Talking about Nirvana

Thanissaro Bhikkhu

There are passages in the Pali Canon where the Buddha states that two things cannot be properly described: unbinding (nibbāna/nirvāṇa) and the post-mortem status of the arahant, a person who has attained total unbinding. However, even though the Buddha is consistent in never describing the post-mortem status of the arahant, there are passages where he does make concessions to his listeners to describe unbinding. The question is, why the distinction?

On the surface, this would appear to be an academic or an esoteric issue, but the Buddha’s apparent reasons for making those concessions are strategic, dealing with the basic purpose of his teachings and the path for getting there. For this reason alone, they’re worth looking into. And, as it happens, there’s a long history of people who—missing the distinction between the Buddha’s two different approaches to these closely related topics—have arrived at some serious misinterpretations of the goal of the practice.

For instance, Nāgārjuna, in chapter 25 of his Mūlamadhyamaka-kārikā (MMK), presents this argument for claiming that there is nothing to distinguish saṁsāra from nirvāṇa:

The statements in verses 17 and 18 seem to be drawn from SN 22:85–86, or versions of those suttas in other early Canons, in which the Buddha argues that because the Blessed One can’t be defined even in this lifetime (“currently”), he can’t be described as existing, not existing, both, or neither, after death. Now, because the MMK is a shorthand summary of arguments, we don’t know the full details of how Nāgārjuna would have fleshed out the logical leap from the statements in verses 17 and 18 to the statement in verse 19. Still, it’s obvious that his argument would have required conflating two things: the Buddha’s treatment of the status of the arahant (in verses 17 and 18) on the one hand, and the status of nibbāna (in verse 19) on the other. From this conflation, many practical problems arise, the foremost being that if there’s nothing to distinguish nibbāna from saṁsāra, how is a meditator to know whether he or she has attained nibbāna?

Nāgārjuna isn’t the only writer to conflate these two issues, or to create misunderstanding based on that conflation. Other thinkers, from a wide variety of Buddhist schools, have made a similar mistake up to the present day. For instance, I have read several modern Theravādins say that given the Buddha’s warnings against perceiving the arahant as existing eternally, we should be careful not to see nibbāna as an unchanging state—even though that’s one of the ways the Buddha himself describes it. If nibbāna weren’t unchanging, it wouldn’t qualify as a proper goal for what the Buddha called the noble search: the search for what is free from aging, illness, and death; sorrow, lamentation, and despair (MN 26).

So it seems worthwhile to go back to the original sources in the Pali Canon to see how the Buddha treated these two topics, and why he treated them in two different ways. That way, we can ferret out what useful lessons could be learned by keeping the Buddha’s treatment of these two topics separate.

A series of verses in the Pali Canon—Upasīva’s questions to the Buddha, recorded in Sn 5:6—indicate that both unbinding (nibbāna) and the arahant who is totally unbound can’t properly be classified. In other words, they can’t properly be described in terms of any of the categories of language or thought.

In the following excerpt from those verses, the Buddha first states that unbinding—the goal (attha)—cannot be classified. Upasīva’s question then conflates the two issues: the nature of the goal as a dhamma—a phenomenon, object, event, or action—and the question of whether a person who has reached the goal exists or not. The Buddha doesn’t bother in this case to explicitly separate the two issues. He simply notes that arahants have no criterion (pamāṇa: limit, measure) by which anyone could say that they exist or not. Then he explains why: Unbinding is actually not a dhamma. It is the doing away with all dhammas, so there is no means by which either the arahant or unbinding could properly be identified or described.

When we look elsewhere in the Canon, we find other passages that convey a similar message about the indescribability both of unbinding and of the arahant. They also indicate that the desire to impose a description either on unbinding or on the existence of an arahant comes from patterns of thinking that have to be abandoned if you are to attain unbinding.

For example, here’s a fragment from a dialogue on the “non-objectified” (apapañca), a synonym for unbinding:

Objectification is a type of thought based on perceptions and classifications derived from the assumption, “I am the thinker” (Sn 4:14). This type of thinking, the Buddha states in many places, leads to conflict and, if clung to, gets in the way of unbinding. So what Ven. Sāriputta is saying here is that the desire to classify and describe unbinding comes from a type of thinking that has to be abandoned for the sake of attaining the goal.

There are other, similar reasons for refusing to describe the arahant after death. A recurring theme in the Canon concerns the Buddha’s refusal to take a stand on the question of whether the arahant, after death, could be described as existing, not existing, both, or neither (DN 9; MN 63; SN 44). In the following passage, Ven. Sāriputta explains that one of the reasons for the Buddha’s refusal is that the questions themselves come from a wrong kind of thinking:

Other suttas give similar reasons for the Buddha’s consistent refusal to take a stand on this issue. SN 44:4, for instance, states that the questions of the arahant’s existence, etc., occur only to those who don’t know and see the aggregates as they have come to be; SN 44:6, that the questions occur only to those with love for the aggregates, becoming, clinging, and craving; SN 44:8, that they occur to those who assume about any of the aggregates that “This is me, this is my self, this is what I am.”

So the questions in both cases—the question seeking a description of unbinding and the question seeking a description of the status of the arahant—try to apply classifications of thought that don’t properly apply in either case. At the same time, they come from states of mind that cause suffering and—if clung to—would get in the way of reaching unbinding. So there are both formal reasons—related to definitions and ways of reasoning—and strategic reasons for refusing to answer these questions.

But when we look at other passages in the Canon, we find that its strategies in treating these two topics—unbinding and the status of the arahant—diverge. On the one hand, the Buddha and his arahant disciples consistently refuse to describe the status of the arahant after death. In fact, this refusal is so insistent that in SN 22:86 the Buddha points out that the arahant can’t be properly defined even when alive. And in the same sutta, the Buddha rejects the idea that the arahant after death could be described not only in any of the four ways—collectively called the tetralemma, because it covers four alternatives—but also in any other way at all.

On the other hand, though, there are times when the Buddha and his arahant disciples do give descriptions of unbinding. Ud 8:1 and Ud 8:2 state unequivocally that unbinding exists. Ud 8:1 also describes it as a dimension or sphere (āyatana). Other passages describe it as a state (pada—Iti 43–44; Dhp 228; ṭhāna—SN 6:1) and as a property or element (dhātu—Iti 44; Ud 5:5 = AN 8:19).

It has been argued that, in the context of the philosophical language of the Buddha’s time, the verb for “exists” used in Ud 8:1 and Ud 8:2—atthi—never denotes ordinary existence; instead, it always denotes permanent existence, but that’s not the case. In DN 1, it’s used in theories that posit a self that perishes at death; in SN 22:94, the Buddha himself uses it when stating that aggregates that are inconstant, stressful, and subject to change do exist.

But even though the verb doesn’t necessarily imply permanent existence, many of the epithets that the Buddha applies to unbinding do state clearly that it is not subject to change. Ud 8:10 and Dhp 228 call it unwavering (acala). Iti 43 and SN 43 call it dhuva, an adjective that can mean “firm,” “constant,” “permanent,” and “lasting.” SN 43 and Ud 8:2 call it unfabricated, and AN 3:48 states that the “unfabricated characteristics of the unfabricated” are that “no arising is discernible, no passing away is discernible, no alteration while staying is discernible.” MN 29 and MN 30 call it “unprovoked (akuppa),” a term with a technical meaning in the Buddha’s time, related to the theory of dhātu, or element. The theory described physical and mental events as the result of the provocation of an underlying property. When the provocation ended, the event would end as well. To say that the release of unbinding is unprovoked, then, would mean that because it was not the result of a provocation, there would be no reason for it to end.

Ud 8:1 goes into even greater detail about the unchanging nature of unbinding:

This lack of the activities that define time—coming, going, staying—is echoed in Ud 8:4, a passage repeated in MN 144 and SN 35:87, which states that unbinding also lacks the coordinates that define space:

The fact that unbinding lacks any of the features that participate in space and time may explain the fact that even though the Canon often makes the point that unbinding is unchanging, it never once describes unbinding as “eternal.” After all, eternity is a measure of time, and so it would not properly apply to anything outside of space and time.

There are, however, many other terms that the Buddha does use to describe unbinding. SN 43, for instance, contains a long list of synonyms and epithets for unbinding that can be sorted into four categories: those denoting (1) lack of change (“unaging,” “undecaying,” “deathless”), (2) bliss, safety, and peace (“bliss,” “island,” “harbor,” “shelter,” “refuge”), (3) freedom (“unbinding,” “release,” “attachment-free”), and (4) excellence and transcendence (“non-objectification,” “the amazing,” “the astounding,” “the exquisite,” “the ultimate,” “the beyond”).

Some of the Buddha’s descriptions of unbinding are paradoxical. For instance, Iti 90 states that the realization of nibbāna is the highest of all dhammas, whether fabricated or not. But Sn 5:6 and AN 10:58 describe unbinding as the ending or abandoning of all dhammas. These latter passages are seconded by Sn 4:6 and Sn 4:10, which state that the arahant has gone beyond not only passion but also dispassion, which Iti 90 equates with the realization of unbinding. So apparently, unbinding is experienced as a dhamma when it is first realized—and can even become an object of clinging on its first realization, in which case awakening remains incomplete (MN 52; AN 9:36). But if awakening is complete, then once the moment of realization is past, unbinding transcends all dhammas.

Another paradox concerns the nature of the happiness of unbinding. Dhp 203, seconded by DN 9, describes unbinding as the foremost bliss, whereas Ud 1:10 states that it is beyond both bliss and pain. MN 59 suggests a way of understanding this paradox: It is possible to experience a type of bliss that is not a feeling of bliss. After all, feelings of bliss are like all feelings: inconstant and stressful, whereas the bliss of unbinding does not change. So unbinding is beyond feelings of bliss, but it is characterized as bliss of a different, undefined sort.

Yet another paradox concerns how unbinding is known. AN 9:34 states that nothing is felt (vedayita) in unbinding, whereas SN 35:117 states that the dimension where the six senses and their objects cease is to be experienced (veditabba). Vedayita and veditabba are two forms of the same verb, vedeti, which means to feel, experience, know, or understand. Because the duty with regard to the cessation of suffering is to realize or verify it, the proper translations of these terms seems to tend more toward “feeling” and “experiencing” than to “understanding.” But that still leaves us with the paradox: There is nothing felt in unbinding, and yet unbinding itself is to be experienced.

SN 12:64 contains an image in a simile that helps to clarify the paradox: a sunbeam going through space and not landing—literally, “not becoming established”—on the surface of any object. The sunbeam is bright in and of itself, but it illuminates nothing. In SN 12:64, this image is used as a simile for unestablished consciousness, and it is apparently also an image for “consciousness without surface” (viññāṇaṁ anidassanaṁ), a term found in MN 49 and DN 11. In both cases, the consciousness is classed as a type of consciousness, but it has no object, not even itself. (If it took itself as its object, it would simply be consciousness as an object of concentration, as in the formless absorptions.) Because MN 49 states that this consciousness is not known through the All (a term for the six senses—see SN 35:23), it is not the same thing as consciousness in dependent co-arising. Because it’s not involved in the dimensions of space or time, it would not rank as consciousness in the aggregates. Both MN 49 and DN 11 state that it is endless and radiant all around. DN 11 adds that it is where the four physical properties have no footing, and where name and form are brought to an end—another indication that this is not simply a reference to consciousness in the formless attainments, which are “name” attainments. But beyond that, the Buddha provides no further explanations of consciousness without surface. Given its paradoxical nature, it would be hard to explain.

It’s been argued that consciousness without surface is not an intrinsic part of unbinding, that it’s simply the arahant’s meditative consciousness of unbinding in this life. But given that this consciousness is independent of the six senses, and that at the death of the arahant the six senses simply grow cold (Iti 44), then such an event should have no effect on it.

At the same time, if unbinding were the ending of all consciousness of every sort, the image of the sunbeam landing nowhere would never have occurred to the Buddha as a useful analogy for unestablished consciousness. Instead, he would have used an image in which the sunbeam—or something else—was annihilated. This suggests that when Ud 8:1 describes unbinding as unestablished and without support, and SN 43 describes it as surfaceless, they are referring to this unestablished, surfaceless consciousness. It’s because of this consciousness that when the Buddha, in MN 38, proclaims the conditioned, changing nature of consciousness, he is careful to limit his statement to dependently co-arisen consciousness. In that way, he exempts this unestablished, surfaceless consciousness from the discussion in that discourse.

So this paradoxical awareness, without object or support, is another feature of the goal.

These are just a few of the descriptions of unbinding found in the Canon. The question is, why did the Buddha refuse to describe the arahant after death and yet give so many descriptions of unbinding? He himself never says, but a number of answers can be inferred from what the Canon itself says about definition and description in general. Those answers fall into two sorts:

The formal reasons are quite straightforward. As SN 23:2 indicates, people are defined as beings when they are caught up on desire, passion, delight, or craving for any of the aggregates. Because arahants are not caught up on these things in any way, they are undefined. The nature of the existence of something undefined cannot be described. This is why, when talking about arahants, the Buddha limits himself to saying what they have done, and not what they are: “arahants, whose effluents are ended, who have reached fulfillment, done the task, laid down the burden, attained the true goal, laid to waste the fetter of becoming, and who are released through right gnosis” (MN 118). When asked what kind of being he might be, the Buddha refused to say even that he was a human being, and recommended instead that he be described simply with a past participle: “awakened (buddha)” (AN 4:36). Because arahants are free from the attachments that would define them and make them describable as beings, the Buddha consistently refused to describe the nature of the arahant’s existence after death, or even when alive, in any way at all.

This formal consideration applying to beings, however, does not apply to states. They are identified, not by their attachments, but by their features. And even though the unfabricated, in its essence, may not be fully describable, it can be differentiated from the fabricated in that it lacks fabricated features. So—even if language is not fully adequate to the task—the unfabricated can at least be described with some legitimacy as being free from x, safe from x, transcending x, or as the ending of x. This means that the Buddha, when making concessions to the objectifying minds of his listeners in describing the features of unbinding as a state, is not stepping outside the boundaries of legitimate ways of speaking. This would not be the only time when he used objectifying language to get his listeners to abandon objectification and to attain an unobjectified goal. See SN 15:13 for a particularly vivid way in which he did this.

And there are strategic reasons for why he would use this kind of language with regard to unbinding. The first is to motivate his listeners to want to follow the path leading there. That path—the noble eightfold path—is a fabricated dhamma (Iti 90). All dhammas are rooted in desire (AN 10:58), and all fabrications are fabricated “for the sake of” something (SN 22:79). People thinking in terms of objectification need a concept of the goal so that they can decide whether it’s “worth it for me” to fabricate the path for the sake of that goal. This is why desire is included in one of the factors of the path—right effort (SN 45:8)—and why concentration based on desire is one of the factors leading to the goal of the end of desire (SN 51:15), as explained in Ven. Ānanda’s image:

So, in describing unbinding as a state of awareness that is unchanging, free, blissful, and excellent, the Buddha is giving his listeners reasons for wanting to follow the path there: to escape the stress and suffering that comes as part and parcel of all things subject to aging, illness, and death (MN 26).

He is also giving them guidance in recognizing whether an experience does or does not count as unbinding. This is his second strategic reason for describing unbinding to his listeners. Without such a description, how would they know whether they had arrived there, or if more work needed to be done? This strategic reason explains the first two of the three perceptions that he applies to all fabricated phenomena: If something is inconstant and stressful, it still falls short of the deathless. As for the third perception, not-self—which applies to all phenomena, fabricated or not: If you feel passion or delight for the phenomenon of the deathless, your awakening is still incomplete, because there is still clinging in the mind. Only when you can perceive the deathless as not appropriately labeled as “me” or “mine”—i.e., as not-self—can you go beyond all clinging to the aggregates, perceptions included (MN 52; AN 9:36). If unbinding were not unchanging and blissful, the first two perceptions would not work as means of judging and letting go of everything that is not unbinding; if unbinding were not attachment-free, neither would the third.

But even though the Buddha had these strategic reasons for describing unbinding to his listeners, he also had strategic reasons for holding to his formal reasons for not describing the arahant after death.

To begin with, if you could describe what arahants are, that would be an attempt to place a measure or a limitation (pamāṇa) on them.

Because arahants are free of obsessions, they cannot be measured, and are free from limitations. This is the reason behind the Buddha’s statement in Sn 5:6 that the arahant has no pamāṇa by which he can be described, and it appears to be the primary reason for the Canon’s repeated refusal to describe what an arahant is. Freed from being a being, the arahant is totally without limit. When, in a concession to the objectifying mind, the Canon uses similes to explain its refusal to describe the arahant, those similes all make the point that the arahant is unlimited and immeasurable: In SN 44:1, Sister Khemā, after rejecting the tetralemma, compares the arahant to the ocean in being incalculable. The Buddha, after rejecting the tetralemma in MN 72, compares the arahant to the ocean in being deep, boundless, and hard to fathom. As with the simile of the sunbeam, we can note that the simile of the ocean would not have occurred either to Sister Khemā or to the Buddha if unbinding were a nothingness or an annihilation. But we still have to be careful to note that it also indicates that the arahant is not limited to being a something.

A second strategic reason for not describing the arahant after death centers on the fact that the question of the arahant’s existence, non-existence (or both or neither) after death is a disguised version of the question as to whether there is or is not a self. If you were to say that the arahant exists, then whatever existed would be his/her self. If you were practicing to become an arahant, such a belief would prevent you from examining all attachments, for you would operate with an underlying assumption that there would be some sort of kernel of a self that would be worth craving, clinging to, and protecting. This craving would lead to attachment, possessiveness, and defensiveness (DN 15). This is why the Buddha rejects eternalism as a particularly pernicious wrong view.

As for the assumption that there is no self, MN 109 shows that if you assume that there is no self, there is no agent responsible for actions, and no one to be affected by actions. This would entirely vitiate the teaching on kamma, which is essential to undertaking the path to the end of suffering (SN 35:145). This would also turn the teaching into a doctrine of non-action (DN 1; AN 3:62). As AN 3:62 states, such a doctrine would undercut any grounds for deciding what should and shouldn’t be done, and so would leave its listeners bewildered and unprotected. This is why the Buddha also rejects annihilationism as another especially pernicious wrong view.

These considerations also help to explain a cryptic passage in SN 44:10 where the Buddha explains why he doesn’t answer the question of whether there is or is not a self (see also MN 2): To say that there’s a self, he says, is to be in company with the eternalists; to say that there is no self is to be in company with the annihilationists. The reason the passage is cryptic is that DN 1 quotes annihilationists who posit the existence of a self, but then maintain that it perishes at death. So, strictly speaking, simply to say that a self exists is not the same thing as espousing eternalism. One possible way of explaining the passage in SN 44:10 in formal terms would be to hold that the verb for “exist” used here—atthi—when used in philosophical contexts, must mean permanent existence. But, as we have already noted above, that’s not the case. And there’s nothing in the passage to indicate that the Buddha’s interlocutor was necessarily asking about a permanent self to begin with.

However, if we combine a formal point with the strategic points made above, we can make sense of the passage. The formal point is that the Buddha does not maintain that saying, “There is a self” is to assert an eternalist doctrine. He simply asserts that saying, “There is a self” is to be in company with the eternalists. The strategic point is that to hold there is a self, no matter how you define it, gives rise to the same defilements as an eternalist view. In other words, the companionship with the eternalists here, even though it’s not formal, is strategic: If, instead of seeing your sense of self simply as an assumption, you identify something as your self in this lifetime, then even if you think that it will perish at death, you’ll believe that that self has to be clung to and protected at all costs. In fact, if you believe that your self will perish, you tend to cling to it and protect it even more fiercely—a tendency that would get in the way of awakening, and one whose detrimental consequences on a gross level we see all around us.

The question sometimes arises: In teaching an unchanging unbinding, isn’t the Canon itself guilty of engaging in eternalism? Isn’t it espousing a pernicious wrong view? The answer is No, and here again there are both formal and strategic reasons for why not.

The formal reason is that eternity is a function of time—a long, unending time. Unbinding, however, lies outside of the confines of space and time entirely, and so the adjective “eternal” doesn’t apply.

The strategic reason is that there is nothing inherently wrong with the idea that something is unchanging. The problem with the wrong view of eternalism is what, precisely, it identifies as unchanging. All of the eternalist views quoted in the Canon—in DN 1 and SN 22:81—posit only two things as eternal: an eternal world and/or an eternal self. For the purpose of putting an end to suffering, though, the cosmos and all assumptions of self have to be seen as fabricated, dependently co-arisen phenomena. Only then can any passion for them be abandoned. To say that they are unchanging and eternal would be to say that they are unfabricated. The perceptions of inconstancy and stress would not apply to them, and so there would be no reason to develop dispassion for them. This would stand in the way of getting beyond them to attain the truly unfabricated goal of unbinding. This is another reason why these eternalist views are pernicious. Unbinding, however, is neither a self nor a world. Thus a belief that unbinding is unchanging would not get in the way of the path. In fact, as we have seen, it’s necessary to perceive unbinding as unchanging in order to be motivated to give up changing pleasures for its sake, and to recognize it when it is attained.

A parallel can be drawn with annihilationism. There is nothing inherently wrong with the idea that something is annihilated. After all, the Buddha himself says that he is an annihilationist when it comes to advocating the annihilation of passion, aversion, and delusion, along with the annihilation of evil, unskillful qualities (AN 8:12). Annihilationism becomes a pernicious wrong view only when referring to the annihilation of an “existing being”—which the examples given in DN 1 show to mean the annihilation of the self, however it is defined. As we have seen, the implications of this sort of annihilationism close the door to the practice.

Now it is true—as mentioned above in connection with MN 52 and AN 9:36—that it’s possible to develop a passion for the deathless on first encountering it, but that passion doesn’t prevent the realization of any of the first three stages of awakening. It simply gets in the way of total awakening. And the teaching that all phenomena—fabricated and unfabricated—should be seen as not-self has been provided to counteract that passion and to open the way to full unbinding.

But a preconceived notion that unbinding means the end of existence would get in the way of even the first level of awakening. As DN 1 and DN 15 point out, there is a level of the cosmos where the beings are totally without perception, and basically unconscious. A person thinking that unbinding is the total end of any and all consciousness could easily mistake that level of the cosmos for unbinding, and—believing it to be the goal—would be stuck there in a dead end for a very long time.

For this reason, the dangers of explaining unbinding as unchanging are far less than the dangers of explaining it as an annihilation—and far less than the danger of not explaining it at all. So even though, as Sn 5:6 indicates, neither unbinding nor the person who has reached unbinding can be properly explained by the categories of language, the Buddha did have good reasons for sketching a description of unbinding in terms that would motivate his listeners to practice so as to attain it, and would help them recognize it once it was attained.

As he explained in MN 29–30, the unprovoked release of the mind is the essence—sāra, literally “heartwood”—of the teaching. And as he further explained in Dhp 11–12,

So it’s for the sake of right resolve—and for the ultimate sake, that his listeners will practice right resolve to the point of reaching the essence—that he was kind enough to bend language in the service of leaving behind descriptions of unbinding as a desirable goal.

Even though we must always keep in mind the fact that descriptions of unbinding are a form of objectification, and for that reason should not be clung to as awakening nears, we can still use them as incentives to get on the path, confident that the goal is more than worth all the effort it requires.

• 
  
• 
  
• 
  
• 
  
•