The Limits of Description

Not-self Revisited

by Ṭhānissaro Bhikkhu

The Buddha once divided his teachings into two categories: those whose meaning has been fully drawn out, and those whose meaning should be inferred (AN 2:24). When dealing with a teaching in the first category, he said, trying to draw further implications from it would actually be an act of slandering him. When dealing with a teaching in the second category, not drawing out its further implications would be an act of slandering him. Unfortunately, the Buddha didn’t give examples of which teachings belonged to which category. Still, the simple fact that he distinguished between these two categories makes an important statement about his teachings as a whole: He wasn’t trying to set out a systematically consistent description of reality. If he had been, the existence of the first category—teachings that shouldn’t have inferences drawn from them—would have been an admission of failure: To try to prevent his listeners from exploring the implications of some of his statements would be an attempt to keep those listeners from seeing that they were inconsistent with the rest of the system.

But as the Buddha said on several occasions, the essence of all his teachings was to lead to release (MN 29MN 30). In other words, his words were never meant to be just descriptive. They were primarily performative: meant to be put to use to have an effect on the mind. In some cases, the proper effect was to be achieved by taking his words just as they were. In others, it was best achieved by exploring the implications of the words. But in no case were the words ends in themselves.

This point relates to the Buddha’s observations about the uses and limitations of language. One of the standard Canonical descriptions of how to ask about the meaning of an expression is “to what extent is this so?” In other words, “how far is this meant to be true?” This could be taken simply as an idiomatic expression with no deeper meaning, except that the realizations leading to release include “having directly known the extent of designation and the extent of the objects of designation, the extent of expression and the extent of the objects of expression, the extent of description and the extent of the objects of description, the extent of discernment and the extent of the objects of discernment” (DN 15). To see the extent of these things means to see both the limitations of language, descriptions, and definitions, and what lies beyond them: the unfabricated dimension of unbinding (nibbāna).

Even a stream-enterer—one who has had his/her first taste of awakening—is said to have seen the drawbacks of the faculty of discernment, which is equivalent to right view, and also the escape from it (SN 48:3). On the way to awakening, a person who applies the highest level of right view to the arising and passing away of contact at the senses is said to enter a mental state where even thoughts of “existence” and “non-existence” with reference to the world of the six senses don’t occur to the mind (SN 12:15). Having been through such an experience—and the resulting release—it’s hard to imagine that such a person would then give total, unlimited approval to statements about the existence or non-existence of anything in the world. Truths, even when true, have their limits. This is why the texts so often speak derisively of sectarians who defend a view saying, “Only this is true; anything otherwise is worthless” (AN 10:93).

The need ultimately to go beyond words and discernment comes from the fact that they are made of perceptions and thought-fabrications, which are aggregates coming under the term “name” in “name-and-form.” As with all aggregates, even the statements of right view, after they have done their work, have to be abandoned for the mind to gain release. But more than that: A perception may be true as far as it goes, but there are limitations to how far it can go, and as DN 15 indicates, there’s a need to see those limitations. In one passage, the Buddha goes to the extent of identifying only one thing as really true: unbinding.

“See the world, together with its devas,

supposing not-self to be self.

Entrenched in name-and-form,

they suppose that ‘This is true.’

In whatever terms they suppose it,

it turns into something other than that,

and that’s what’s false about it:


it’s deceptive by nature.

Undeceptive by nature

is unbinding:

That the noble ones know

as true.

They, through breaking through

to the truth,


are totally unbound.” (Sn 3:12)

This, however, doesn’t mean that the Buddha meant for his words only to be performative without trying to make them accurate as descriptions. He never dealt in “useful fictions.” As he said in MN 58, his words were always true, beneficial, and timely. In his analysis of what that meant, he gave no room to the possibility that any statement could be either beneficial or timely if it were false. But having seen what lies beyond language, and making the dimension beyond language the goal of his teaching, he must have been very sensitive to the limits of how far a statement could be true. This is why, as a teacher, his main concern was to use true statements in such a way that they would lead the listener to act in such a way as to lead to release. And this is why he would avoid answering questions on topics where statements of any kind would not lead in that direction. It’s possible to find at least 60 questions in the suttas that the Buddha or his arahant disciples put aside on the grounds that any attempt to answer them would actually get in the way of awakening (see Skill in Questions, chapter eight).

So when interpreting the Buddha’s teachings, it’s important not to fall into the scholarly bias that tries to capture the views of an awakened person in the net of its language. This applies both to attempts to draw implications from his words to answer questions that he put aside—which, as AN 2:24 notes, would be akin to slandering him—and to attempts to depict the practice as a process of leading the meditator simply to give full assent to the accuracy of the Buddha’s teachings as a description of reality. To capture the practice in a net of words in these ways is to miss the meaning and purpose of the Buddha’s teachings entirely.

These reflections were sparked recently by reading a critique of an article I wrote in 1993, called “The Not-self Strategy.” The thesis of that article (available in the essay collection Noble & True)—which I revised in 2013 both to tighten and to expand the presentation—was that the Buddha intended his teaching on not-self (anattā), not as an answer to the metaphysical/ontological question, “Is there a self?” but as a strategy for cutting through clinging to the five aggregates and so to put an end to suffering. The main argument I presented in support of this thesis in both versions of the article was that the one time the Buddha was asked point-blank, “Is there a self?”… “Is there no self?” he remained silent (SN 44:10). Similarly, in MN 2, he stated that such questions as “Do I exist?” “Do I not exist?” and “What am I?” are not worthy of attention because they lead to conclusions that fetter a person in a “thicket of views” and a “fetter of views,” including the views that “I have a self” and “I have no self.” In other words, any attempt to answer these questions constituted a side road away from the path of right practice.

The critique—“Anattā as Strategy and Ontology,” written by Ven. Bhikkhu Bodhi—was brought to my attention just over a month ago, even though it has apparently been around for some time. It takes issue both with the thesis and with the argument of my article, but in doing so it displays the scholarly bias mentioned above: that the practice of the Buddha’s teachings is primarily a process of leading the meditator to give full assent to the accuracy of those teachings as a description of reality, and that this assent is what frees the mind from suffering. Because this bias is not only the bias of the critique, but of so much thought in the Buddhist world, I thought it might be useful to explore how both the thesis of the critique and the arguments used in support of that thesis display this bias, so that it can be recognized for what it is not only in this case but also in other Buddhist writings.

For ease of reference, I will state the critique’s main points in a numbered format. In the response to those points, I will avoid—except where necessary—repeating arguments already made in “The Not-self Strategy.” If you are interested in the full argument presented in that article, I recommend that you read the 2013 version.

The basic thesis of the critique is actually an ancient one, with a long history in the Buddhist philosophical tradition. It can be summarized in the form of a syllogism:

1. For the Buddha, the term “self” has to mean a substantial, permanent, unchanging essence.

2. Personal identity—what you really are—is composed of conditioned elements that are constantly changing.

Therefore: 3. There is no self.

The critique admits that the Buddha never assented to the statement “There is no self,” but maintains that he had two pragmatic reasons for not directly stating this truth that is implicit in his teachings.

4. The first is that, because the view of an underlying substantial self is so deeply ingrained in the unawakened mind, the simple, direct statement that there is no self would not uproot it. Instead, the meditator would come to that conclusion only through the indirect means of examining each element of his/her personal identity to see that none of those elements were permanent in themselves or had an essential relationship to anything permanent.

5. The second reason is that the annihilationists—sectarians who argued that death was the end of consciousness and personal identity—also taught that there is no self, so to simply state this truth might mislead people into thinking that the Buddha was siding with the annihilationists.

Despite the potential drawback cited in point five, the critique argues that, with proper explanation, it can be avoided, and that there is still practical value in stating the abstract principle lying implicitly behind the Buddha’s indirect approach for three reasons:

6. The fact that there is no self is what makes the teaching on not-self work as a strategy.

7. The attainment of stream-entry is what frees the meditator from the mistaken belief that there is an unchanging core to personal identity.

8. Therefore, to help a person aiming at stream-entry, it is important to state that the not-self teaching is not only a strategy but also a statement of an ontological truth: There is no self.

9. Finally, the author asserts that the not-self teaching cannot be said to have only a strategic purpose because the right view that there is no permanent self is not just a factor of the path for those in training, but is also an inalienable endowment for the arahant.

The critique cites several passages from the Pāli Canon to support these points. However, when we examine these points both on their own merits and in relationship to the passages meant to support them, we find that the scholarly bias behind them turns the Buddha’s teachings into the thicket of views that the Buddha expressly warned against entering.

By Definition

1. The first point in the critique’s thesis makes its case through a definition: One’s identity has to be permanent to count as a “self.” The Buddha, however, never defined “self” in this way. Before looking at the critique’s textual argument for inferring this definition from a passage in the Canon, it’s worth looking at the historical and practical reasons for calling the inference into question.

a. Historical. There is a popular belief, promoted by many scholars, that the Buddha formulated his not-self teaching primarily in response to the Upaniṣadic doctrine of a permanent, unchanging self, identical with the ground of being for the cosmos. Thus it is only natural that “self,” in the time of the Buddha, meant a permanent unchanging essence lying at the core of one’s identity.

However, this belief misses two important facts.

The first is that the Pāli Canon cites a wide variety of beliefs about the self current in the Buddha’s time, and many of them proposed a self that was finite—i.e., it comes to an end—and subject to change. DN 15 provides a framework for classifying the different possible views about self, starting with four types of self: possessed of form and finite, possessed of form and infinite, formless and finite, and formless and infinite. Further, beliefs about each of these four types state that the self is either already that way, or that it naturally becomes that way (for instance, at death or when falling asleep), or that it can be made to become that way (through practice of one sort or another). Combining these two lists gives altogether 12 types of self-doctrines, only two of which teach an unchanging self: the self already possessed of form and infinite, and the self already formless and infinite. In addition, DN 1 cites seven annihilationist views about the self—three defining the self as possessed of form, four defining it as formless—that perished at death.

Moreover, there are two instances where the Buddha, when mentioning the view of a permanent, unchanging self, identical with the cosmos (MN 2; SN 22:81), mentions it alongside other views of the self, implying that it is simply a particular instance of self-view, and not the only one he is trying to refute. In MN 2, he mentions it as a special case of the view, “I have a self.” In SN 22:81, he mentions it as an additional case after discussing twenty ways in which a self-view can be constructed around the five aggregates.

So it’s clear that the permanent, unchanging self mentioned in some of the Upaniṣads (such as Bṛhad-āraṇyaka I.4.7–10 and Chāndogya III.14.2–3) was not the only self-view the Buddha was addressing with his not-self teaching.

The second fact missed by the popular belief about the primacy of the Upaniṣadic view of the self at the Buddha’s time is that the major Upaniṣads are not unanimous in the ways they define the self. It’s impossible to know whether all of these Upaniṣads existed at the time of the Buddha, but it’s enlightening to note that the major ones offer a variety of doctrines of the self that fall into at least eight, and perhaps nine, of the categories listed in DN 15, including doctrines that describe the self as already possessed of form and finite (Bṛhad-āraṇyaka II.5.1; Maitrī VI.11) and naturally becoming possessed of form and finite (Bṛhad-āraṇyaka IV.3.19–21). In other words, even some of the Upaniṣads taught the self was impermanent. So it might have been the case that the Buddha derived many of the categories of his framework in DN 15 at least partly from the wide variety of self-views in the Upaniṣads.

So the idea of a permanent self did not have a monopoly in the time of the Buddha. This means that if he were going to insist arbitrarily, as a crucial assumption, that a self had to be permanent to qualify as a self, he would have had to present a case to defend that definition. But he never did. So it’s unlikely that this assumption should be inferred from his teachings.

b. Practical. One of the fetters abandoned at stream-entry is the fetter of identity views. The Canon shows that these views relate to various ways of conceiving the self in relation to the five aggregates. However, to restrict the definition of “self” in this case to a permanent, unchanging essence raises many practical questions: Why would a belief in a permanent self be any more of a fetter than a belief in an impermanent self? And if the annihilationists, as DN 1 shows, believe in an impermanent self, does that mean that they have already dropped the fetter of identity views? If so, why does the Buddha single them out as holding to a view that is particularly evil (pāpaka)? Practical experience shows that people who define their body as their self, knowing that it will perish at death, are even more attached to it, and will do worse things to ensure its survival, than do people who believe that the self survives death. And the recent embrace of gender fluidity has shown that people will cling just as firmly to the fluidity of an identity they know to be fluid as they will to an identity that they think is permanent.

So to insist that a self-view has to posit a permanent self in order to be a fetter makes no practical sense.

c. Textual. The passage the critique uses to infer that the Buddha assumed implicitly that the term “self” had to mean a permanent, unchanging essence comes from MN 148. In this passage, the Buddha argues that it’s not tenable to view the senses, their objects—along with consciousness, contact, feeling, and craving based on the senses and their objects—as self. The reasoning in each case follows the same pattern, and can be illustrated with the Buddha’s argument focused on the first sense, the eye:

“If anyone were to say, ‘The eye is the self,’ that wouldn’t be tenable. The arising and falling away of the eye are discerned. And when its arising and falling away are discerned, it would follow that ‘My self arises and falls away.’ That’s why it wouldn’t be tenable if anyone were to say, ‘The eye is the self.’ So the eye is not-self.”

Although it might be possible to infer from this passage that the Buddha assumes that self must be defined as something permanent—not subject to arising and falling away—the above-mentioned difficulties that would follow from this inference suggest that there must be a better way to construe the Buddha’s reasoning here. And there is, one inherent in any idea of self: The self, whether permanent or not, can’t watch itself arise and pass away. To discern its arising, it would have to be there before its arising; to discern its passing way, it would have to survive its passing away. This means that whatever it’s discerning as arising and passing away can’t be the same thing that it is. Which means that that “whatever” isn’t its self.

This interpretation avoids the above difficulties of insisting that “self” has to mean a permanent, unchanging essence because it focuses on a fact inherent in every idea of self, and is not an arbitrary assumption with little practical value. It also allows for the fact that clinging to the idea of an impermanent self can be a fetter. So it’s a preferable way of interpreting this passage.

At the same time, this interpretation is in line with the meaning of the term, anicca, which the Buddha frequently used in connection with the teaching on not-self, and which is all too often translated as “impermanent.” Actually, the term is the negative form of nicca, or “constant.” To say that the aggregates are anicca is not to say that they don’t last forever, but that they’re inconstant: unreliable and fluctuating. Their arising and passing away is discernible in the present moment. This observable fact is what can lead to the value judgment that they are not worthy of regarding as self.

So there’s no basis in the Canon for supporting the first point of the critique, that the word “self” in the Buddha’s teachings has to mean a permanent unchanging essence.

A Distinction without a Difference

2. Given that “self” in the Buddha’s teachings doesn’t have to mean a permanent essence, it’s a mistake to distinguish—as the critique’s second premise does—between “self” and “constituents of personal identity”—i.e., what you are. “What you are” is the same thing as your self, regardless of whether that identity is permanent or not.

The fact that, for the Buddha, this would count as a distinction without a difference is shown by the questionnaire he frequently used to lead his listeners to the conclusion that the aggregates are not-self. With each aggregate, he would ask, “Is it constant or inconstant?” The answer: “Inconstant.” The questionnaire would then proceed as follows:

“And is that which is inconstant easeful or stressful?”

“Stressful, lord.”

“And is it fitting to regard what is inconstant, stressful, subject to change as: ‘This is mine. This is my self. This is what I am’? [or: ‘I am this’?]”

“No, lord.”

He would then conclude, taking form as an example:

“Every form is to be seen with right discernment as it has come to be: ‘This is not mine. This is not my self. This is not what I am.’ [or: ‘I am not this.’]” (SN 22:59)

Notice that the Buddha here treats these two sentences—‘This is my self’ and ‘This is what I am’ [or: ‘I am this’]—as equivalent. In other words, the five aggregates are not your self, nor are they what you are. The critique, however, wants to make a distinction here, saying that the five aggregates are what you are, even though they don’t qualify as a self, forcing a distinction where the Buddha doesn’t see one. From his point of view, to say that the five aggregates comprise your identity—i.e., what-you-are—even though you realize that they are impermanent, is the same thing as saying that they’re your self. This, of course, would go directly against the point he is trying to make with this questionnaire, that the aggregates are not worth identifying with in any way at all.

The fact that the Canon treats “This is my self” as equivalent to “I am this” is also shown by the ways it defines the fetter of identity view. MN 44 defines identity view as any one of 20 ways of assuming a self around the aggregates. SN 22:89 equates the assumption of a self around the aggregates with the assumption of “I am this” around those aggregates. (This is why the annihilationists who say that the self perishes at death (DN 1) are still fettered with identity view.) So, as far as the Canon is concerned, the critique’s distinction between “self” and “what you are” indicates no real difference at all. We find that this false distinction runs into even more serious problems when we consider what it would mean for the postmortem fate of the arahant. Because the aggregates end with the attaining of total unbinding at the death of the arahant, to say that the changing aggregates that constitute your personal identity through many lifetimes would end at death if you have attained unbinding would be a self-view falling under one of the twelve categories set out in DN 15: the self possessed of form that is not already finite, and does not naturally become that way on its own, but can be made to become that way through practice. Although this view is not identical with the annihilationist view that every self perishes at death, it does constitute a type of annihilationism when it comes to the death of an arahant: An arahant, whose identity consisted of the five aggregates prior to death, would no longer exist after death.

The Buddha, however, was always extremely careful to avoid the position that an arahant does not exist after death. In fact, SN 22:85 goes so far as to label it an “evil (pāpaka) viewpoint.” When presented with the fourfold question as to whether an awakened one—called a Tathāgata, meaning a Buddha or an arahant disciple—existed, didn’t exist, both existed and didn’t exist, or neither existed nor didn’t exist after death—he refused to agree to any of the alternatives. If he held the unspoken assumption that there really is no self, then he wouldn’t have had to take such pains to avoid taking a stand on the issue: The arahant, being composed of the five aggregates, simply would not exist after death. But because the Buddha was so careful not to take that position, and to even regard it as evil, shows that he did not view the five aggregates as constituting one’s identity, and did not hold to the unspoken assumption that there is no self.

So these are some of the textual inconsistencies that come from identifying the aggregates as the constituents of personal identity. They can all be avoided, however, by following the Buddha’s example:

a) by regarding the aggregates not as the constituents of your personal identity, but as the raw material from which, through the activities of ahaṅkāra and mamaṅkāra, “I-making” and “my-making,” you construct your identity; and

b) by at the same time paying no attention to the question of whether or not there is a self lying behind that activity.

In following the Buddha’s strategy here, we avoid not only the textual inconsistencies cited above, but also some very practical problems that would come from assuming either the existence or the non-existence of a self lying behind the activity of I-making and my-making. As the Buddha notes in SN 44:10, to assume that there is a self lying behind that activity would get in the way of applying the perception of not-self to all phenomena. You’d be continually looking for that self, and would protect it when you thought you had found it. That way, there would still be an area of experience subject to clinging—and subtle suffering—that would never get abandoned.

On the other hand, if you assumed that there was no self lying behind your I-making and my-making, you’d fall into either of the two extremes listed in Iti 49: Either you would fear that, with the ending of I-making and my-making, there would no longer be any you, and so you’d be afraid to put an end to your creation of a sense of self; or else, eager for the destruction of your I-made self, you’d fall into the extreme of craving for non-becoming. As the Buddha noted in MN 49 and SN 56:11, craving for non-becoming paradoxically leads to more becoming and its attendant suffering. So in either case, your assumptions about the existence or non-existence of a self would get in the way of release.

Iti 49 describes briefly the way out of this dilemma: seeing what has come to be as come to be—in other words, seeing what is actually occurring simply as actually occurring—and developing dispassion for it. SN 12:15, noted above, helps to explain how this works: By focusing on the origination and passing away of events at the six senses, the mind enters a state where thoughts of “existence” and “non-existence” don’t occur to it. In that state, questions of the existence or non-existence of the self also don’t arise, as you’re focused purely on the suffering inherent in whatever phenomena are originated and pass away. This makes it easier to let go of the desire fueling those phenomena with no concern for what this contemplation would do to the existence of a “self,” and in that way the mind can gain release.

As for the question of the status of the arahant after death, the Buddha notes in SN 23:2 that a “being” is defined by attachment to the aggregates. Where there is no attachment, no being can be located. And when no being can be located to define what it is, nothing can properly be said about it. This is why SN 22:85SN 22:86 make the point that, when you can’t even define a fully awakened person in the present life, how can you predicate anything about awakened people after they die?

Further Implications

3. Because the two premises of the syllogism lying at the heart of the criticism of “The Not-self Strategy” are false, the conclusion based on them is unfounded. In other words, it’s a mistake to attribute to the Buddha an unspoken assumption that there is no self. This means that the remaining points dependent on the syllogism also don’t follow. However, some important practical and interpretive lessons can be drawn from considering exactly where some of those points go wrong. Here I will focus only on the points that are useful to consider in this way: 5, 6, and 9.

5. Point five claims that the Buddha avoided saying that there is no self because it would have confused some of his listeners into thinking that he was siding with the annihilationist view that death is automatically annihilation.

On an immediate level, it’s hard not to be amazed at modern interpreters who think that, although the Buddha refused to state that there is no self for fear that this statement would cause confusion among his listeners, they can make this statement at present on his behalf without causing confusion among theirs.

However, that point aside, the critique bolsters its claim here with an assertion that has to be addressed. The assertion is this: When, in SN 44:10 and MN 2, the Buddha refuses to accept the statement that there is no self, his refusal can be explained because “there is no self” is an annihilationist thesis and he can’t consent to the consequences that the annihilationists draw from that thesis, that there is no conscious survival beyond the present life.

This interpretation is mistaken on two points.

• Although SN 44:10 does state that to say there is no self would be to conform with the annihilationists, MN 2 makes no mention of annihilationists or annihilationism at all. So there is no reason to assert that in that sutta he is rejecting the statement “I have no self” only because he wanted to avoid sounding like an annihilationist. As he says there, simply the view, “I have no self” gets one involved in a thicket of views. And the tangled history of Buddhist philosophy—ever since interpreters of the Dhamma began interpreting the not-self teaching as based on the assumption that there is no self—has borne this statement out.

• The Buddha had a systematic strategy for classifying questions into four types, as to whether they deserved a categorical answer, an analytical answer, whether they might first require cross-questioning the listener before answering, or whether they should be put aside and left unanswered. In both SN 44:10 and MN 2, he leaves the question of the existence of a self unanswered. If he had an analytical view of the non-existence of the self—that, for example, there is no permanent self, but that there is a continuum of personal identity that does not automatically end with death—he could have easily stated it. But he didn’t. He had it totally within his power to have said, “There is no categorical answer to that question,” his typical way of beginning a response to a question deserving an analytical answer. But he didn’t. If, in SN 44:10 he had wanted to state such an analytical position to Ven. Ānanda, who was present at the conversation and who surely would have understood him, he would have. But he didn’t. And, as noted above, in MN 2 he states in no uncertain terms that questions related to the existence or non-existence of the self aren’t worth paying attention to at all.

As noted in the introduction to this essay, to draw inferences from the Buddha’s teachings that would provide answers to questions he deliberately put aside—even if they are analytical answers—has to count as a form of slander as mentioned in AN 2:24.

Why Does the Not-self Strategy Work?

6. Point six raises a valuable question: Why does the not-self strategy work in liberating the mind from clinging? Rather than following the critique’s strategy of trying to find the answer to this question by inferring from the suttas a position that the Buddha refused to endorse—that there is no self—it’s more fruitful to look for the answer in the Buddha’s express statements about how and why clinging to a self-view happens in the first place. When we understand how self-view is fabricated, how clinging is fabricated around that, and why that clinging constitutes suffering, we can understand the Buddha’s strategy for bringing these fabrications to an end.

MN 44 notes that all self-identity views revolve around one or more of the five aggregates: form, feeling, perception, fabrication, and consciousness. In each case, the self can be defined as being identical with the aggregate, as possessing the aggregate, as being inside the aggregate, or as containing the aggregate within it. This gives twenty types of identity-view in all.

Acts of fabrication, i.e., intentional choices, play a many-layered role in shaping the aggregates and any of the identity views that cluster around them. As SN 22:79 notes, fabrication plays a role in fabricating each aggregate for a purpose. And as SN 22:81 further notes, the act of assuming a self around any of the aggregates is also a fabrication, based on craving, which in turn is based on ignorance. SN 22:60 identifies the purpose underlying both layers of fabrication: It’s because of the pleasure to be obtained from the aggregates that beings are infatuated with them. We fabricate both the aggregates and the sense of self around them in order to obtain that pleasure. In other words, the pleasure to be found in the aggregates is the root cause of why we desire them and cling to them, building a sense of self around them. Even without having to think that the aggregates are permanent, if we think that the pleasures that can be derived from them are worth the effort that goes into clinging to them, we’ll choose to cling.

All of this is in line with two observations from AN 10:58: That all phenomena are rooted in desire, and are brought into being through attention. In this case, the desire is for pleasure, and the act of attention is that of attending inappropriately to questions about the past, present, and future existence of the self (MN 2) in hopes that the answer will help realize our desire and maximize the pleasure.

The problem is that clinging to a self-view counts as a form of suffering as defined in the first noble truth (SN 56:11). So in our ignorant pursuit of pleasure, we end up constructing suffering instead. In most cases, this clinging entails suffering because it tries to latch on to things that will change (SN 22:1). But it’s also possible for the mind, on its first encounters with the deathless, to cling to that experience (MN 52; AN 9:36). Even though the deathless is not fabricated, and so is not subject to change, the act of clinging to it is fabricated, and so entails suffering nevertheless.

Because all of this clinging and fabrication is driven by desire, sparked by an ignorant value judgment—seeing that it’s worth the effort to fabricate aggregates and self-views for the sake of the pleasure—the strategy to undercut it has to replace it with a more accurate value judgment: That the pleasure is not really worth the effort at all.

This is where the not-self strategy comes in: to focus attention on how much effort actually goes into fabricating the aggregates and the self-views based around them, and on how the results don’t really repay the effort that goes into them. In other words, its purpose is to accentuate the fact of the effort required by fabrication and to raise the question of its value: whether it’s worth the effort to keep fabricating.

SN 22:57 outlines the general approach of this strategy in seven inter-related steps. The first four steps follow the pattern of the four noble truths: directly knowing each aggregate, directly knowing the origination of the aggregate, i.e., what gives rise to it; directly knowing the cessation of that aggregate; and directly knowing the path of practice leading to the cessation of that aggregate, i.e., the noble eightfold path.

The first two of these steps—in which the aggregates are observed as they actually occur (yathābhūtaṁ)—is meant to draw attention to how the aggregates do not simply happen and maintain themselves on their own. A lot of desire and effort go into shaping them and trying to keep them going. This is why, in the first step, the term “origination” (samudaya) is important. It doesn’t denote just the act of arising; it denotes the process of causation: what makes the aggregate arise. To see this requires more than bare awareness of events. You learn about causation not by simply watching things come and go, but by trying to make them come and make them go. That’s when you learn what’s a causal factor and what’s not. SN 22:5, taken together with AN 9:36, states that the ideal way to learn about the origination of the aggregates is to turn them into a state of concentration. And SN 45:8 notes that desire is an essential part of the right effort leading to right concentration. Thus, the act of focusing your desire on giving rise to right concentration—which is part of the noble eightfold path, the fourth step—is the test case in which the aggregates are fabricated in a way that allows you to see clearly how they originate in step one.

The fifth and sixth steps expand on the role of right view and appropriate attention in the fourth step: directly knowing the allure of the aggregate, i.e., the pleasure that can be found in the aggregate; and directly knowing the drawbacks of the aggregate, i.e., the pain and suffering involved in clinging to the aggregate.

This sixth step is where the Buddha’s not-self strategy is applied. The perception of not-self is actually one of several perceptions that he says can be applied to the aggregates to drive home the point that the drawbacks of fabricating the aggregates far outweigh the allure of continuing to fabricate them. AN 9:36 lists eleven perceptions that can perform this function: perceiving the aggregates as inconstant, stressful, a disease, a cancer, an arrow, painful, an affliction, alien, a disintegration, an emptiness, not-self. These fall under three main perceptions: inconstancy, stress/suffering, and not-self. And as we have seen from the Buddha’s not-self questionnaire, these three are intimately related. If something is inconstant, it’s stressful. If it’s stressful, it’s not worth identifying as “mine,” “my self,” or “what I am.” As noted above, SN 12:15 says that this contemplation, when applied to events at the senses as they are happening, leads to a state where there are no thoughts of “existence” or “non-existence,” so there’s no concern for what this contemplation will do to the existence or non-existence of the self. In this way, the Buddha’s questionnaire, and the resulting value judgment, can be applied without fear to every aggregate as it’s experienced.

When this value judgment hits home as it catches the mind in the act of fabricating even the most desirable fabrication possible—right concentration—it leads to the seventh step, the escape from the aggregates, which is dispassion. This seventh step is actually identical with the third: the cessation of the aggregates. Because fabrication is driven by passion and desire, dispassion puts an end to the drive, and fabrication ceases. When fabrication ceases, the aggregates and everything constructed around them cease as well, and the mind, relinquishing everything, attains the unfabricated: release.

The nature of the value judgment leading to this release is reflected in the words the Buddha uses to describe clinging on the one hand, and disenchantment—the step just prior to dispassion (SN 56:11)—on the other. Upādāna, the word for clinging, also denotes sustenance and the act of taking sustenance. In other words, to feed is to cling is to suffer: This is the import of the Buddha’s first noble truth. Nibbidā, disenchantment, is a word used to describe the sense of having had enough of a particular food, and not wanting to eat it any more. So: To cling to something as self is a way of feeding on it; perceptions of not-self—along with the other perceptions focusing on the drawbacks of the mind’s sustenance and taking sustenance—are meant to turn an avid eater into one with no more appetite. The good news of the Buddha’s teachings is that in losing your desire to feed in this way, you don’t starve. Instead, you’re brought to a dimension where there’s no need to feed. As Sn 3:12 and many other texts affirm, the freedom of unbinding is hunger-free.

The dynamic underlying this change of heart depends on more than simply agreeing to arbitrary definitions of terms. It has to come from a value judgment, as you catch the mind in the process of shaping the food on which it wants to feed, and see that the anticipated rewards are simply not worth it. Only a judgment of this sort, focused on the mind’s activities as they are in the course of actually happening (yathābhūtaṁ) can break the mind of its ignorant, unskillful habits.

Because the mind’s habits are the main factor shaping its experience—this is the point of the famous first line of the Dhammapada, that the mind precedes all phenomena—getting it to change its habits will change its experience. To get it to stop fabricating entirely will allow it to experience the unfabricated. And it’s precisely in the power of the Buddha’s teachings to steer the mind, the chief instigator, in this direction that their performative function lies.

The exact nature of the difference between the approach outlined here and the one offered in the critique can be highlighted by exploring a seemingly small issue of translation. As part of his critique, Ven. Bodhi cites a passage from SN 22:126 to the effect that ignorance can be ended by observing that the aggregates are subject to arising and ceasing. However, the term he translates as “arising” is the same term used above in SN 22:57: samudaya, or origination. To translate it as “arising” gives the impression that ignorance can be ended by witnessing, through bare awareness, the arising and ceasing of the aggregates and concurring with the general principle that, yes, they do arise and cease.

But this misses an important dynamic in the practice, which lies in seeing the extent to which your own desires and efforts play a complicit causal role in that arising and that, in fostering a passion for fabricating, you’ve been fooling yourself all along. It’s only when you stop fabricating—on realizing that the allure of the aggregates is not worth the effort of fabricating—that the unfabricated can appear. The perception of not-self is one of the Buddha’s strategic, performative teachings for inducing the value judgment that can bring this necessary change of heart about.

It’s useful to note here that because the perception of not-self is a value judgment, it allows for different judgments at different stages of the path. This is important, for on the beginning stages of the path, a skillful perception of self is actually worth cultivating. If used appropriately, it can get you to start on the path and to stick with it (AN 4:159; AN 3:40). You start on the path because you see that you’ll benefit from it and that you have within you the ability to follow it; you stick with it out of a continuing concern for your own well-being. On these stages, the perception of “not-self” is applied to things that would pull you away from the practice of virtue, concentration, or discernment. Only when these practices have been mastered (AN 9:36) can the perception of not-self be applied to all phenomena, for at that point the strategy of thinking in terms of a self is no longer needed. The ultimate happiness (MN 13) has been attained.

That’s what we can gather from the Canon’s express statements as to how and why the not-self strategy works.

The Right View of One Beyond Training

9. In addition to the arguments based on the syllogism given in points 1–3, the critique of “The Not-self Strategy” bolsters its position by making three observations to the effect that arahants are endowed with right view. This, the argument goes, means that right view does not merely serve a strategic function on the path. It states a truth about the non-existence of the self that arahants continue to see as true.

• The first point is that MN 65 and MN 78 state that an arahant is endowed with the “right view of one beyond training.” The critique claims that because this term is nowhere defined, its meaning must be identical with the right view of one on the path: that all phenomena are not-self.

• The second point is that in SN 22:122 Ven. Sāriputta states that arahants should attend to the five aggregates as not-self:

“An arahant should attend in an appropriate way to these five clinging-aggregates as inconstant, stressful, a disease, a cancer, an arrow, painful, an affliction, alien, a dissolution, an emptiness, not-self. Although, for an arahant, there is nothing further to do, and nothing to add to what has been done, still these things—when developed and pursued—lead both to a pleasant abiding in the here and now, and to mindfulness and alertness.”

This, the critique concludes, means that the perception of not-self serves purposes beyond the path, and that even though an arahant no longer has to develop right view, the right view with which he/she is inalienably endowed is that all phenomena are not-self.

• The third point is in response to the fact that “The Not-self Strategy” quoted passages from the Canon stating that arahants are beyond views, and are not attached to ideas of “true” and “false.” In response, the critique notes that those passages come from poems in the Canon: passages from the Sutta Nipāta and the concluding poem of AN 4:24. Being poems, it says, these passages are only suggestive rather than lucid, and so are not as reliable a guide to the Dhamma as the prose passages. Because the above prose passages show that arahants in no way discard right view even though they don’t cling to it, those passages should be taken as more authoritative.

To respond to these three observations:

• First, there are many stages of right view even prior to awakening. As noted three times above, every arahant has gone through an advanced stage of right view where notions of “existence” and “non-existence” don’t occur to the mind:

“By and large, Kaccāna, this world is supported by [takes as its object] a polarity, that of existence and non-existence. But when one sees the origination of the world [i.e., the six sense media] as it has come to be with right discernment, ‘non-existence’ with reference to the world does not occur to one. When one sees the cessation of the world as it has come to be with right discernment, ‘existence’ with reference to the world does not occur to one.

“By and large, Kaccāna, this world is in bondage to attachments, clingings [sustenances], and biases. But one such as this does not get involved with or cling to these attachments, clingings, fixations of awareness, biases, or obsessions; nor is he resolved on ‘my self.’ He has no uncertainty or doubt that mere stress, when arising, is arising; stress, when passing away, is passing away. In this, his knowledge is independent of others. It’s to this extent, Kaccāna, that there is right view” (SN 12:15).

SN 22:94 shows that the Buddha, after awakening, would use concepts of existence and non-existence to talk about the world. But you have to wonder: After having developed the right view described above, and seen the release that comes from it, to what extent would he hold to concepts of “existence” and “non-existence” within his own mind? There’s no way of knowing apart from attaining full awakening yourself. Even Anāthapiṇḍika, a stream-enterer, when asked about the Buddha’s views, replied, “I don’t know all of the Blessed One’s view” (AN 10:93). And although, for an awakened one, statements of right view may be true as far as they go, only one who, like an arahant, has known the limits of description and what lies beyond those limits of description (DN 15) would be in a position to know how far that “true” actually goes.

As SN 47:4 states, arahants still develop the establishings of mindfulness after their awakening, but they do it in a way that they are disjoined from the frames of reference on which those establishings are based. This includes the framework of dhammas:

“Monks, even those who are arahants—whose effluents are ended, who have reached fulfillment, done the task, laid down the burden, attained the true goal, totally destroyed the fetter of becoming, and who are released through right gnosis—even they remain focused on dhammas in and of themselves—being ardent, alert, unified, clear-minded, concentrated, and single-minded, disjoined from dhammas.”

Because “dhammas” here includes not only the five clinging-aggregates, but also the four noble truths—and thus the fourth truth, the path, and the factor of right view within the path—arahants experience right view disjoined from it, just as they are disjoined from all of the six sense media and their objects (MN 140).

The prose section of AN 4:24 contains this interesting passage:

Whatever in this world—with its devas, Māras and Brahmās, its generations with their contemplatives and brahmans, rulers and common people—is seen, heard, sensed, cognized, attained, sought after, pondered by the intellect: That I directly know. That has been realized by the Tathāgata, but in the Tathāgata it has not been established.”

So, apart from an actual experience of full awakening, it’s hard to know what the experience of being disjoined—or of a truth’s not being established in one’s mind—is actually like. But it’s certainly not identical to the way a person on the path relates to right view, as AN 4:24 makes clear:

“Whatever is seen or heard or sensed

and fastened onto as true by others,

One who is Such—among the self-fettered—

would not further claim to be true or even false.”

Even if we were to regard this passage as only be a suggestion, it still suggests some important things. One of them is that it would be foolhardy to say that, from the Canon, we can confidently infer the nature of an arahant’s relationship to a right view about things true and false.

• In response to the second observation: Although it is true that SN 22:122 describes a strategic use for the perception of not-self beyond the path, it’s still just a strategic use: for the purpose of a pleasant abiding in the here-and-now and for mindfulness and alertness. And as the passage clearly states, the arahant has no need for this contemplation. It’s simply a pleasant way for an awakened person to spend the time, mindful and alert. There is nothing in SN 22:122 to indicate that this contemplation performs any function for an arahant beyond serving that strategic purpose. And given what we have noted under the first point, it would be hard to say how far the truth of that perception goes in the eyes of someone who has seen the limitations of perception and what lies beyond perception.

• As for the third observation: Not all the Canon’s statements about the limitations of language in describing the arahant’s relationship toward right view are contained in the poems. Some prose passages speak of these limitations as well. For instance, a prose discourse, SN 48:4, clearly states that the arahant has realized the escape from discernment, which is equivalent to right view. The passages in SN 12:15, DN 15, and AN 10:93, cited above to support this point, are in prose, as is the first half of AN 4:24, cited in the original article.

Secondly, the prose passages of the Canon never suggest that the poems recorded in the Canon are to be dismissed as “only suggestive.” On the contrary, several prose passages are devoted to ferreting out the meaning of verses that they treat as particularly profound. (See, for example, SN 22:3, AN 3:32, AN 3:33, AN 4:41, and AN 6:61.) In AN 4:231, the Buddha distinguishes among four kinds of poets, one of them being the meaning-poet. Although he doesn’t define any of the four, the implication is that he himself was a meaning-poet, one whose verses had meaning and were not to be dismissed, in the words of AN 2:46, as being merely “elegant in sound and elegant in rhetoric.” Given that the Buddha and the compilers of the Canon took their poetry seriously, it’s an act of scholarly arrogance for modern interpreters to dismiss that poetry simply because its message lies outside the categories of our own thought and language. This is especially true in the case of a teaching, like the Buddha’s, whose whole point lies beyond the boundaries of description.

One of the important lessons of the Aṭṭhaka Vagga, a chapter in the Sutta Nipāta devoted to the theme of not-clinging, is that language is slippery. Even though it has a strategic purpose—in the Buddha’s case, to convey lessons of the Dhamma—it falls short of the highest dhammas, and even further short of the ending of dhammas, unbinding (AN 10:58).

This is why—when dealing with all lessons of the Dhamma, including the lessons of not-self—it’s important to view the language of perceptions and thought-fabrications as performative and to use it strategically: to get the mind to what lies beyond perceptions and thought-fabrications, and not to develop a scholarly fixation on perceptions and thought-fabrications as expressions of truth in and of themselves. Otherwise, we risk wasting our time trying to catch in the net of language something that no words can catch.

Effluents ended,

independent of nutriment,

their pasture—emptiness

& freedom without sign:

their trail,

like that of birds through space,

can’t be traced. (Dhp 93)