A proper appreciation of the teaching on not self as a meditative strategy requires keeping in mind that the absence of a permanent self does not imply that there is nothing at all. It only implies that everything, without exception, is changing and conditioned. The dependently arisen and impermanent flow of subjective experience by way of the five aggregates (bodily form, vedanā, perception, volitional formations, and consciousness) certainly is present. But this is just an impermanent and dependently arisen process, which is devoid of some sort of unchanging and self-sufficient essence or entity.
Similarly, targeting the notion “I am” for meditative scrutiny is not meant to imply that we are no longer allowed to use the word “I.” Awakened ones can still use the first person singular pronoun and the discourses show the Buddha to have had no qualms about saying, for example, “I have backpain; I want to rest.” The task is only to become aware of the tendency toward “selfing,” the self-referentiality that is almost continuously present in the background of unawakened experience.
In the same vein, classifying things as “not mine” does not imply that we can no longer own anything at all. Even monastics, in spite of having renounced worldly possessions, own things like their bowls or robes, etc. After having reached full awakening, the knowledge that a certain bowl or robe is “mine” will still be possible and even meaningful for them. What the meditative strategy puts into question is only the sense of possessiveness, the clinging to things as “mine.” This clinging makes it difficult, if not impossible, to let go of what has been appropriated in this way with attachment, which inevitably leads to sorrow and affliction if anything happens to those things. A broken bowl and a torn robe can only upset those who own them with clinging.
How these different avenues of insight into not self can be related to vedanā emerges from a passage in the Mahānidāna-sutta (DN 15). Although this part of the discourse is somewhat complex and requires some effort to be understood, it can yield powerful support for insight meditation.
The relevant passage begins with the
Buddha (here referred to as the Tathāgata) declaring that there are
three ways in which the notion of a self could be related to vedanā,
followed by taking up each of these options for closer scrutiny to show
their unconvincing nature. For a better appreciation of this
exposition, in what follows I provide a translation of the Chinese Āgama parallel to the Mahānidāna-sutta.
This is to enable the reader to compare this with existing translations
of the Pāli discourse and thereby come to a personal understanding of
the often-minor differences between these parallel versions, both of
which are the product of centuries of oral transmission. In the Dīrgha-āgama parallel to the Mahānidāna-sutta, the first option taken up for such closer scrutiny identifies the self with vedanā:
Ānanda, as to a person who has the view of self, declaring that ‘vedanā is the self,’ one should tell that person: ‘The Tathāgata has taught three [types of] vedanā: pleasant vedanā, painful vedanā, and neutral vedanā.’
At the time when there is pleasant vedanā, there is no painful vedanā or neutral vedanā. At the time when there is painful vedanā, there is no pleasant vedanā or neutral vedanā. At the time when there is neutral vedanā, there is no painful vedanā or pleasant vedanā.
Ānanda, the reason for this is that pleasant vedanā arises in dependence on pleasant contact. If pleasant contact ceases, the vedanā also ceases. Ānanda, painful vedanā arises in dependence on painful contact. If painful contact ceases, the vedanā also ceases. [Ānanda,] neutral vedanā arises in dependence on neutral contact. If neutral contact ceases, the vedanā also ceases.
Ānanda, it is like rubbing together two firesticks to kindle a fire. Putting each of them in a different place, there will consequently be no fire. This is also like that: Because pleasant vedanā arises in dependence on pleasant contact, if pleasant contact ceases, that vedanā also ceases completely. Because painful vedanā arises in dependence on painful contact, if painful contact ceases, that vedanā also ceases completely. Because neutral vedanā arises in dependence on neutral contact, if neutral contact ceases, that vedanā also ceases completely.
Ānanda, these three vedanās are conditioned and impermanent, being born of conditions, and they are of a nature to end, of a nature to cease, of a nature to become decayed. They do not belong to a self and there is no self in them. With right knowledge one should contemplate them as they really are. Ānanda, as to the person who has the view of self, by taking vedanā to be the self, that is consequently to be rejected.
parallel versions agree in highlighting the mutually exclusive nature
of the three vedanās, which is what counters identifying a particular
vedanā as a self. The corresponding examination in the Mahānidāna-sutta directly targets the main problem resulting from the fact that three different types of vedanā exist. Which of these three is supposed to be the self? A parallel in the Madhyama-āgama proceeds similarly:
You have three [types of] vedanā: pleasant vedanā, painful vedanā, and neutral vedanā. Of these three [types of] vedanā, which vedanā do you view as the self?
The question posed in this way clarifies the main problem with the notion that equates vedanā with the self (the alternative idea that the self experiences all these different vedanās will be taken up later). Which of these three mutually exclusive vedanās is supposed to be the self? If one of them were to be identified as the self, what happens when one of the others manifests?
The passage translated above from the Dīrgha-āgama discourse stands alone in providing additional support for establishing this mutually exclusive nature. It achieves this by directing attention to contact as the condition for vedanā. This can conveniently be related back to an earlier part of the Mahānidāna-sutta and its parallels, which in the context of a detailed examination of dependent arising highlights the role of contact as what precedes vedanā and serves as the condition for its arising.
Contacts through any of the six sense doors occur one after the other. It is not the case that contacts through different sense doors occur simultaneously. Although this may sometimes seem to be the case from a subjective perspective, closer inspection shows that what really happens is a quick succession of different contacts. Each type of contact in turn conditions the type of vedanā that arises in dependence on it. This form of presentation clarifies why the three types of vedanā are seen as mutually exclusive.
The Dīrgha-āgama version also
offers an illustration of the conditioned dependence of vedanā on
contact with the help of a simile of two fire sticks, which is not found
in the parallel versions. The fire sticks will only produce fire if
they are rubbed against each other. Just as anyone in the ancient Indian
setting would have known from personal experience that rubbing two fire
sticks together provides the necessary condition for making a fire, in
the same way contact should be seen as providing the necessary condition
for the arising of vedanā. It follows that, with the cessation of
contact, the corresponding vedanā ceases. Although the simile of the two fire sticks is not found in the Mahānidāna-sutta, it does occur in other Pāli discourses and their parallels.
Having in this way established the mutually exclusive nature of the three types of vedanā, the parallel versions continue by pointing out that vedanā is impermanent and of a nature to cease. Since positing a self in the ancient Indian setting implied a permanent and inherently blissful entity, it follows that vedanā cannot be identified as such a self. Bhikkhu Bodhi (1984/1995: 35) clarifies the underlying reasoning:
If feeling is self, whatever attributes belong to feeling also belong to self and whatever happens to feeling also happens to self. Since feeling is impermanent, conditioned, dependently arisen, and subject to destruction, it would follow that the same pertains to self. This is a conclusion the theorist could not accept, as it contradicts his conception of self as permanent, unconditioned, independent, and indestructible; yet his initial thesis forces it upon him. Further, all feeling ceases and disappears, so if one identifies a particular feeling as self, with the ceasing of that feeling one would have to assert that self has disappeared—for the theorist an unthinkable situation, as it would leave him without the self he is seeking to establish.
Exploring possible practice applications, the idea of identifying vedanā as a self can go beyond explicitly formulated views. In fact, whereas the view of self is left behind with stream-entry, selfing and conceit are only abandoned once full awakening has been reached. Hence the matter broached above is not yet settled once we no longer uphold explicit views of an eternal self. This in turn means that the same teaching can also have a bearing on tendencies of the untrained mind that are more subtle than the proposing of philosophical standpoints based on positing a permanent self.
A practical example would be the case of an excruciating pain in some part of the body. The tendency of the untrained mind is to become completely drawn into and absorbed in the pain, to the extent of fully identifying with the pain. The pain has become so intense that its experience captures the entire range of awareness such that it is no longer possible to discern anything else but the pain.
Mindfulness can offer a cure for the tendency to become overwhelmed by the pain. A first step could simply be to make a conscious effort to direct attention to parts of the body that are not in pain right now. If, for example, the back is hurting terribly, what about the soles of the feet? The feet? The lower legs? The upper legs? By gradually introducing recognition of areas of the body that are not directly impacted by the pain, awareness broadens and the pain no longer captivates the whole field of attention.. After all, it is just a particularly strong vedanā, nothing more.
In case the pain is of the type that affects the whole body, an approach would be to inquire if other possible diseases are also there. Take the case of a strong fever that affects the whole body. What about a toothache? Are any bones broken? Stomach cramps? It turns out that there is quite a range of possible afflictions that are not present. Noticing which particular illnesses are currently absent can have a similar effect of preventing the pain from captivating the whole field of attention.
In each case, the idea is to find ways of lessening identification with the pain, of diminishing the tendency to equate oneself and become completely engulfed in the experience of that particular vedanā. In one way or another, there will always be aspects of subjective experience that need not be dominated by that particular pain. Moreover, any vedanā will eventually change; none will last forever. It is all just a process. No need to make it into more than what it actually is, and in particular no need to identify with it.
Returning to the exposition in the Mahānidāna-sutta and its parallels, after having dismissed the notion that vedanā could be identified as a self, the parallels continue with the possibility of instead positing a relationship of the self to vedanā. This presumably intends to prevent the belief in a self from being contradicted by the mutually exclusive nature of the different vedanās.
According to the Dīrgha-āgama version, someone holding this view would express it as follows: “vedanā is not the self; the self is the one who feels.” In the Madhyama-āgama parallel,
this position takes the similar form of someone who “does not view
vedanā as the self, but views the self as being able to feel, it being
the nature of the self to be able to feel.”
The chief problem with this proposition is the impermanent nature of vedanās. The experiencing of something that is impermanent must also be a changing phenomenon itself, otherwise it would be forever frozen in the condition of knowing just one thing. If the ability to know is considered a permanent entity, its very permanency implies that it could no longer know change. The very knowing of change implies that the knowing itself has also changed and will continue to change. In other words, to turn a function (here the function of feeling vedanās) into a permanent entity makes it become dysfunctional.
The only possibility to avoid that the proposed permanent entity is not affected by the changing nature of experiences is to posit it as something completely apart from the function of experiencing. As a result of such a move, however, this entity would become unable to feel vedanās (a position to be examined below). But the ability to feel impermanent vedanās must itself be amenable to change in order to be able to perform its function.
A practical application of the above could take the form of directing attention to the tendency to appropriate and own vedanā. This can also be applied to the case of physical pain, of course, where a shift from “my pain” to just “pain” and from “my body” being afflicted to just “the body” being afflicted can make a world of difference. Since pain has already been taken up above, however, it seems appropriate to explore also the case of pleasant vedanā. After all, it is particularly with agreeable experiences that the tendency to appropriation can manifest strongly, by way of wanting to own the pleasure and keep it forever. To counter such a tendency, it can be helpful to make it a continuous habit to share any joy or happiness with others. Such sharing is not confined to material sharing, although that is of course a prominent way to implement the basic attitude of sharing. A traditional practice that exemplifies the basic attitude is sharing one’s merits. In this or any other way, the task is simply to counter any tendency to appropriate what is liked by immediately sharing it with others.
Returning once again to the exposition in the Mahānidāna-sutta and its parallels, yet another possibility of positing a self would be to completely dissociate this self from vedanā, in the sense of proposing that the two are just unrelated to each other and the proposed self does not feel any vedanā. The Madhyama-āgama version presents this as involving the view that “the self is without feeling.” Yet, this attempt also fails to provide a satisfactory solution. Bhikkhu Bodhi (1984/1995: 36) explains:
Fundamental to the notion of selfhood is an inherent capacity for self-affirmation; as the autonomous subject of experience, self should be able to affirm its own being and identity to itself without need for external referents. Yet, the theorist is forced to admit that, with the cessation of feeling, in the complete absence of feeling, the idea ‘I am this’ could not be conceived. The assumed self can only identify itself as ‘this,’ e.g. ‘I am the experiencer of feeling,’ by reference to its psychophysical adjuncts. If these are removed, all points of reference for self to conceive its identity are removed and it then becomes a conceptual cipher … The question clinches the point that the supposed self, being incapable of identifying itself without reference to its adjuncts, becomes totally dependent upon them for its identity—a strange predicament for an autonomous self to get into.
In other words, the notion of a self is intrinsically bound up with experience. A self that is dissociated from experience, being unable to feel or be conscious, becomes a meaningless position. Precisely because it has been completely dissociated from actual experience, the notion of a self becomes somewhat irrelevant to understanding and dealing with actual experience.
From a practical perspective, developing a meditative approach related to this position could take the form of noting the extent to which vedanā is integral to experience. For such exploration, neutral vedanā would be of particular importance, as its contemplation shows that the absence of pleasure and pain is not the absence of vedanā. Once the all-pervasive and continuous role of vedanā in experience has been fully appreciated in this way, it becomes clear that any type of experience necessarily will involve some vedanā or the other.
An exception to this rule, in the sense of being a type of experience where vedanā is absent, is the realization of Nibbāna. Yet, this is at the same time the type of experience where any basis for the notion of a self completely disappears. In a way, selfing and conceit are intimately bound up with vedanā. Their cessation comes through the direct experience of the ceasing of vedanā that occurs when the entire chain of dependent arising ceases.
Looking back on the positions surveyed in the Mahānidāna-sutta and its parallels, a natural starting point for contemplation of not self, besides the more specific practical dimensions already surveyed above, is the mindful contemplation of the impermanent nature of vedanā. In the absence of a proper understanding, the experience of vedanā can easily become interwoven with a subjective sense of an enduring ‘I’ or ‘me’ as the one who feels. The tendency for this subjective sense to manifest needs to be noticed alongside awareness of the constant change of vedanā from one feeling tone to another. The impermanent nature of vedanā, evident in this way, clearly shows that this sense of a permanent ‘I’ at the center of experience cannot be identified with vedanās, because they keep changing all the time. It also undermines the idea of a permanent agent that feels vedanā, because vedanās that change from one type to another inevitably involve a variety of experiences. Hence, their knowing needs to be something changing as well. Nor does it make sense to posit a permanent experiencer completely apart from the felt experience of change.
The key doctrine of not-self in the sense of an absence of a permanent self anywhere in subjective experience may well be considered the most marked departure of the Buddha’s teaching from ancient Indian thought. The potential of contemplating the impermanent nature of vedanās to lead to insight into not self can be strengthened by catching within the purview of attention the sense of an ‘I’ lurking at the background of the meditative experience. With the groundwork laid by previous reflection along the lines of the above discourse passages, the implications of not self can emerge in an intuitive way that involves only a minimal amount of conceptual input. All it takes is applying the explorations proposed in the Mahānidāna-sutta and its parallels to actual practice in such a way that understanding deepens and freedom from attachment grows. It is in this way that the path to liberation naturally unfolds.
 For example, MN 22 at MN I 136,6 (translated by Ñāṇamoli 1995/2005: 229), and its parallel MĀ 200 at T I 765c7.
 Anālayo 2011: 35n55.
 DĀ 13 at T I 61c6, parallel to DN 15 at DN II 66,14 (translated by Walshe 1987: 227), in addition to which there are a few relevant Sanskrit fragments in Waldschmidt 1932: 9f (see also 56f), Sander 1987: 149f, and Bechert and Wille 1989: 41 and 60, Chinese discourse parallels in T 14 at T I 243c12, MĀ 97 at T I 580a11 (translated in Anālayo and Bucknell 2020: 215), and T 52 at T I 845b29, and a parallel preserved in Tibetan in Up 4068 at D 4094 ju 231a7 or P 5595 tu 264a7.
 The translation “rubbing” is based on adopting a variant reading; the original speaks instead of “accumulating.”
 DN 15 at DN II 66,15.
 MĀ 97 at T I 580a12.
 One such occurrence is MN 140 at MN III 242,31 (translated by Ñāṇamoli 1995/2005: 1091), with counterparts in the Chinese and Tibetan parallels MĀ 162 at T I 691b27, T 511 at T XIV 780c3, and Up 1041 at D 4094 ju 39a6 or D 5595 tu 42b6. Another occurrence is SN 12.62 at SN II 97,9 (translated by Bodhi 2000: 597), again found also in the Sanskrit and Chinese parallels Tripāṭhī 1962: 120 and SĀ 290 at T II 82a20. Another occurrence is SN 36.10 at SN IV 215,21 (translated by Bodhi 2000: 1270), but in this case the illustration does not occur in the parallel SĀ 466 at T II 119a11. Yet another occurrence is in SN 48.39 at SN V 212,21 (translated by Bodhi 2000: 1683), of which no parallel is known.
 DĀ 13 at T I 61c22.
 MĀ 97 at T I 580a9.
 MĀ 97 at T I 580a11 (the presentation of this view in DĀ 13 appears to have suffered from a textual corruption and for this reason is not taken up here).
CBETA Chinese Buddhist Electronic Text Association
D Derge edition
DĀ Dīrgha-āgama (T 1)
MĀ Madhyama-āgama (T 26)
P Peking edition
SĀ Saṃyukta-āgama (T 99)
T Taishō edition (CBETA)
Anālayo, Bhikkhu 2011: A Comparative Study of the Majjhima-nikāya, Taipei: Dharma Drum Publishing Corporation.
Anālayo, Bhikkhu and R. S. Bucknell 2020: The Madhyama Āgama (Middle-Length Discourses), Volume II, Moraga, California: Bukkyō Dendō Kyōkai America.
Bechert, Heinz and Klaus Wille 1989: Sanskrithandschriften aus den Turfanfunden, Teil 6, Stuttgart: Franz Steiner.
Bodhi, Bhikkhu 1984/1995: The Great Discourse on Causation, The Mahānidāna Sutta and its Commentaries, Translated from the Pali, Kandy: Buddhist Publication Society.
Bodhi, Bhikkhu 2000: The Connected Discourses of the Buddha, A New Translation of the Saṃyutta Nikāya, Boston: Wisdom Publication.
Ñāṇamoli, Bhikkhu 1995/2005: The Middle Length Discourses of the Buddha, A Translation of the Majjhima Nikāya, Bhikkhu Bodhi (ed.), Boston: Wisdom.
Sander, Lore 1987: Nachträge zu ‚Kleinere Sanskrit-Texte Heft III-V’, Wiesbaden: Franz Steiner.
Tripāṭhī, Chandrabhāl 1962: Fünfundzwanzig Sūtras des Nidānasaṃyukta, Berlin: Akademie Verlag.
Waldschmidt, Ernst 1932: Bruchstücke Buddhistischer Sūtras aus dem Zentralasiatischen Sanskritkanon, Herausgegeben und im Zusammenhang mit ihren Parallelversionen bearbeitet, Leipzig: F.A. Brockhaus.
Walshe, Maurice 1987: Thus Have I Heard; The Long Discourses of the Buddha, London: Wisdom.