Chua Duoc Su

Anger is the most powerful of the kle"sas that not only "plant seeds" for suffering but also "cut the roots of virtue" for periods of up to a thousand aeons per instance. This article examines and assesses the exegesis by Tsongkhapa, founder of the Tibetan Gelukba order, of Indian sources on the topic of anger. It argues that despite Tsongkhapa's many careful qualifications he may not be successful in avoiding the conclusion that if the sūtras are to be accepted literally, there almost certainly will be persons for whom liberation from saṃsāra is precluded.

Among the six root afflictive emotions (nyon mongs, kle"sa) identified in the Buddhist Abhidharma literature as the causes for episodes or entire lifetimes of suffering, anger (Tibetan: khong khro, Sanskrit: pratigha) holds a singular place. It is one of a few mental states (1) that not only establish "seeds" or "roots" of nonvirtue, but also nullify the seeds or roots of individual virtue planted by exemplary actions such as giving and patience. Among these states, anger is uniquely destructive. The Mañju"srīvikrīḍita Sūtra warns that a single moment of anger can result in a person's loss of a hundred aeons of virtue.(2) "Sāntideva, the ninth century author of the greatly influential Bodhisattvacaryāvatāra multiplies this dire warning tenfold (chapter 6, verse 1): anger wipes out not just a hundred, but a thousand aeons of virtue.

Since most people lose their tempers with dismaying frequency, it seems reasonable to wonder how, from a Buddhist perspective, it is possible simultaneously to contend that a mere outburst can have such an extraordinarily negative effect and to assert, as Mahāyāna Buddhists generally do, that all sentient beings will gain merit sufficient to attain liberation. It appears that apologists for the Mahāyāna tradition have a heavy burden--they must either interpret statements about anger's effect on the stores of virtue as gross exaggerations spun out as a matter of "skill in means" (thabs la mkhas pa, upāya-kau"salya), delimit the range of persons to whom they are said to apply, or indicate ways in which anger's effects can be ameliorated.

Tsongkhapa Losang Drakba (3) (tsong kha pa blo bzang grags pa, 1357-1419), founder of the Tibetan Gelukba (dge lugs pa) order, uses the latter two routes in his extensive analysis of anger. I will examine portions of his Great Exposition of the Stages of the Path (completed in 1402) and his Illumination of the Thought, Extensive Explanation of (Candrakīrti's) Madhyamakāvatāra (1418) where Tsongkhapa meticulously scrutinizes the Indian sources. In these works he assesses the importance of the status of the recipient of anger and attempts to explain what precisely it does and does not mean to "cut the roots" of virtue for the incredible spans indicated in the Indian sources.(4) In the process, he manages to limit significantly the scope of the Indian sources, explaining that they refer only to anger at bodhisattvas and that "cutting" the roots of virtue means something far less than "destroying." However, it is not clear that, in the end, Tsongkhapa has succeeded in demonstrating that anger does not, at least in some cases, prevent salvation.

Although clearly the Buddha regarded anger as a massively destructive force, sūtra sources that quantify its effect are noticeably scarce. Tsongkhapa cites the Upāliparipṛcchā Sūtra, the Mañju"srīvikrīḍita Sūtra, and the Sañchayagāthāprajñāpāramitā Sūtra. The Upāliparipṛcchā Sūtra (5) warns that there is no greater cause for elimination of the roots of virtue "than when one spiritual adept (brahmacārya) abuses another," but it does not specify how great that loss might be. For that, the locus classicus appears to be the aforementioned Mañju"srīvikrīḍita Sūtra which warns that one may lose a hundred aeons of virtue in a moment of anger. Candrakīrti (7th cent.), the Mādhyamaka interpreter through whom Tsongkhapa views nearly all important matters of Buddhist doctrine, possibly basing his estimate on this source, also states that anger destroys a hundred aeons of virtue (Madhyamakāvatāra 3.33):

Candrakīrti clarifies the sūtra by indicating that hundred-aeon anger is directed at a "Conqueror Child," or bodhisattva--a person who, for Tsongkhapa, has an aspiration to Buddhahood both altruistic and spontaneous (but who is not necessarily someone who has amassed significant amounts of merit or wisdom). This, of course, greatly reduces the probable instances of hundred-aeon anger by an ordinary person. Tsongkhapa also cites "Sāntideva,(6) who without specifying the recipient of anger, says (Bodhisattvacaryāvatāra 6.1):

Aware that Candrakīrti has specified that the recipient of hundred- aeon anger is a bodhisattva, he surmises that the recipient of the thousand-aeon anger mentioned by "Sāntideva must also be a bodhisattva and, moreover, that the angry person must be a non- bodhisattva.(7) If this is what "Sāntideva meant, we might suppose that a thousand-aeon penalty would be a rather rare occurrence. Given a bodhisattva's generally benign behavior, presumably a bodhisattva would rarely commit acts that would spur the wrath of others.(8)

Continuing with this line of reasoning, Tsongkhapa concludes that if the supreme penalty for anger involves a non-bodhisattva's anger with a bodhisattva, then Candrakīrti's reference to a lesser penalty that also involves anger with a bodhisattva can only mean that one bodhisattva is angry with another. It is surprising to learn that bodhisattvas ever get angry, since they are, for Tsongkhapa, persons always able to rouse their bodhicitta, the altruistic aspiration to Buddhahood. However, although bodhicitta can arise spontaneously, it is not continuously present in non-buddhas, and at least some bodhisattvas are susceptible to anger for nearly all of a period of "uncountable" aeons. This is the length of the paths of "accumulation" (tshogs lam, saṃbhāramārga) and "preparation" (sbyor lam, prayogamārga), the first two of the five paths concluding in Buddhahood.(9) Anger is not precluded until one is well into the path of preparation, the second part of which is called "peak" (rtse mo, mūrdhan) because it is the end of the period in which one can generate anger that will sever the roots of virtue. At least one contemporary scholar says that a bodhisattva may become angry even after that point, but the anger is weaker than the anger to which the quotations refer and will not sever the roots of virtue.(10)

Tsongkhapa is very specific about the consequences of being an angry bodhisattva. A mature bodhisattva who is angered by one who is lesser (11) loses a hundred aeons of virtue; on the other hand, a bodhisattva angry with a greater one loses an aeon of virtue for each instant of the anger's duration. In the latter case, Tsongkhapa has a source in the Sañchayagāthāprajñāpāramitā Sūtra, which states: (12)

Tsongkhapa interprets this to mean that a bodhisattva's anger with one who has received the prophesy of Buddhahood from a Buddha will impede the former's progress for many aeons. For example, someone about to progress from the path of accumulation to the path of preparation would be set back for as many aeons as there were instants of anger. Presumably the number of instants would swiftly rise above one hundred, since anger has more serious consequences for lower persons than high ones and otherwise the greater bodhisattva would pay a higher price than a lesser.

According to Tsongkhapa, it does not matter whether one knows as a bodhisattva the person with whom one is angry. This is unexpected. Tibetan discussions of karma virtually always classify correct identification of the recipient of an action as a primary consideration in the determination of a specific act's weightiness. It is of lesser consequence, for instance, to shoot a gun at a coiled rope in the corner of a darkened room that one believes mistakenly to be a snake than to shoot at an actual snake. However, perhaps Tsongkhapa would answer that even if one does not realize that the person at whom one is angry is an actual bodhisattva, one certainly would have had experience of that person's compassion; one therefore would have correctly identified the fundamental character of the person even if one did not realize that the person merited the title "bodhisattva." If so, it would support the view of the contemporary Gelukba scholar, Geshe Kelsang Gyatso (14) , who contends that anger toward anyone who has shown one great kindness is a source of "limitless destruction of merit." To become angry even at an equal, he continues, may cost roots of virtue collected over many lifetimes. To speculate, perhaps this is because anger mixed with ingratitude contributes to pride and other kle"sas. This modern interpretation seems consonant with the thrust of the Indian sources.

Whoever is the recipient of one's anger, clearly anger is considered an immensely negative force. We would not be surprised to learn that anger could result in rebirth in a hell for thousands of years or that it might give one who had an otherwise fortunate birth an ugly countenance. But anger is far worse. What makes anger different from most other nonvirtues is that it not only contributes to the store of causes for miserable future experiences but also affects the store of causes for fortunate experiences.

Tsongkhapa calls the principal effect of anger "cutting" the "roots of virtue" (dge rtsa, ku"sulamūla).(15) Ways to cultivate and "plant" roots of virtue were a major concern in early Buddhism, as Robert Buswell has recently shown.(16) For instance, roots of virtue are a major topic in the Abhidharmamahāvibhāṣā (second cent.), the document from which comes the name of the Vaibhāṣika school that, according to Tibetan doxographers, is one of the two principal Hīnayāna systems. Subsequent theoreticians of karma retained the horticultural metaphor but switched to the image of "seeds" rather than roots; Tsongkhapa treats "roots" and "seeds" as synonymous terms. Both refer to the establishment in an individual continūm of a potential for future effects. Both virtuous intentional actions and nonvirtuous intentional actions infuse an individual continūm with potentials. The nature of these potentials --are they physical? mental? neither?--has long been debated in Buddhist scholasticism.

Again, the Indian texts seem to warn clearly and unambiguously that even a moment of anger can wipe out the virtue one has accumulated over the course of aeons. What else might it mean to "destroy" (bcom) virtue "from the roots?" When a plant's roots are cut, it usually dies. Alternately, when its seeds are destroyed it can no longer bear fruit. However, some plants, such as the sweet potato, do not die when their roots are cut; they lie dormant until the conditions exist for their regeneration, or they slowly produce new root systems. Tsongkhapa, it seems, considers virtue to be a sweet potato. He explains that when anger "cuts" virtue's "roots," it is not destroyed, although aeons will roll on before it again becomes capable of producing the sweet fruit of a pleasant rebirth. Therefore, "destruction of the roots of virtue" (dge pa'i rtsa la bcom pa) is not equivalent to "totally cutting the roots of virtue" (dge pa'i rtsa ba kun tu chad pa, samucchinnaku"salamūla) which for some early Buddhists meant a permanent disbarment from liberation.(17)

This is the picture that emerges from Tsongkhapa's reflections in the "Patience" chapter of his Lam rim chen mo. It arises as he addresses himself to certain unnamed scholars, apparently (18) the followers of Bud"on (bu ston, 1290-1364), the prolific scholar of the Sagya (sa skya) sect whose influential works were still reverberating when Tsongkhapa began his Buddhist studies. He affirms Bud"on's basic interpretation: despite the presence in the Indian sources of apparently unambiguous language such as "destruction" or "elimination," the "seeds" (sa bon, bīja) established by virtuous actions are certainly not destroyed by negative emotions such as anger; they are merely incapacitated. They cannot be destroyed by anger because only wisdom--consciousnesses at the level of the path of seeing (mthong lam, dar"sanamārga) and above--can eliminate karmic seeds. That is, until one has experienced emptiness (stong pa nyid, "sūnyatā) mystically--without any dualities, without conceptuality--liberation from any sort of karma and its results is impossible. Hence, the language of the Indian texts is not literal, but must be interpreted in the following way: because the seeds of virtue cannot reach fruition, for the angry person it is as though the roots of virtue were destroyed.

Although it is not a question Tsongkhapa addresses explicitly, we can see that by interpreting "cutting" as something less than destruction, Tibetan exegetes seek to avoid a serious challenge to the Mahāyāna doctrine of universal salvation (namely, that all sentient beings will eventually reach Buddhahood). If anger can be so potent, and as we know too well ourselves, occur so frequently, then certainly how could they ever have fortunate rebirths in which to make progress toward Buddhahood? Asa"nga (fourth cent.), in his Abhidharmasamuccaya, asks just this question.(19) He makes a distinction between "roots" and "seeds" of virtue and nonvirtue such that it might be possible for someone to have lost roots but not seeds and therefore retain the possibility of future regeneration of the roots of virtue. However, he contends that there are some among those whose roots of virtue are eradicated who also have no seeds of virtue and therefore have no "dharma of parinirvāṇa." They make saṃsāra truly endless, for they themselves will never escape it. Tsongkhapa makes no distinction like Asa"nga's between "roots" and "seeds" and does not admit the possibility that some are doomed to endless saṃsāra. He appears to think that since the roots of virtue can be regenerated, and since, moreover (as we will see below), their period of dormancy can be abbreviated by the application of proper antidotes to the poison that has deadened them, no such result need be entailed. As I state later, it is not clear that his explanation succeeds.

Although Tsongkhapa agrees with Bud"on and his followers that the roots of virtue continue to exist despite anger, he disagrees with them over whether this will entail adverse consequences. The problem, they think, with asserting that virtue might still exist despite having been "cut" is that it might then seem to follow that if certain precise conditions were to occur, virtue's seeds might yet sprout; therefore, anger would not actually have had a deleterious effect on virtue. For example, if virtue continued to exist, could not a wayward monk whose temper too often bested him somehow still experience the effects of past virtue? Tsongkhapa's response falls under several heads.

In the first place, Tsongkhapa wishes to establish that karmic seeds can exist without ripening even in the presence of conditions that ordinarily would cause them to "sprout." He relies on Bhāvaviveka's (sixth cent.) Madhyamakahṛdayavṛttitarkajvālā to assert that, for example, even an "ordinary" person (one who has not had the mystical experience of emptiness) can use the "four opponent powers"--remorse, restraint, the cultivation of specific "antidotes," and cultivating bodhicitta (20) -- to suppress the issuance of the effects of nonvirtue. One might regret a harsh utterance, pledge not to repeat the behavior, cultivate loving-kindness, and so forth. Just as anger cannot destroy the seeds of virtue, so the four powers do not destroy the seeds of non-virtue, but they do prevent their unpleasant effects from being issued.(21) This suppression of the maturation of negative karmic seeds is commonly called "purification," which one might incorrectly assume entailed complete destruction or elimination, but as we have seen can mean only temporary incapacitation.

Another of Tsongkhapa's examples involves a far more advanced person who has attained the path of preparation, that level at which, according to Tsongkhapa, there has been an inferential understanding of emptiness.(22) For such a person, the attainment of a higher path consciousness ensures that even the presence of what ordinarily would be proper ripening conditions will still not lead to the maturation of those seeds of non-virtue that could ripen as wrong views (log lta, mithyāḍṛṣṭi) or birth in the miserable realms of animals, hungry ghosts, and hell-beings. As in the previous instance, although the seeds cannot yet be destroyed, they can be incapacitated.

Indeed, all "heavy" karma, the sort that results in particularly fortunate or miserable birth, suppresses the issuance of effects that are contrary to it. For instance, a hell-being never experiences pleasure, nor does a god experience pain (until, after vast stretches of time, his or her birth-impelling karma approaches exhaustion). Therefore, in Buddhist cosmology, the incapacitation of seeds of nonvirtue or virtue is a common occurrence.(23)

Tsongkhapa's final example is not as obvious as the others. Among seeds that exist without ripening are those that have already ripened, yet continue to exist. Commenting (in Illumination) on a passage in the Akṣayamatinirde"sa Sūtra that compares virtue to a drop of water placed in the ocean, remaining as long as the ocean endures, Tsongkhapa says,(24) "Virtuous roots are not consumed through the emergence of their effects; however, it is not the case that anger does not consume them." In Great Exposition he says, "Even with regard to virtuous and nonvirtuous actions that have ceased upon issuing their own maturation, there has not been an elimination of their seeds."(25) In brief, he says that actions can cause effects without being "used up."

How could "ripened seeds" continue to have any kind of existence? We must recall that seeds established by virtue (or nonvirtue) cannot be destroyed by anything other than wisdom of the path of seeing or above; therefore, they are not destroyed even if their effects have already issued forth. This point is, perhaps, counter-intuitive: once a seed has produced its effect, what sense can be made of saying that it continues to exist? It is as though one were to say that despite the fact that a seed had developed into a tree, the seed continued to exist (although it could not, of course, produce yet another tree). I think, however, that Tsongkhapa's point is considerably more subtle. He expands upon it in Illumination, commenting on Candrakīrti's statement in Madhyāmakāvatāra (6.33) that:

Tsongkhapa comments (27) In the [non-Prāsa"ngika] systems, they think: "When a thing such as a sprout has disintegrated, everything that is part of the sprout is obliterated." Since one does not get any other thing that is different from a sprout, such as a pot, they assert that disintegratedness (zhig pa) is utterly not a thing. In the [Prāsa"ngika] system, for example, one cannot designate (1) Upagupta's individual five aggregates (phung bo, skandha), (2) their collection, or (3) that which is a different entity from those two as an illustration of Upagupta, and Upagupta is also unsuitable to be an illustration of those three. However, it is not contradictory that despite that, what is designated as Upagupta in dependence on his aggregates is a thing. Similarly, even though disintegratedness also cannot be an illustration of either the thing (dngos po, bhāva) that has been destroyed or anything that is the same type as that, it is a thing because it is produced in dependence on a thing that has been destroyed.

In Tsongkhapa's view, karmic "seeds" are neither physical presences nor even mental phenomena that persist over time. Then, what are they? They are "disintegratednesses" (zhig pa, naṣṭa--there is no graceful English term). Because all impermanent phenomena disintegrate moment-by-moment, when an action disintegrates, its state of having disintegrated--its disintegratedness--arises. It too, disintegrates, giving rise to the "disintegratedness of the disintegratedness" of the action, and so on, until a fruition occurs. Hence, "seed" really refers to the present moment of "disintegratedness" of an original action. Asserting that "disintegratedness" is a functioning entity but denying that it is substantially existent allows Tsongkhapa to avoid either the absurdity of saying that karma persists unchanged or of proposing a substantially existent entity like the Vaibhāṣika "acquisition" (thob, prāpti) to account for the continuing link between a mind-stream and a karma.(28)

Based on his understanding of Prāsa"ngika philosophy, Tsongkhapa describes all phenomena as mere imputations designated in dependence on certain bases. In his example, a man named Upagupta is not identical with the body and mind in dependence on which "Upagupta" is designated. This is Upagupta's mode of existence because he is empty of inherent existence (rang bzhin gyis grub pa, svabhāvasiddhi). Nevertheless, Upagupta exists. Similarly, says Tsongkhapa, the "disintegratedness" of a virtuous action exists upon the action's disintegration. Although there is nothing to which one can point that is the "disintegratedness" (just as there was nothing to which one could point which was Upagupta), nevertheless there is a basis--the disintegrated action--in dependence on which "disintegratedness" can be designated (just as there is a basis--a body and mind--in dependence on which "Upagupta" can be designated).

The consequence of this is that Tsongkhapa feels that it is possible to assert that even when an actions's fruition has been experienced, the action's disintegratedness, which functions as its "seed," does not cease. Of course, how could "disintegratedness" ever cease to exist? Once something has disintegrated, it will always be true that it has disintegrated. Thus, there is no way that anger could destroy the seeds of virtuous actions.(29) (It may also be that this manner of explaining the persistence of virtue's "seeds" even when virtue has been "ripened" has to do with denying that the accumulation of merit is a "zero-sum game." That is, although virtue might ripen in fortunate rebirths, it continues to "count" toward the store of merit that comprises half -- the other half being the store of wisdom -- of the requisite for Buddhahood.) (30)

We might have expected the opposite, namely, that one result of the incapacitation of virtue would be a neutralization or reversal of its "habitual" effect, the establishment of propensities for further virtuous action.(31) Apparently Tsongkhapa feels that although the seeds are incapacitated, the habits are not necessarily broken. This makes sense because even angry persons may have had much conditioning to predispose them to virtuous behavior. Certainly this would be true in the cases of the bodhisattvas who become angry with each other or with common beings. It would contradict what we observe daily to maintain that a moment of anger dramatically and permanently alters an otherwise balanced or even benevolent personality.

If, then, one can accumulate more virtue, does this mean that there are "fresh seeds" that might ripen as fortunate rebirth or pleasant experiences? If so, does this not considerably reduce the negative effect of anger? Since this would otherwise constitute a major loophole in Tsongkhapa's formulation, we can perhaps presume that these seeds, too, are incapacitated by anger. This assumption is consistent with the basic thrust of Tsongkhapa's interpretation of the meaning of "cutting the roots of virtue," since it looks forward toward aeons in which there will be no ripening of the seeds of virtue rather than looking backwards at so many aeons of virtue ruined. Of course, it leads to the apparent paradox that the stores of virtue may be increased during the same period in which virtue's roots are "cut" and raises questions such as: if new roots of virtue are produced -- and incapacitated -- does this mean that some dormant "older" roots are activated? In other words, does anger affect a certain quantity of virtuous roots?

Besides tempering what the Indian texts say about the existence of virtue and affirming the possibility of its performance, Tsongkhapa also distinguishes between degrees of anger, only the worst of which truly "cuts" the roots of virtue. Thus, he appears to think that although in general, the seeds of virtue cannot ripen, there may be exceptions. He says: (32) The overcoming of a virtue does not mean that a virtue in one's continūm ceases to exist after one generates anger; rather, anger harms the virtue's capacity to issue forth an effect. The extent to which later fruition is harmed accords with the amount of harm done, causing a small, middling, or great extinguishment of virtue as explained above.

Tsongkhapa is referring to the Upāliparipṛcchā Sūtra, which said: Upāli, I have not seen such a drawing of a wound or maiming as when a trainee in the pure life (brahmacārya) abuses [another] trainee in the pure life. Upāli, then those great roots of virtue become diminished, thoroughly reduced, and eliminated. Upāli, if you would not try to attack with your mind things such as burning logs, what can we say about a body with consciousness?

Tsongkhapa interprets "diminished," "reduced," and "eliminated" respectively as small, middling, and complete elimination. That is, he argues that although it is true that anger cuts the roots of virtue, it may do so only partially. It is not clear whether this means that in "small" or "middling" eliminations only some roots of virtue are touched or whether it means that all roots of virtue are diminished significantly, so that only partial fortunate results are possible.

In short, Tsongkhapa argues that although the Indian texts warn of draconian consequences to even a moment of anger--the loss of a thousand aeons of virtue, for instance--this really means, in most cases, that there is a partial incapacitation of that virtue for a long future period. The result is that some of the seeds of virtue might actually ripen as a good body with good resources, etc., and because of this one could probably continue to make progress as a bodhisattva on the paths to Buddhahood. However, one's progress will be slow. A novice bodhisattva's anger at a mature bodhisattva, for instance, will not de-commission her, but it will impede her development. Therefore, "cut" not only means just "incapacitation" but also just "mostly incapacitated." Perhaps anger incapacitates those roots of virtue that would have ripened as lifetimes with superb conditions for the study of Dharma, enabling only those roots of virtue that could ripen as lifetimes or circumstances that are relatively mediocre.

According to Tsongkhapa's own reckoning, the journey over the paths to Buddhahood requires no less than three periods of "countless" great aeons. In a sense, then, a moment of anger amounts to no more than a stumble on the path. On the other hand, who gets angry only once in a great while -- like a thousand aeons, for instance? Even with Tsongkhapa's modifications, it seems likely that an ordinary person would have little virtue not incapacitated by anger. The most serious problem with any of the accounts of the effect of anger, then, is that they seem to leave open the possibility that there might be persons who would be the karmic equivalent of indentured servants, unable ever to be born into a body from which they could seek liberation. This would contradict a deeply held dogma about the possibility of universal salvation, which Tsongkhapa supports.(33) A single lifetime's episode of anger (particularly if that life is spent largely being jealous of one or more real bodhisattvas) could easily dig a hole so deep that even innumerable aeons seem too brief to permit escape. For, one of the principal reasons why vice is vicious is that it impels one into life after suffering life in which anger, among other negative emotions, is the norm rather than the exception. The Tibetan tradition uses a famous analogy to saṃsāra which compares the chances of being born as a human who can hear the Dharma to the odds that a blind sea turtle, surfacing only once in a hundred years, will stick its head through a golden yoke floating on a vast ocean. Adding multiple-aeon calculations on anger is like changing the setting of this scenario to outer space. What odds remain? In short, Tsongkhapa's efforts at moderation notwithstanding, the Indian sources seem to lead to an untenable conclusion.

Second, an apparent self-contradiction in Tsongkhapa's interpretation is that he himself maintains that regarding the roots of virtue, "cut" cannot be equated with "delay" or else there would be no great difference between anger and other negative emotions such as jealousy or gossiping, which also can delay the issuance of the effects of virtue. He says: "The mere temporary postponement of maturation is not appropriate to be the meaning of destroying the roots of virtue; otherwise, all the nonvirtuous actions that have power would have to be set forth as destroyers of the roots of virtue."(34) Based on our analysis, it is difficult to see how his understanding of the destruction of roots of virtue amounts to anything other than delay, since anger, though much more potent than any of these other kle"sas, seems to be different only by degree. However, Tsongkhapa focuses upon the way in which the other kle"sas cause a delay: (35)

The fruition of the seeds of any virtuous or nonvirtuous action can result in a birth that prevents the maturation of seeds of its opposite. For instance, a seed established by nonvirtue might ripen as a birth in one of the hells. Because such a life is devoid of pleasure, seeds formerly established by virtue would lack the necessary conditions for their maturation. These seeds would not have been rendered ineffective in exactly the same way that anger renders seeds ineffective; they would not have been "cut" or "scorched" or "withered" or otherwise directly neutralized. They would be like patrons in line for a film who do not know that around the corner, near the box office, others are cutting in. But what difference does it make that anger and pride, for instance, operate differently? In practice, the result is the same.

By focussing on the narrow issue of how and to what degree anger affects the stores of virtue, we have not yet asked several obvious questions that probably should be raised before leaving the topic. First, why is anger seen to be so incredibly destructive? There is no other religious tradition that approaches Buddhism in its negative assessment of the consequences of a moment's angry outburst. What is special about anger for Buddhists?

Let us look at Tsongkhapa's arguments against anger (in which he follows the lead of "Sāntideva's Bodhisattvacaryāvatāra).(36) (1) Anger against others is irrational because others lack autonomy. They are helpless against their own conditioning, which leads them to commit acts that provoke us to anger. It is obvious that they lack autonomy because even though they themselves want happiness, they commit acts that lead to suffering. (2) Similarly, if one thinks that others are inherently annoying, they cannot rationally be blamed, since they are merely doing what is their nature. (3) If, on the other hand, their annoying qualities are not inherent, then those qualities are a merely adventitious product of conditioning and should not be held against them. (4) The provocative person is only indirectly responsible for annoyance; he or she is being used by hate in the same way that a person uses a stick. One should oppose the annoyance, not the person ("love the sinner, hate the sin"). (5) Whatever makes one angry is the result of one's own past actions. Annoying persons are nothing other than the agents of one's own previous misdeeds. And (6) only a provocative person gives one the opportunity to amass merit that can be helpful for spiritual progress. Therefore, one ought to be grateful for the provocation.

Note that focus is on what happens to a person who gets angry, not on the immediate consequences to the recipient of the anger. In other words, Tsongkhapa does not argue that anger ought to be avoided because it leads to violence against others or because it tends to provoke the recipient of one's anger into an equally angry state. These would be legitimate arguments, but Tsongkhapa's concern is for the mental state of the person who gets angry. He wishes to convince us that anger is simply irrational and that forbearance is beneficial, not that anger is wrong because it leads to physical or verbal acts (as he might argue if, for instance, he were addressing the faults of intoxication). It is a reminder that karma is primarily about intention (sems pa, cetanā), not act.

What is noteworthy about these arguments is that most of them revolve around the angry person's assumption of autonomy with respect to a provocateur--around the sort of ignorance that Buddhists identify as the "root of saṃsāra." To be angry with someone implies that one falsely imputes to that person an autonomous self, and the dynamics of anger serve to reify that misconception. Tsongkhapa also demonstrates that anger involves ignorance about oneself, for it indicates that one does not understand that harms, real or imagined, arise only in dependence upon one's own continūm.(37) Because anger reifies ignorance, it is strongly contrary not only to the development of wisdom but also to the development of compassion, which grows only where the distinction of self-and-other has been weakened. Perhaps, then, anger is felt to be in a different class than other nonvirtues because even more than desire, etc., it solidifies that most vicious of all vices, ignorance? That is why anger joins ignorance and desire to comprise the "three poisons" functioning as the hub of the wheel of saṃsāra.

Moving to a second question, why do Buddhists say that anger affects virtue instead of simply saying that anger is a nonvirtuous act that carries tremendous potential for future suffering? Why place anger (and a few other nonvirtues, as described below) in a different category than any other act? Perhaps the answer is that anger does not merely set in motion a future retribution and habituate the one it grasps to further outbursts; it creates a mood, or is one, which undermines positive thoughts and actions. It would not be sufficient on the plane of ordinary experience to describe anger's effect only in terms of future negative effects. We would surely also want to add that anger diminishes positive movement. Thinking homologically, it must seem necessary in karmic theory to claim that anger produces not only roots of nonvirtue but affects the roots of virtue as well.

This is equally true of weighty virtues, such as giving. They establish roots or seeds for future pleasant lives or experiences, but they also "purify" nonvirtues (as we saw above when we considered the four powers that can temporarily nullify nonvirtues). The language of cleaning, rather than that of destruction, is used; for instance, we are not told that generosity "cuts the roots of nonvirtue." With virtues, what Buddhist teachers emphasize are ways in which the fruition of the virtues will enhance the attainment of liberation for oneself and others.(38)

Finally, one question that might be raised with regard to the purification of nonvirtue is what consequence this might have for virtue. We have seen that according to Tsongkhapa, anger can be nullified by the four opponent powers of remorse, restraint, etc.(39) But if anger is nullified by remorse, etc., is its nullification of virtue similarly cancelled? Are the roots of virtue then freed? Or does one just establish roots of liberation? Tsongkhapa, commenting in Great Exposition on Bhāvaviveka's statement that even though there is purification by the four powers, there is no destruction of seeds, concludes that "even though your accumulation of sins is washed away through purification by the four powers, this does not contradict the fact that you are slow to produce higher paths."(40) In Illumination he is even more explicit; referring to the Sarvavaidalyasaṃgraha Sūtra, he says: (41) If one abandons the doctrine as set forth in the sūtra but confesses the fault three times daily for seven years, the fruition of that deed is purified, but even at the fastest ten aeons are necessary to attain endurance [i.e., to progress to the next path]. Thus, even though confession and restraint in many ways does not restore a path that has become slower, it will purify experience of the fruition.

Anger, identified along with ignorance and desire as a "poison" that generates saṃsāra, is singled out by Tsongkhapa as a particularly destructive emotion. It is founded on ignorance and reifies it. It not only establishes potentials for future occasions of suffering and habituates its subject to react similarly in future provocative circumstances, but also has a considerable impact on the store of previously accomplished virtue. The magnitude of its effect on virtue is dependent on (1) the degree of anger, (2) the status of the person with toward whom it is directed, (3) the status of the person who is angry, and (4) whether it is "purified" by the four opponent powers. To expand briefly on these points:

Although the effect of anger--or, at least, intense anger--is to "cut the roots (or, destroy the seeds) of virtue," this does not actually mean that virtue is destroyed, for nothing other than a wisdom consciousness can destroy karma. Rather, the roots or seeds of virtue are incapacitated. Consequently, one may be reborn many times in the miserable realms below the level of humans, or, if born a human, will be unable to make much spiritual progress.

Tsongkhapa's attempt to explain and moderate the position of the Indian texts is not wholly convincing. On the one hand, since anger only temporarily incapacitates the roots or seeds of virtue, it is not clear how it differs from other kle"sas such as pride. Tsongkhapa himself says "cut" must mean more than "delay" but in the final analysis it appears to mean that and nothing more. On the other hand, even if anger means only incapacitation, its extraordinary damage spreading over many aeons, based on as little as a moment's outburst, seems to make liberation a practical impossibility for most persons. Tsongkhapa's interpretation would have to be even bolder--or, anger of the root-cutting variety would have to be clearly restricted to only the most extraordinary moments of rage--to avoid this untenable conclusion.

Candrakīrti nḍ. Madhyāmakāvatāra. P 5261, P 5262, vol. 98. Edited Tibetan: Louis de La Vallee Poussin, Madhyamakāvatāra par Candrakīrti. Bibliotheca Buddhica IX (Osnabr™ck: Biblio Verlag, 1970). English translations: C. W. Huntington, Jr., The Emptiness of Emptiness (University of Hawai'i Press, 1989). Chapters 1-5: Jeffrey Hopkins, Compassion in Tibetan Buddhism (Valois, NY: Gabriel/Snow Lion, 1980). Chapter 6: Stephen Batchelor, in Geshe Rabten, Echoes of Voidness (London: Wisdom, 1983): 47-92.

-1973 ('Jams dbyang bzhad pa). Dbu ma chen mo.Great Exposition of the Middle Way/Analysis of (Candrakīrti's) "Entrance to (Nāgārjuna's ) 'Treatise on the Middle Way'", Treasury of Scripture and Reasoning, Thoroughly Illuminating the Profound Meaning [of Emptiness], Entrance for the Fortunate (dbu ma chen no/ dbu ma 'jug pa'i mtha' dpyod lung rigs gter mdzod zab don kun gsal skal bzang 'jug ngogs). In Collected Works of 'Jam-dbya"ns-bzad-pa'i-rdo-rje, Vol. 9, Drashikyil edition (New Delhi: Ngawang Gelek Demo). English translation (section on two truths): Guy Newland, The Two Truths, PhḌ. Dissertation, University of Virginia, 1988 (not included in 1992 book published by Snow Lion.)

Great Exposition of Tenets/Explanation of "Tenets", Sun of the Land of Samantabhadra Brilliantly Illuminating All of Our Own and Others' Tenets and the Meaning of the Profound [Emptiness], Ocean of Scripture and Reasoning Fulfilling All Hopes of All Beings (grub mtha' chen mo/grub mtha'i rnam bshad rang gzhan grub mtha' kin dang zab don mchog tu gsal ba kun bzang zhing gi nyi ma lung rigs rgya mtsho skye dgu'i re ba kun skong). Drashikyil edition (Mundgod: Gomang). English translation (beginning of the Prāsa"ngika chapter): Jeffrey Hopkins in Meditation on Emptiness (London: Wisdom, 1983).

-1985 Great Exposition of the Stages of the Path/Stages of the Path to Enlightenment Thoroughly Teaching All the Stages of Practice of the Three Types of Beings (lam rim chen mo/skyes bu gsum gyi rnyams su blang ba'i rim pa thams cad tshang bar ston pa'i byang chub lam gyi rim pa). P 6001, vol. 152. The edition used for this article is the Lhasa bya khyung edition published by Mtsho sngon mi rigs Printing Press. Other editions have been published in Dharamsala (Shes rig par khang, 1964) and Delhi (Ngawang Gelek, 1975).

(1) Anger is not unique as a negative emotion that can "cut the roots" of virtue. In Illumination of the Thought Tsongkhapa cites sūtra passages collected in "Sāntideva's "Sikṣāsamuccaya that identify other extremely counterproductive notions such as disbelief in cause and effect, boasting about spiritual attainments one does not have, etc. as root-cutters. He also mentions that the Aakā"sagarbha Sūtra identifies root infractions of bodhisattva vows as root-cutters. See Illumination: 57a.5-57b.1. Of course, none of these are said to have the force of anger.

(3) To represent Tibetan names, I use a modified form of Jeffrey Hopkins' "essay phonetics" (Hopkins 1983: 19-22) system for Lhasa dialect. I drop the hyphens and tonal marks, and I make an exception for the name Tsongkhapa, which has become widely known in that form (it would otherwise be spelled Dzongkaba). I use the phonetic form because it is well known that we can't remember what we can't pronounce, and I think that the names of the best Tibetan scholars are worth remembering. Tibetan book titles are translated into English to avoid the consonant-cluster nightmare of transliterated Tibetan that alienates those who are not Tibetanists and to avoid the phonetic form that alienates Tibetanists. The translation of titles also reminds readers of book contents.

(4) Great Exposition of the Stages of the Path (lam rim chen mo) is the common name for Stages of the Path to Enlightenment Thoroughly Teaching All the Stages of Practice of the Three Types of Beings (skyes bu gsum gyi rnyams su blang ba'i rim pa thams cad tshang bar ston pa'i byang chub lam gyi rim pa). It is Tsongkhapa's grand synthesis of Indian materials pertaining to the enlightenment path. It has been partially translated by Alex Wayman (Calming the Mind and Discerning the Real in 1978 and Ethics of Tibet in 1991). A new translation of the entire text is in preparation by a team working under the auspices of the Tibetan Buddhist Learning Center in Washington, New Jersey and is scheduled for publication in three volumes by Harper in 1996-98. My own interest in the topic of anger was raised by my translation of the "ethical discipline" and "patience" chapters for the project. Illumination of the Thought (dbu ma dgongs pa rab gsal) is the common name for Illumination of the Thought, Extensive Explanation of (Candrakīrti's) "Entrance to (Nāgārjuna's) 'Treatise on the Middle Way'" (dbu ma la 'jug pa'i rgya cher bshad pa dgongs pa rab gsal). It is Tsongkhapa's attempt, late in life, to clarify the thought of Candrakīrti, who he saw in turn as the most important of Nāgārjuna's Mādhyamika successors. Since Candrakīrti's discussion in the Madhyamakāvatāra revolves around the ten bodhisattva grounds, Illumination is also concerned with many of the same issues as Great Exposition and is also characterized by copious citations from Indian texts. It has been partially translated by Hopkins (1980; chapters 1-5) and Klein and Hopkins (Klein 1994; first part of chapter 6). Tsongkhapa makes similar statements in both sources (in fact, much of the text of Illumination on this topic has simply been lifted from Great Exposition. The principal difference is that in the later Illumination he clarifies a few matters (for instance, the precise parties to whom he believes the Indian texts refer).

(9) The extensive Gelukba sa lam (=bhūmi and mārga, grounds and paths) literature is based on Maitreya's Abhisamayālaṃkāra (which in turn is based on the Prajñāpāramitā literature, where the five-path scheme can be dimly discerned) and Haribhadra's commentary; it also uses the five-path scheme of Kamala"sīla's Bhāvanākrama (following a much older tradition evident even in Sarvāstivādin texts--Hirakawa 208ff.). In brief, the bodhisattva path of accumulation begins with the initial generation of bodhicitta, and the path of preparation with a union of calm abiding (zhi gnas, "samatha) and special insight (lhag mthong, vipa"syanā) with emptiness (stong pa nyid, "sūnyatā) as the object.

(11) The angry bodhisattva must still be a relatively low one since a bodhisattva who has progressed past the third of the ten bodhisattva bhumis (a pre-Mahāyāna system adapted to Mahāyāna in, for instance, the Da"sabhūmika Sūtra) is no longer ever subject to anger. This qualification can be found in Maitreya's Abhisamāyalaṃkāra and elaborated in subsequent treatments of the bodhisattva path (cf. Candrakīrti's Madhyamakāvatāra 3.13). The Gelukba scheme would place such a person even lower. The second of the four parts of the path of preparation is called "peak" (rtse mo, mūrdhan) because it is the end of the period in which one can generate anger that will sever the roots of virtue. Also, Tsongkhapa makes a distinction between bodhisattvas whose faculties are sharp and those whose are not. The former are the sort who needed to convince themselves that the Buddha's teaching on emptiness was true, and therefore that Buddhahood was attainable, before they could make the extraordinary commitment to strive for countless aeons to free all sentient beings. Such persons might have attained a level of understanding equivalent to the path of preparation even before they generated bodhicitta, the effect of which would be to undermine (though not, of course, destroy) their predispositions to anger and desire. A later commentator, the fifteenth century Jaydz™n Cho/oogyi Gyeltsen (rje btsun chos kyi rgyal mtshan) who wrote the Mādhyamaka textbook--based on Illumination--still in use by Sera Jay (se ra rje) Gelukba monastery, went even further. He claimed that not only do intelligent persons realize emptiness prior to generating bodhicitta, but most dull ones do also (see Newland: 43).

(15) It may not be the case that all instances of anger cut the roots of virtue. As we will see, instances of anger may be differentiated on the basis of their recipients, but are there other factors that make one instance worse than another? Tenzin Gyatso, Dalai Lama XIV, says that it is still possible for someone past the path of seeing to experience anger; however, since root-cutting anger is no longer experienced past the second part of the even earlier path of preparation, it is clear that this anger would not impel lifetimes of suffering. The implication is that a higher bodhisattva's anger is not as serious, perhaps because to some extent its root cause, ignorance, has been undermined. Does this also mean that not all instances of anger would result in severance of the roots of virtue? Would Tsongkhapa agree with the Dalai Lama's conclusion?

(19) Pradhan: 35/Rahula: 78:58. Cited in Buswell: 119-20. According to Tsongkhapa's Gelukba order, the mind's emptiness of inherent existence is a "natural lineage" (rang bzhin gnas rigs) that is the Buddha nature of each sentient being, and hence there is no one who will fail, eventually, to attain Buddhahood. (For a review of reasons why some of Tsongkhapa's followers found difficulties with these doctrines, see Lopez.) They interpret Asa"nga to mean that he sees five lineages (rigs, gotra) for sentient beings, respectively those who follow the path of the "Srāvakas, Pratyekabuddhas, and Bodhisattvas, those who switch from one of the former to the latter, and those without a lineage for liberation.

(23) This paragraph is not Tsongkhapa's example, but my own, which I include because it seems parallel to the example he furnishes. As I point out below, Tsongkhapa wants to distinguish between the temporary suppression of fruitions by the ripening of other, contrary, karmas, and the incapacitation of fruitions by anger. That is, anger is qualitatively different from most other nonvirtues. That is why I think that he himself would not use this as an example. However, there seems to be no difference between the practical effects of these nonvirtues.

(26) Tsongkhapa and his followers consistently interpret the "not other than" statements in Indian Mādhyamaka as meaning "not inherently other" since, of course, things such as seeds and sprouts are different from each other. On the other hand, they are individually not inherently existent (rang bzhin gyis grub pa, svabhāvasiddhi) and do not have a relationship of inherent otherness--a relationship that is not merely imputed by thought.

(28) Prāpti and other means to account for the continuation of karma, such as the ālayavijñāna of Yogacārā texts, are rejected by Tsongkhapa as entities not included in the conventions of the world (which he thinks are, in contrast, upheld by sūtras of definitive meaning and in the ultimate commentarial tradition of Prāsa"ngika-Mādhyamika), not to mention the fact that as described by their proponents they could be established only by ultimate analysis. This is a major topic of the "unique tenets of Prāsa"ngika" section of Jamyang Shayba's Great Exposition of Tenets (grub mtha' chen mo), which I translated as part of my dissertation. For a recent discussion of Vaibhāṣika positions and how they are critiqued by Vasubandhu's Abhidharmako"sa-bhāṣya, see Hayes.

(29) Of course, it also raises the question of how wisdom could destroy seeds. This is reminiscent of a discussion by Jamyang Shayba (Great Exposition of the Middle Way: 628.3-.5) on the "disintegratedness" of the obstructions to omniscience (shes sgrib, jñeyāvaraṇa) for Buddhas. To become Buddhas, of course, necessitated the destruction of those obstructions, but Jamyang Shayba, wishing to avoid saying that Buddhas have anything like a taint in their continūms, maintains that the disintegratedness of obstructions to omniscience does not exist. His reasoning: in order to be a functioning entity, something must be capable of producing an effect, and this disintegratedness cannot. Instead, the obstructions to omniscience are completely "extinguished into the dharmadhātu." I have discussed arguments for and against Jamyang Shayba's position in my dissertation.