exclude the place of doubt. This is done for that meaning. This is stated in dependence on the distinction of dharma and dharmin. Although the desired is not stated, it is understood from the occasion. The example has no fault of concomitance. As when one says "I am other," that is not established in the example. That obstructs the desired. If one is disputing something, one lists its proof; if that is not the thesis, what is the thesis? If it is something else, it is not desired, or it is without result. In this, in two forms, one shows no concomitance and various contradictions. Suppose one says that the general is accepted as the thesis. Because it has no other meaning, the body is unestablished, unestablished. Those who say the expressed object is empty are deluded by ignorance. Those who imagine a distinction between sound and pot, and that destruction is the same—no, because sound is produced in that way. Thus, having another meaning, the pot also possesses that. If, because of a distinction of sound, impermanence does not follow, then because the unconnected is excluded, it is not. Not possessing and possessing the rest—its exclusion has two kinds; because only the non-possessing is excluded, it does not turn to another. That general is the thesis, therefore it is not establishing what is already established. That dharmin distinction has no fault of non-concomitance. This shows that all refutations of dharmin distinctions and dharma distinctions and dharmin distinctions are exclusions. The dharmin is not established in that way. In the exclusion of the dharmin, the collections are not contradictory. It is not desired to be established in that way; only the dharma is the thesis. Although in regard to one dharmin many dharmas are established in dispute, one's own desire is called the thesis; therefore one speaks of "according to one's own word." If one says that because of accepting the treatise, one accepts everything, then if one meaning is obstructed, reason and thesis both have faults. To establish that sound is impermanent, because it excludes the quality of smell and earth, it should become a contradictory reason. Because of the occasion, it is not so. The rest is also the same. This dharmin is the occasion. That too harms the meaning of the treatise. If one says that by the proof, the dharma of the dharmin and so on are established by the establisher's desire. Sometimes the desired is accomplished, and through that it is also understood. To insist "you desire this" shows it is an independent action. A reason said to have no result is obstructed by confusion. It is not that the nonexistent is not produced; although it obstructs that meaning, how does it obstruct the rest? Inference, moving in its own object, is said not to depend on teaching. When that is established, that is well established; then it does not depend on the treatise. Then one abandons the dispute. Because there is no acceptance, it is not abandonment. This is a means of acceptance; although it exists, it is not a branch. Thus the two objects are pure; that is the time of upholding the treatise. Then the treatise can harm. That meaning and that are established; because it cannot be conceived as contradictory. Turning to a third place, accepting the treatise is reasonable. That and the established dharma—obstruction does not fail to arise; one should abandon it. It is not abandoning the rest, because the fault of infinite regress would follow. It is said that all examination must take the treatises; by what is this school made? Those without a school do not understand fire from smoke. That person is born alone; not seeing the virtues and faults, this school of evil spirits—alas, who sent it? If one says that in the similar example, all treatises display a single pervading nature of proof, that is the proof. This rule transcends the world. A connected dharma is not established; how is it not established? When that is established, stating the proof, how can the reason have a fault? The dharmin has no remaining dharma; the example is only non-comprehensive grasping. By words and smoke and so on, one infers mind, fire, and so on. Not confused, the proof, intrinsic nature and the meaning of the reason, because it harms what is stated, does not exclude the intrinsic nature. The meaning desired as the thesis—understanding does not fail to arise; how do others not understand by that reason? Whether or not the reason is stated, if it is contradictory, what does the establisher do here? It is not by that statement that it becomes a fault; that gives rise to harm to the treatise. Because harm is stated, there is a fault. If one does not state it, is there then no harm? Not doing what is unreasonable—how is there a fault? A pseudo-reason, for others, because it establishes one's own desire, is said to be a fault of meaningless establishment. Because that is established, this is not. If one says that because there is a treatise, something is unreasonable, then refuting others who state reasoning is the custom of kshatriyas. All meanings are the same as one's own; one cannot state the proof, so all proofs and theses are destroyed in their establishment. In regard to one dharmin, contradiction is impossible. In regard to contradictory decision, obstruction—this is not present; thus the fault of contradiction. If one says that although that is not the harmer and the harmed, because it harms the meaning of the treatise, the unconnected also harms, then everything harms everything. If one says that because there is connection, it should harm that, then no—all reasons, in regard to their own thesis, because of observing virtues and faults. Without a thesis, it does not arise. That connection—this is not understood. This only harms the treatise; that is the fault. The rest is also the same. If one says that because of accepting the treatise, the treatise sees the entire thesis. If the reason and example are not established, it should also become the thesis. Because stating the proof is not so, because stating the desired. The thesis is obtained by reasoning, not excluded by words. What is not desired is not the thesis. What the establisher does not desire, the remaining dharma is not the thesis. Then it is not the thesis; what contradiction does it harm? This is a meaning outside the characteristic of the thesis; one's own word is also meaningless. For those who wish to enter into dispute, from what does this doubt arise? That, which no valid means of knowledge refutes, how can upholding obstruct it? If a valid means of knowledge refutes it, how can it enter by words? Those who first held the thesis by their own desire, why can they not later hold another as they wish? Desire is not harmed, not established, because it shows infinite regress. To eliminate wrong constructions, this is done according to one's own word. Others say it is abandonment, the distinction made by the teaching, the commonly accepted dharmin; therefore one speaks of "according to one's own word." Because of the time of examination, the commonly accepted is based on what is already established. One's own desire to establish a distinction—because things are without dispute. If, by stating what is already established, the dharmin is not the thesis. By stating its own nature, this understands that meaning. Stating what is already established is the proof. If there is doubt, stating and showing the nature of the thesis—this has the benefit of result. By stating correct inference, it takes the general as its object. Even if this is not stated here, the thesis has not the slightest contradiction. The dharmin is the thesis; what fault does that have? That has no function. Why? Because the reason has no concomitance. Are those not faults? In dependence on later branches, it is not accepted as a fault of the thesis. If so, the faults of reason and so on should also become faults of the thesis. Because they all harm the thesis. Therefore only that which is connected is accepted as a fault of the thesis. Not the rest. As contradiction with perception and so on. Not in dependence on removal, the reason and so on are excluded by characteristics. Because non-pervasion and reversal are cut off, one states the characteristic of the thesis. "According to one's own," "only," "nature," and "not excluded," and so on, remove the reverse of its characteristic. The word "desired" removes non-pervasion. This is accepted as the characteristic of the thesis. Those without the characteristic of the thesis—when excluded, there is harm; because the other characteristic does not operate. One says "according to one's own" and "desired"; although it is understood as a specific meaning, because it is connected with the extreme characteristic, and because it excludes other times, it is stated. Desire focused on this is not a branch; the statement of the desired and so on is a branch. That is, by that, one refutes unestablished reasons and so on. Because it is not expressed, it is unreasonable; therefore "according to one's own" is one's own desire, not depending on all treatises; that desire is the thesis. Therefore what is connected with non-desire—the desired can also harm. As established, because there is no harm, thesis and reason both have no fault. What they accept is also not refuted by any valid means of knowledge. Because the reason is stated in response to doubt, one excludes what is not the basis of the reason. By the distinction of inference, one states four kinds of harm. What is accepted is a resultant reason; general renown is a nature reason. Wishing to state correct decision, because self and other are alike in belief, one combines one's own words with the treatise and states them as one kind. If one's own is not a valid means of knowledge, speech does not operate. Thus, not relying on the treatise, that which is called meaning examination is also said to be the same. At the time of that treatise, there is obstruction; to show this meaning, this one's own words are stated together. Therefore in these examples, sameness is also shown. If there are no valid means of knowledge, treatise and speech are impossible. One's own words contradict clear examples; this points to a direction. That teaching of dharma in the future does not produce pleasure—this is the example. Not relying on that stated matter, even if it contradicts the treatise, it is not harmed. Because the two functions are equal, they can obstruct. The example is like one's own words being contradicted. That too is the nature of one's own words. What valid means of knowledge is there in that, and how does it harm others? Not so—apart from reasoning, how can a thesis with equal function harm a genuine inference? Those who take the teachings as a valid means of knowledge—as extensively refuted above. Therefore, having accepted, one examines and accepts this fault. Because of harm, acceptance is not apart from inference; because it shows the distinction of objects, it makes a distinction. Not so—there would be too much, or stating a distinction would be meaningless. One's own words are also distinct; mere speech can obstruct. If by that acceptance the treatise is a valid means of knowledge, it should harm all things. If one does not accept that, how can it harm? If one's own words are contradicted, then one's own words and the treatise should be no different in contradiction. If a person desires to act, this is the perfect valid means of knowledge. Therefore in regard to commonly accepted meanings, there is no fault in abandoning the treatise. Not present and not accepted as a treatise, examination does not operate. In this, well examined, many contradictory meanings are stated. Greed and so on are the root of non-dharma; bathing can purify non-dharma. In regard to seen and unseen matters, reasoning and one's own words—whatever treatise is without harm should be taken; examination operates. A person's desire is connected; the meaning expressed by the word "desire," because it is not refuted, is not harmed; it is the belief accepted by the many. Phenomena born from names—the meaning expressed by names and sounds. What is measured by perception and so on—the valid means of knowledge is stated as the measure. Born from that, because desire operates, there is no exclusion. What is done and not done and so on—all kinds of natures are possible. Only the meaning is connected; what has arisen or will arise—sound is possible. That is, by that, it can harm various dissimilar instances. Although there is no name, in regard to the matter, possibility exists. By its power, all sounds connected with that are established in regard to each. Others, then, in this, are harmed by the reason, not the uncommon. Refuting that should be by inference; sound, meaning, and sense organ do not operate. If there is no concomitance, those reasons are called uncommon. Thus stating reason and example is accepted as having the benefit of result. Sound relies on name; that again relies only on desire. Sound is established, not unestablished; thus one states sound and name. The contradiction of what is established by inference is not without error. Thus that is indeed the inference of general renown. Also, what is said in the world is said to have no inference. Because it is apart from inference, the name has a different object. Therefore, whether things exist or not, the inference that follows differently does not operate independently; this shows that it is born from desire. The rabbit does not accept the moon; what does that name accept? Therefore it has no similar example; it is accepted as uncommon. Not only by this example is it stated thus. Thus the statement "answer to the lord" is also harmed, because it is not a tree. This dragon's brain, mercury, and so on are also seen in the world. What is turned by name—is that also uncommon? If that is established, it should be established by the power of the matter. If one accepts that it is established by name, then in regard to the non-existent rabbit, there is also no ability to exclude. If it is established in the matter, the rabbit is also not excluded. If that matter does not exist, although the rabbit is excluded, there is no fault. Therefore things are not decided; having a name born from sound, all dharmas can be expressed in meaning, because there is no obstruction to desire. If that is established, it should be established by the power of the matter. If one accepts that it is established by name, the rabbit is not excluded. Refuting that expressibility, born from an unrefuted name, one says that name inference can have an object that harms. This states that following the name, the meanings of sounds are undetermined. If they are refuted in regard to that, it becomes contradictory. Therefore, whether the meaning of a reason sound exists, or the meanings of sounds, those who refute the ultimate nature show no harm. Because it shows the distinction of objects, although the name is not apart from inference, it makes a distinction, as before. Establishing what is stated as different is for a purpose. These two, and the cause extensively stated, are not harmed by perception and inference.