Classic and Contemporary Buddhist Works

The Connected Discourses of the Buddha - Introduction

A Translation of the Saṃyutta Nikāya


The Saṃyutta Nikāya is the third great collection of the Buddha’s discourses in the Sutta Piṭaka of the Pāli Canon, the compilation of texts authorized as the Word of the Buddha by the Theravāda school of Buddhism. Within the Sutta Piṭaka it follows the Dıgha Nikāya and Majjhima Nikāya, and precedes the Aṅguttara Nikāya. Like the other Pāli Nikāyas, the Saṃyutta Nikāya had counterparts in the canonical collections of the other early Buddhist schools, and one such version has been preserved in the Chinese Tripiṭaka, where it is known as the Tsa-a-han-ching. This was translated from the Sanskrit Saṃyuktāgama, which the evidence indicates belonged to the Sarvāstivāda school. Thus, while the Saṃyutta Nikāya translated in the present work has its locus within the Theravāda canon, it should never be forgotten that it belongs to a body of texts—called the Nikāyas in the Pāli tradition prevalent in southern Asia and the Āgamas in the Northern Buddhist tradition—which stands at the fountainhead of the entire Buddhist literary heritage. It was on the basis of these texts that the early Buddhist schools established their systems of doctrine and practice, and again it was to these texts that later schools also appealed when formulating their new visions of the Buddha’s way.

As a source of Buddhist doctrine the Saṃyutta Nikāya is especially rich, for in this collection it is precisely doctrinal categories that serve as the primary basis for classifying the Buddha’s discourses. The word saṃyutta means literally “yoked together,” yutta (Skt yukta) being etymologically related to our English “yoked” and saṃ a prefix meaning “together.” The word occurs in the suttas themselves with the doctrinally charged meaning of “fettered” or “bound.” In this sense it is a past participle related to the technical term saṃyojana, “fetter,” of which there are ten that bind living beings to saṃsāra, the round of rebirths. But the word saṃyutta is also used in a more ordinary sense to mean simply things that are joined or “yoked” together, as when it is said, “Suppose, friend, a black ox and a white ox were yoked together by a single harness or yoke” (35:232; S IV 163,12–13). This is the meaning relevant to the present collection of texts. They are suttas—discourses ascribed to the Buddha or to eminent disciples—yoked or connected together. And what connects them, the “harness or yoke” (damena vā yottena vā), are the topics that give their titles to the individual chapters, the saṃyuttas under which the suttas fall.

The Groundplan of the Saṃyutta Nikāya

Despite the immense dimensions of the work, the plan according to which it is constructed is fairly simple and straightforward. The Saṃyutta Nikāya that has come down in the Pāli tradition consists of five major Vaggas, parts or “books,” each of which corresponds to a single volume in the Pali Text Society’s roman-script edition of the work. Between them, these five volumes contain fifty-six saṃyuttas, chapters based on unifying themes. The longer saṃyuttas are in turn divided into subchapters, also called vaggas, while the smaller saṃyuttas can be considered to consist of a single vagga identical with the saṃyutta itself. Each vagga, in this sense, ideally contains ten suttas, though in actuality the number of suttas in a vagga can range from as few as five to as many as sixty. Thus we find the word vagga, literally “a group,” used to designate both the five major parts of the entire collection and the subordinate sections of the chapters.

The two largest saṃyuttas, the Khandhasaṃyutta (22) and the Saḷāyatanasaṃyutta (35), are so massive that they employ still another unit of division to simplify organization. This is the paññāsaka or “set of fifty.” This figure is only an approximation, since the sets usually contain slightly more than fifty suttas; indeed, the Fourth Fifty of the Saḷāyatanasaṃyutta contains ninety-three suttas, among them a vagga of sixty! Most of these suttas, however, are extremely short, being merely variations on a few simple themes.

Unlike the suttas of the first two Nikāyas, the Dıgha and the Majjhima, the suttas of SN do not have proper names unanimously agreed upon by all the textual traditions. In the old ola leaf manuscripts the suttas follow one another without a clean break, and the divisions between suttas have to be determined by certain symbolic markings. Each vagga ends with a short mnemonic verse called the uddāna, which sums up the contents of the vagga by means of key words representing its component suttas. In modern printed editions of SN these key words are taken to be the titles of the suttas and are placed at their head. As the uddānas often differ slightly between the Sinhalese and the Burmese textual traditions, with the PTS edition following now one and now the other, the names of the suttas also differ slightly between the several editions. Moreover, the most recent Burmese edition, that prepared at the Sixth Buddhist Council, sometimes assigns the suttas titles that are fuller and more meaningful than those derivable from the mnemonic verses. In this translation I have generally followed the Burmese edition.

The titles of the vaggas also occasionally differ between the traditions. Whereas the Burmese-script edition often names them simply by way of their numerical position—e.g., as “The First Subchapter” (paṭhamo vaggo), etc.—the Sinhala-script Buddha Jayanti edition assigns them proper names. When the titles of the vaggas differ in this way, I have placed the numerical name given in the Burmese-script edition first, followed parenthetically by the descriptive name given in the Sinhala-script edition. The titles of the vaggas are without special significance and do not imply that all the suttas within that vagga are related to the idea expressed by the title. Often these titles are assigned merely on the basis of one sutta within the vagga, often the first, occasionally a longer or weightier sutta coming later. The grouping of suttas into vaggas also appears largely arbitrary, though occasionally several successive suttas deal with a common theme or exemplify an extended pattern.

In his commentaries to the Pāli Canon, Ācariya Buddhaghosa states that SN contains 7,762 suttas, but the text that has come down to us contains, on the system of reckoning used here, only 2,904 suttas. Due to minor differences in the method of distinguishing suttas, this figure differs slightly from the total of 2,889 counted by Léon Feer on the basis of his roman-script edition. Table 1 shows how these figures are arrived at, with the divisions into Vaggas, saṃyuttas, and vaggas; the variant figures counted by Feer are given next to my own. The fact that our totals differ so markedly from that arrived at by Buddhaghosa should not cause alarm bells to ring at the thought that some 63% of the original Saṃyutta has been irretrievably lost since the time of the commentaries. For the Sāratthappakāsinı, the SN commentary, itself provides us with a check on the contents of the collection at our disposal, and from this it is evident that there are no suttas commented on by Buddhaghosa that are missing from the Saṃyutta we currently possess. The difference in totals must certainly stem merely from different ways of expanding the vaggas treated elliptically in the text, especially in Part V. However, even when the formulaic abridgements are expanded to the full, it is difficult to see how the commentator could arrive at so large a figure.

The five major Vaggas or “books” of the Saṃyutta Nikāya are constructed according to different principles. The first book, the Sagāthāvagga, is unique in being compiled on the basis of literary genre. As the name of the Vagga indicates, the suttas in this collection all contain gāthās or verses, though it is not the case (as Feer had assumed at an early point) that all suttas in SN containing verses are included in this Vagga. In many suttas of Part I, the prose setting is reduced to a mere framework for the verses, and in the first saṃyutta even this disappears so that the sutta becomes simply an exchange of verses, presumably between the Buddha and an interlocutor. The other four Vaggas contain major saṃyuttas concerned with the main doctrinal themes of early Buddhism, accompanied by minor saṃyuttas spanning a wide diversity of topics. Parts II, III, and IV each open with a large chapter devoted to a theme of paramount importance: respectively, the chain of causation (i.e., dependent origination, in SN 12), the five aggregates (22), and the six internal and external sense bases (35). Each of these Vaggas is named after its opening saṃyutta and also includes one other saṃyutta dealing with another important topic secondary to the main one: in Part II, the elements (14); in Part III, philosophical views (24); and in Part IV, feeling (36). The other saṃyuttas in each of these collections are generally smaller and thematically lighter, though within these we can also find texts of great depth and power. Part V tackles themes that are all of prime importance, namely, the various groups of training factors which, in the post-canonical period, come to be called the thirty-seven aids to enlightenment (sattatiṃsa bodhipakkhiyā dhammā). The Vagga concludes with a saṃyutta on the original intuition around which the entire Dhamma revolves, the Four Noble Truths. Hence this book is called the Mahāvagga, the Great Book, though at one point it might have also been called the Maggavagga, the Book of the Path (and indeed the Sanskrit version translated into Chinese was so named).

The organization of SN, from Parts II to V, might be seen as corresponding roughly to the pattern established by the Four Noble Truths. The Nidānavagga, which focuses on dependent origination, lays bare the causal genesis of suffering, and is thus an amplification of the second noble truth. The Khandhavagga and the Saḷāyatanavagga highlight the first noble truth, the truth of suffering; for in the deepest sense this truth encompasses all the elements of existence comprised by the five aggregates and the six internal and external sense bases (see 56:13, 14). The Asaṅkhatasaṃyutta (43), coming towards the end of the Saḷāyatanavagga, discusses the unconditioned, a term for the third noble truth, Nibbāna, the cessation of suffering. Finally, the Mahāvagga, dealing with the path of practice, makes known the way to the cessation of suffering, hence the fourth noble truth. If we follow the Chinese translation of the Skt Saṃyuktāgama, the parallelism is still more obvious, for this version places the Khandhavagga first and the Saḷāyatanavagga second, followed by the Nidānavagga, thus paralleling the first and second truths in their proper sequence. But this version assigns the Asaṅkhatasaṃyutta to the end of the Mahāvagga, perhaps to show the realization of the unconditioned as the fruit of fulfilling the practice.

I said above that what makes the suttas of this collection “connected discourses” are the themes that unite them into fixed saṃyuttas. These, which we might consider the “yokes” or binding principles, constitute the groundplan of the collection, which would preserve its identity even if the saṃyuttas had been differently arranged. There are fifty-six such themes, which I have distinguished into four main categories: doctrinal topics, specific persons, classes of beings, and types of persons. Of the two saṃyuttas that do not fall neatly into this typology, the Vanasaṃyutta (9) is constructed according to a fixed scenario, generally a monk being admonished by a woodland deity to strive more strenuously for the goal; the Opammasaṃyutta (20) is characterized by the use of an extended simile to convey its message.

In Table 2 (A) I show how the different saṃyuttas can be assigned to these categories, giving the total numbers of suttas in each class and the percentage which that class occupies in the whole. The results of this tabulation should be qualified by noting that the figures given are based on a calculation for the whole Saṃyutta Nikāya. But the Sagāthāvagga is so different in character from the other Vaggas that its eleven saṃyuttas skew the final results, and thus to arrive at a more satisfactory picture of the overall nature of the work we might omit this Vagga. In Table 2 (B) I give the results when the Sagāthāvagga is not counted. Even these figures, however, can convey a misleading picture, for the classification is made by way of titles only, and these provide a very inadequate indication of the contents of the actual saṃyutta. The Rāhulasaṃyutta and the Rādhasaṃyutta, for example, are classified under “Specific Person,” but they deal almost exclusively with the three characteristics and the five aggregates, respectively, and give us absolutely no personal information about these individuals; thus their content is properly doctrinal rather than biographical. Moreover, of the eleven chapters named after specific persons, nine are almost entirely doctrinal. Only saṃyuttas 16 and 41, respectively on Mahākassapa and Citta the householder, include material that might be considered of biographical interest. Since the chapters on the main doctrinal topics are invariably much longer than the other chapters, the number of pages dealing with doctrine would be immensely greater than those dealing with other themes.

The Saṃyutta Nikāya and the Saṃyuktāgama

The Pāli commentaries, and even the canonical Cullavagga, give an account of the First Buddhist Council which conveys the impression that the participating elders arranged the Sutta Piṭaka into essentially the form in which it has come down to us today, even with respect to the precise sequence of texts. This is extremely improbable, and it is also unlikely that the council established a fixed and final recension of the Nikāyas. The evidence to the contrary is just too massive. This evidence includes the presence in the canon of suttas that could only have appeared after the First Council (e.g., MN Nos. 84, 108, 124); signs of extensive editing internal to the suttas themselves; and, a weighty factor, the differences in content and organization between the Pāli Nikāyas and the North Indian Āgamas preserved in the Chinese Tripiṭaka. It is much more likely that what took place at the First Council was the drafting of a comprehensive scheme for classifying the suttas (preserved only in the memory banks of the monks) and the appointment of an editorial committee (perhaps several) to review the material available and cast it into a format conducive to easy memorization and oral transmission. Possibly too the editorial committee, in compiling an authorized corpus of texts, would have closely considered the purposes their collections were intended to serve and then framed their guidelines for classification in ways designed to fulfil these purposes. This is a point I will return to below. The distribution of the texts among groups of reciters (bhāṇakas), charged with the task of preserving and transmitting them to posterity, would help to explain the divergences between the different recensions as well as the occurrence of the same suttas in different Nikāyas.

Comparison of the Pāli SN with the Chinese Saṃyuktāgama is particularly instructive and reveals a remarkable correspondence of contents arranged in a different order. I already alluded just above to some differences in organization, but it is illuminating to examine this in more detail. The Chinese version contains nine major Vaggas (following Anesaki, I use the Pāli terms and titles for consistency). The first is the Khandhavagga (our III), the second the Saḷāyatanavagga (our IV), the third the Nidānavagga (our II), which latter also contains the Saccasaṃyutta (56) and the Vedanāsaṃyutta (36), departing markedly from SN in these allocations. Then follows a fourth part named Sāvakavagga, without a counterpart in the Pāli version but which includes among others the Sāriputta- (28), Moggallāna- (40), Lakkhaṇa- (19), Anuruddha- (52), and Cittasaṃyuttas (41). The fifth part, whose Pāli title would be Maggavagga, corresponds to SN Mahāvagga (our V), but its saṃyuttas are arranged in a sequence that follows more closely the canonical order of the sets making up the thirty-seven aids to enlightenment: Satipaṭṭhāna (47), Indriya (48), Bala (50), Bojjhaṅga (46), and Magga (45); this part also includes the nāpānasati- (54) and Sotāpattisaṃyuttas (55), while a series of small chapters at the end includes a Jhānasaṃyutta (53) and an Asaṅkhatasaṃyutta (43). The sixth Vagga of the Saṃyuktāgama is without a Pāli parallel but contains the Opammasaṃyutta (20) and a collection of suttas on sick persons which draws together texts distributed among various chapters of SN. Then, as the seventh book, comes the Sagāthāvagga (our I), with twelve saṃyuttas—all eleven of the Pāli version but in a different order and with the addition of the Bhikkhusaṃyutta (21), which in this recension must contain only suttas with verses. Finally comes a Buddha- or Tathāgatavagga, which includes the Kassapa- (16) and Gāmanisaṃyuttas (42), and an Assasaṃyutta, “Connected Discourses on Horses.” This last chapter includes suttas that in the Pāli Canon are found in the Aṅguttara Nikāya.

The Role of the Saṃyutta among the Four Nikāyas

Prevalent scholarly opinion, fostered by the texts themselves, holds that the principal basis for distinguishing the four Nikāyas is the length of their suttas. Thus the largest suttas are collected into the Dıgha Nikāya, the middle length suttas into the Majjhima Nikāya, and the shorter suttas are distributed between the Saṃyutta and the Aṅguttara Nikāyas, the former classifying its suttas thematically, the latter by way of the number of items in terms of which the exposition is framed. However, in an important groundbreaking study, Pāli scholar Joy Manné has challenged the assumption that length alone explains the differences between the Nikāyas. By carefully comparing the suttas of DN with those of MN, Manné concludes that the two collections are intended to serve two different purposes within the Buddha’s dispensation. In her view, DN was primarily intended for the purpose of propaganda, to attract converts to the new religion, and thus is aimed mainly at non-Buddhists favourably disposed to Buddhism; MN, in contrast, was directed inwards towards the Buddhist community and its purpose was to extol the Master (both as a real person and as an archetype) and to integrate monks into the community and the practice. Manné also proposes that “each of the first four Nikāyas came about in order to serve a distinct need and purpose in the growing and developing Buddhist community” (p. 73). Here we shall briefly address the question what purposes may have been behind the compilation of SN and AN, in contradistinction to the other two Nikāyas.

In approaching this question we might first note that the suttas of these two Nikāyas provide only minimal circumstantial background to the delivery of the Buddha’s discourses. With rare exceptions, in fact, a background story is completely absent and the nidāna or “setting” simply states that the sutta was spoken by the Blessed One at such and such a locale. Thus, while DN and MN are replete with drama, debate, and narrative, with DN especially abounding in imaginative excursions, here this decorative framework is missing. In SN the whole setting becomes reduced to a single sentence, usually abbreviated to “At Sāvatthı, in Jeta’s Grove,” and by the fourth book even this disappears. Apart from the Sagāthāvagga, which is in a class of its own, the other four books of SN have little ornamentation. The suttas themselves are usually issued as direct proclamations on the doctrine by the Buddha himself; sometimes they take the form of consultations with the Master by a single monk or group of monks; occasionally they are framed as discussions between two eminent monks. Many suttas consist of little more than a few short sentences, and it is not unusual for them simply to ring the permutations on a single theme. When we reach Part V whole chains of suttas are reduced to mere single words in mnemonic verses, leaving to the reciter (or to the modern reader) the task of blowing up the outline and filling in the contents. This indicates that the suttas in SN (as also in AN) were, as a general rule, not targetted at outsiders or even at the newly converted, but were intended principally for those who had already turned for refuge to the Dhamma and were deeply immersed in its study and practice.

On the basis of its thematic arrangement, we might postulate that, in its most distinctive features as a collection (though certainly not in all particulars), SN was compiled to serve as the repository for the many short but pithy suttas disclosing the Buddha’s radical insights into the nature of reality and his unique path to spiritual emancipation. This collection would have served the needs of two types of disciples within the monastic order. One were the doctrinal specialists, those monks and nuns who were capable of grasping the deepest dimensions of wisdom and took upon themselves the task of clarifying for others the subtle perspectives on reality opened up by the Buddha’s teachings. Because SN brings together in its major saṃyuttas the many abstruse, profound, and delicately nuanced suttas on such weighty topics as dependent origination, the five aggregates, the six sense bases, the factors of the path, and the Four Noble Truths, it would have been perfectly suited for those disciples of intellectual bent who delighted in exploring the deep implications of the Dhamma and in explaining them to their spiritual companions. The second type of disciples for whom SN seems to have been designed were those monks and nuns who had already fulfilled the preliminary stages of meditative training and were intent on consummating their efforts with the direct realization of the ultimate truth. Because the suttas in this collection are vitally relevant to meditators bent on arriving at the undeceptive “knowledge of things as they really are,” they could well have formed the main part of a study syllabus compiled for the guidance of insight meditators.

With the move from SN to AN, a shift in emphasis takes place from comprehension to personal edification. Because the shorter suttas that articulate the philosophical theory and the main structures of training have found their way into SN, what have been left for inclusion in AN are the short suttas whose primary concern is practical. To some extent, in its practical orientation, AN partly overlaps with SN Mahāvagga, which treats the various groups of path factors. To avoid unnecessary duplication the redactors of the canon did not include these topics again in AN under their numerical categories, thereby leaving AN free to focus on those aspects of the training not incorporated in the repetitive sets. AN also includes a notable proportion of suttas addressed to lay disciples, dealing with the mundane, ethical, and spiritual concerns of life within the world. This makes it especially suitable as a text for the edification of the laity.

From this way of characterizing the two Nikayas, we might see SN and AN as offering two complementary perspectives on the Dhamma, both inherent in the original teaching. SN opens up to us the profound perspective reached through contemplative insight, where the familiar consensual world of persons and things gives way to the sphere of impersonal conditioned phenomena arising and perishing in accordance with laws of conditionality. This is the perspective on reality that, in the next stage in the evolution of Buddhist thought, will culminate in the Abhidhamma. Indeed, the connection between SN and the Abhidhamma appears to be a close one, and we might even speculate that it was the nonsubstantialist perspective so prominent in SN that directly gave rise to the type of inquiry that crystallized in the Abhidhamma philosophy. The close relationship between the two is especially evident from the second book of the Pāli Abhidhamma Piṭaka, the Vibhaṅga, which consists of eighteen treatises each devoted to the analysis of a particular doctrinal topic. Of these eighteen, the first twelve have their counterparts in SN. Since most of these treatises include a “Suttanta Analysis” (suttantabhājaniya) as well as a more technical “Abhidhamma Analysis” (abhidhammabhājaniya), it is conceivable that the Suttanta Analyses of the Vibhaṅga were the primordial seeds of the Abhidhamma and that it was among the specialists in SN that the idea arose of devising a more technical expository system which eventually came to be called the Abhidhamma.

The Aṅguttara Nikāya serves to balance the abstract philosophical point of view so prominent in SN with an acceptance of the conventional world of consensual realities. In AN, persons are as a rule not reduced to mere collections of aggregates, elements, and sense bases, but are treated as real centres of living experience engaged in a heartfelt quest for happiness and freedom from suffering. The suttas of this collection typically address these needs, many dealing with the practical training of monks and a significant number with the everyday concerns of lay followers. The numerical arrangement makes it particularly convenient for use in formal instruction, and thus it could be easily drawn upon by senior monks when teaching their pupils and by preachers when preparing sermons for the lay community. AN is replete with material that serves both purposes, and even today within the living Theravāda tradition it continues to fulfil this dual function.

The preceding attempt to characterize each Nikāya in terms of a ruling purpose should not be understood to imply that their internal contents are in any way uniform. To the contrary, amidst a welter of repetition and redundancy, each displays enormous diversity, somewhat like organisms of the same genera that exhibit minute specific differences absolutely essential to their survival. Further, it remains an open question, particularly in the case of SN and AN, whether their blueprints were drawn up with a deliberate pedagogical strategy in mind or whether, instead, the method of arrangement came first and their respective tactical applications followed as a matter of course from their groundplans.

Relationship with Other Parts of the Canon

Due partly to the composition of the suttas out of blocks of standardized, transposable text called pericopes, and partly to common points of focus throughout the Sutta Piṭaka, a considerable amount of overlapping can be discovered between the contents of the four Nikāyas. In the case of SN, parallels extend not only to the other three Nikāyas but to the Vinaya Piṭaka as well. Thus we find three SN suttas of great importance also recorded in the Vinaya Mahāvagga, represented as the first three discourses given by the Buddha at the dawn of his ministry: the Dhammacakkappavattana, the Anattalakkhaṇa, and the Ādittapariyāya (56:11; 22:59; 35:28). In the Vinaya, too, there are parallels to the SN suttas on the Buddha’s encounters with Māra (4:4, 5), on his hesitation to teach the Dhamma (6:1), on his first meeting with Anāthapiṇ˜ika (10:8), on the secession of Devadatta (17:35), and on the tormented spirits seen by Mahāmoggallāna (19:1–21). While it is possible that both the Vinaya and SN received this material via separate lines of oral transmission, in view of the fact that the narrative portions of the Vinaya Piṭaka appear to stem from a later period than the Nikāyas, we might conjecture that the redactors of the Vinaya drew freely upon texts preserved by the Saṃyutta reciters when composing the frameworks for the disciplinary injunctions.

SN includes as individual suttas material which, in DN, is embedded in larger suttas. The most notable instances of this are segments of the Mahāparinibbāna Sutta (e.g., at 6:15; 47:9; 47:12; 51:10), but we find as well a few snippets shared by the Mahāsatipaṭṭhāna Sutta (47:1, 2; 45:8) and a short (cÒḷa) version of the Mahānidāna Sutta (12:60). The latter shares with its larger counterpart (DN No. 15) only the opening paragraph but thereafter diverges in a completely different direction. Again, any solution to the question of borrowing can only be hypothetical.

The compilers of the canon seem to have laid down stringent rules governing the allocation of texts between SN and AN, intended to avoid extensive reduplication when a doctrinal theme is also a numerical set. Still, within the bounds set by that condition, a certain amount of overlapping has taken place between the two Nikāyas. They hold in common the suttas on Rohitassa’s search for the end of the world (2:26), on the lion’s roar (22:78), on the ten qualities of the stream-enterer (12:41 = 55:28), on the death of Kokālika (6:9–10), on the five hindrances (46:55, but in AN without the section on the enlightenment factors), as well as several large blocks of text that in SN do not constitute separate suttas.

It is, however, between SN and MN that the boundary appears to have been the most permeable, for SN contains five whole suttas also found in MN (22:82; 35:87, 88, 121; 36:19), as well as the usual common text blocks. We cannot know whether this dual allocation of the suttas was made with the general consent of the redactors responsible for the whole Sutta Piṭaka or came about because the separate companies of reciters responsible for the two Nikāyas each thought these suttas fitted best into their own collections. But in view of the fact that in SN several suttas appear in two saṃyuttas, thus even in the same Nikāya, the first alternative is not implausible. Suttas from SN have also found their way into the smaller works of the Khuddaka Nikāya—the Suttanipāta, the Udāna, and the Itivuttaka—while the correspondence between verses is legion, as can be seen from Concordance 1 (B).

Literary Features of the Saṃyutta

Of the four Nikāyas, SN seems to be the one most heavily subjected to “literary embellishment.” While it is possible that some of the variations stemmed from the Buddha himself, it also seems plausible that many of the more minute elaborations were introduced by the redactors of the canon. I wish to call attention to two distinctive features of the collection which bear testimony to this hypothesis. We might conveniently call them “template parallelism” and “auditor-setting variation.” The texts that exhibit these features are collated in Concordances 3 and 4 respectively. Here I will explain the principles that lie behind these editorial devices and cite a few notable examples of each.

Template parallels are suttas constructed in accordance with the same formal pattern but which differ in the content to which this pattern is applied. The template is the formal pattern or mould; the template sutta, a text created by applying this mould to a particular subject, the “raw material” to be moulded into a sutta. Template parallels cut across the division between saṃyuttas and show how the same formula can be used to make identical statements about different categories of phenomena, for example, about the elements, aggregates, and sense bases (dhātu, khandha, āyatana), or about path factors, enlightenment factors, and spiritual faculties (maggaṅga, bojjhaṅga, indriya). The recurrence of template parallels throughout SN gives us an important insight into the structure of the Buddha’s teaching. It shows that the teaching is constituted by two intersecting components: a formal component expressed by the templates themselves, and a material component provided by the entities that are organized by the templates. The application of the templates to the material components instructs us how the latter are to be treated. Thus we are made to see, from the template suttas, that the constituent factors of existence are to be understood with wisdom; that the defilements are to be abandoned; and that the path factors are to be developed.

The templates are in turn sometimes subsumed at a higher level by what we might call a paradigm, that is, a particular perspective offering us a panoramic overview of the teaching as a whole. Paradigms generate templates, and templates generate suttas. Thus all one need do to compose different suttas is to subject various types of material to the same templates generated by a single paradigm.

SN abounds in examples of this. One prevalent paradigm in the collection, central to the Dhamma, is the three characteristics of existence: impermanence (anicca), suffering (dukkha), and nonself (anattā). This paradigm governs whole series of suttas both in SN 22 and SN 35, the royal saṃyuttas of Parts III and IV, respectively; for it is above all the five aggregates and the six pairs of sense bases that must be seen with insight in order to win the fruits of liberation. The “three characteristics paradigm” generates four common templates: impermanent, etc., in the three times; the simple contemplation of impermanence, etc.; impermanent, etc., through causes and conditions; and, most critical in the Buddha’s soteriological plan, the “what is impermanent is suffering” template, which sets the three characteristics in relation to one another.

Another major paradigm is the triad of gratification, danger, and escape (assāda, ādınava, nissaraṇa), which generates three templates. At AN I 258–60 we find these templates used to generate three suttas in which the material content is the world as a whole (loka). SN, apparently drawing upon certain ways of understanding the concept of the world, contains twelve suttas churned out by these templates—three each in the saṃyuttas on the elements and the aggregates (14:31–33; 22:26–28), and six in the saṃyutta on the sense bases (35:13–18; six because the internal and external sense bases are treated separately). This paradigm is in turn connected to another, on the qualities of true ascetics and brahmins, and together they give birth to three more recurrent templates on how true ascetics and brahmins understand things: by way of the gratification triad; by way of the origin pentad (the gratification triad augmented by the origin and passing away of things); and by way of the noble-truth tetrad (modelled on the Four Noble Truths: suffering, its origin, its cessation, and the way to its cessation). These templates generate suttas on the four elements, gain and honour, the five aggregates, feelings, and the faculties. The last template is also applied several times to the factors of dependent origination, but strangely they are all missing in the Saḷāyatanasaṃyutta.

The main cause of suffering, according to the Buddha, is craving (taṇhā), also known as desire and lust (chanda-rāga). In SN the task of removing craving serves as a paradigm which generates another set of templates, arrived at by splitting and then recombining the terms of the compound: abandon desire, abandon lust, abandon desire and lust. These are each connected separately to whatever is impermanent, whatever is suffering, and whatever is nonself (intersecting with the three characteristics paradigm), thereby giving rise to nine templates. These are then extended to the aggregates and to the internal and external sense bases, generating respectively nine and eighteen suttas (22:137–45; 35:168–85).

Some templates must have emerged from the conversations into which the monks were drawn in their everyday lives, such as the one based on the question why the holy life is lived under the Blessed One (35:81, 152; 38:4; 45:5, 41–48). Part V, on the groups pertaining to the path, employs still new templates, though without a single dominant paradigm. Many of the templates occur in the repetition series, which are elaborated in full only in the Maggasaṃyutta and thereafter abbreviated in mnemonic verses. But more substantive templates generate suttas in the bodies of these saṃyuttas, which will be discussed at greater length in the introduction to Part V.

If we closely inspect the concordance of template parallels, we would notice that certain templates are not employed to generate suttas in domains where they seem perfectly applicable. Thus, as noted above, we do not find the “ascetics and brahmins” templates applied to the six sense bases, or the “noble and emancipating” template applied to the five spiritual faculties, or the “seven fruits and benefits” template applied to the four establishments of mindfulness. This raises the intriguing question whether these omissions were made by deliberate design, or because the applications were overlooked, or because suttas got lost in the process of oral transmission. To arrive at cogent hypotheses concerning this question we would have to compare the Pāli recension of SN with the Chinese translation of the Saṃyuktāgama, which would no doubt be a major undertaking requiring a rare combination of skills.

The second distinctive editorial technique of SN is what I call “auditor-setting variation.” This refers to suttas that are identical (or nearly identical) in content but differ in regard to the person to whom they are addressed, or in the protagonist involved (in a sutta involving a “plot”), or in the circumstances under which they are spoken. The most notable example of this device is the sutta on how a bhikkhu attains or fails to attain Nibbāna, which occurs seven times (at 35:118, 119, 124, 125, 126, 128, 131), in exactly the same words, but addressed to different auditors, including the deva-king Sakka and the gandhabba Pañcasikha. As the Buddha must have reiterated many suttas to different inquirers, the question arises why this one was selected for such special treatment. Could it have been a way of driving home, to the monks, what they must do to win the goal of the holy life? Or were there more mundane motives behind the redundancy, such as a desire to placate the families of important lay supporters?

Under this category fall several instances where a sutta is spoken by the Buddha a first time in response to a question from Ānanda, a second time to Ānanda on his own initiative, a third time in response to a question from a group of bhikkhus, and a fourth time to a group of bhikkhus on his own initiative (e.g., 36:15–18; 54:13–16). Again, the Rādhasaṃyutta includes two vaggas of twelve suttas each identical in all respects except that in the first (23:23–34) Rādha asks for a teaching while in the second (23:35–46) the Buddha takes the initiative in speaking.

A third literary embellishment, not quite identical with auditor-setting variation, is the inclusion of chains of suttas that ring the permutations on a simple idea by using different phrasing. Thus the Diṭṭhisaṃyutta (24) contains four “trips” (gamana) on speculative views differing only in the framework within which the exposition of views is encased (partial exception being made of the first trip, which for some unclear reason lacks a series of views included in the other three). In the Vacchagottasaṃyutta (33), the wanderer so named approaches the Buddha five times with the same question, about the reason why the ten speculative views arise in the world, and each time the answer is given as not knowing one of the five aggregates; each question and answer makes a separate sutta. Not content with this much, the compilers of the canon seem to have felt obliged to make it clear that each answer could have been formulated using a different synonym for lack of knowledge. Thus the saṃyutta is built up out of ten variants on the first pentad, identical in all respects except for the change of synonyms. The Jhānasaṃyutta (34) exhibits still another literary flourish, the “wheel” (cakka) of permutations, whereby a chain of terms is taken in pairwise combinations, exhausting all possibilities.

Technical Notes

Here I will discuss a few technical matters pertaining to the translation, emphasizing particularly why my renderings here sometimes differ from those used in MLDB. For the sake of precision, I usually refer to SN by volume, page, and line numbers of Ee (Ee1 in references to Part I), and use the saṃyutta and sutta numbers only when the whole sutta is relevant.

The Repetitions

Readers of the Pāli suttas are invariably irked, and sometimes dismayed, by the ponderous repetitiveness of the texts. In SN these are more blatant than in the other Nikāyas, even to the extent that in whole vaggas the suttas might differ from one another only in regard to a single word or phrase. Besides this type of reiterative pattern, we also come across the liberal use of stock definitions, stereotyped formulas, and pericopes typical of the Nikāyas as a whole, stemming from the period when they were transmitted orally. It is difficult to tell how much of the repetition stems from the Buddha himself, who as an itinerant teacher must have often repeated whole discourses with only slight variations, and how much is due to zealous redactors eager to ring every conceivable change on a single idea and preserve it for posterity. It is hard, however, not to suspect that the latter have had a heavy hand in the redaction of the texts.

To avoid excessive repetitiveness in the translation I have had to make ample use of elisions. In this respect I follow the printed editions of the Pāli texts, which are also highly abridged, but a translation intended for a contemporary reader requires still more compression if it is not to risk earning the reader’s wrath. On the other hand, I have been keen to see that nothing essential to the original text, including the flavour, has been lost due to the abridgement. The ideals of considerateness to the reader and fidelity to the text sometimes make contrary demands on a translator.

The treatment of repetition patterns in which the same utterance is made regarding a set of items is a perpetual problem in translating Pāli suttas. When translating a sutta about the five aggregates, for example, one is tempted to forgo the enumeration of the individual aggregates and instead turn the sutta into a general statement about the aggregates as a class. To my mind, such a method veers away from proper translation towards paraphrase and thus risks losing too much of the original text. My general policy has been to translate the full utterance in relation to the first and last members of the set, and merely to enumerate the intermediate members separated by ellipsis points. Thus, in a sutta about the five aggregates, I render the statement in full only for form and consciousness, and in between have “feeling … perception … volitional formations …,” implying thereby that the full statement likewise applies to them. With the bigger sets I often omit the intermediate terms, rendering the statement only for the first and last members.

This approach has required the frequent use of ellipsis points, a practice which also invites criticism. Several consulting readers thought I might improve the aesthetic appearance of the page (especially in Part IV) by rephrasing repetitive passages in a way that would eliminate the need for ellipsis points. I accepted this suggestion in regard to repetitions in the narrative framework, but in texts of straight doctrinal exposition I adhered to my original practice. The reason is that I think it an important responsibility of the translator, when translating passages of doctrinal significance, to show exactly where text is being elided, and for this ellipsis points remain the best tool at hand.


Rather than embark on the quest for a single English rendering that can capture all the meanings of this polyvalent Pāli word, I have settled for the more pragmatic approach of using different renderings intended to match its different applications. When the word denotes the Buddha’s teaching, I have retained the Pāli “Dhamma,” for even “teaching” fails to convey the idea that what the Buddha teaches as the Dhamma is not a system of thought original to himself but the fundamental principles of truth, virtue, and liberation discovered and taught by all Buddhas throughout beginningless time. This is the Dhamma venerated by the Buddhas of the past, present, and future, which they look upon as their own standard and guide (see 6:2). From an internal “emic” point of view, the Dhamma is thus more than a particular religious teaching that has appeared at a particular epoch of human history. It is the timeless law in which reality, truth, and righteousness are merged in a seamless unity, and also the conceptual expression of this law in a body of spiritual and ethical teachings leading to the highest goal, Nibbāna, which is likewise comprised by the Dhamma. The word “Dhamma,” however, can also signify teachings that deviate from the truth, including the erroneous doctrines of the “outside” teachers. Thus the Jain teacher Nigaṇṭha Nātaputta is said to “teach the Dhamma to his disciples” (IV 317,25)—certainly not the Buddha’s teaching.

In one passage I render Dhamma as “righteousness” (at the Se counterpart of IV 303,21). This is in the epithet dhammarājā used for a universal monarch, where “king of righteousness” fits better than “king of the Dhamma,” the significance the epithet has relative to the Buddha. The corresponding adjective, dhammika, is “righteous.”

When dhamma occurs as a general term of reference, often in the plural, I usually render it “things.” As such, the word does not bear the narrow sense of concrete material objects but includes literally every-thing, such as qualities, practices, acts, and relationships. Thus the four factors of stream-entry are, as dhammas, things; so too are the twelve factors of dependent origination, the five aggregates, the six pairs of sense bases, and the diverse practices leading to enlightenment. Used in the plural, dhammā can also mean teachings, and so I render it at III 225,9 foll., though the exact sense there is ambiguous and the word might also mean the things that are taught rather than the teachings about them. One expression occurring in two suttas (II 58,3–4; IV 328,21–22), iminā dhammena, can be most satisfactorily rendered “by this principle,” though here dhamma points to the Dhamma as the essential teaching. Again, at I 167,9 (= I 168,25, 173,10), we have dhamme sati, “when this principle exists,” a rule of conduct followed by the Buddha.

When plural dhammā acquires a more technical nuance, in contexts with ontological overtones, I render it “phenomena.” For instance, paṭicca-samuppannā dhammā are “dependently arisen phenomena” (II 26,7), and each of the five aggregates is loke lokadhamma, “a world-phenomenon in the world” that the Buddha has penetrated and taught (III 139,22 foll.). When the word takes on a more psychological hue, I render it “states.” The most common example of this is in the familiar pair kusalā dhammā, wholesome states, and akusalā dhammā, unwholesome states (found, for example, in the formula for right effort; V 9,17–27). The enlightenment factor dhammavicaya-sambojjhaṅga is said to be nurtured by giving careful attention to pairs of contrasting mental states (among them wholesome and unwholesome states; V 66,18), and thus I render it “the enlightenment factor of discrimination of states.” But since the dhammas investigated can also be the four objective supports of mindfulness (V 331–32), dhammavicaya might have been translated “discrimination of phenomena.” Sometimes dhammā signifies traits of character more persistent than transient mental states; in this context I render it “qualities,” e.g., Mahākassapa complains that the bhikkhus “have qualities which make them difficult to admonish” (II 204,3–4).

As a sense base and element, the dhammāyatana and dhammadhātu are the counterparts of the manāyatana, the mind base, and the manoviññāṇadhātu, the mind-consciousness element. The appropriate sense here would seem to be that of ideas and mental images, but the commentaries understand dhammas in these contexts to include not only the objects of consciousness but its concomitants as well. Thus I translate it “mental phenomena,” which is wide enough to encompass both these aspects of experience. As the fourth satipaṭṭhāna, objective base of mindfulness, dhammā is often translated “mind-objects.” So I rendered it in MLDB, but in retrospect this seems to me unsatisfactory. Of course, any existent can become an object of mind, and thus all dhammas in the fourth satipaṭṭhāna are necessarily mind-objects; but the latter term puts the focus in the wrong place. I now understand dhammas to be phenomena in general, but phenomena arranged in accordance with the categories of the Dhamma, the teaching, in such a way as to lead to a realization of the essential Dhamma embodied in the Four Noble Truths.

Finally, -dhamma as a suffix has the meaning “is subject to” or “has the nature of.” Thus all dependently arisen phenomena are “subject to destruction, vanishing, fading away, and cessation” (khayadhamma, vayadhamma, virāgadhamma, nirodhadhamma; II 26,9 foll.). The five aggregates are “of impermanent nature, of painful nature, of selfless nature” (aniccadhamma, dukkhadhamma, anattadhamma; III 195–96).


In MLDB I had changed Ven. Ñāṇamoli’s experimental rendering of saṅkhārā as “determinations” back to his earlier choice, “formations.” Aware that this word has its own drawbacks, in preparing this translation I had experimented with several alternatives. The most attractive of these was “constructions,” but in the end I felt that this term too often led to obscurity. Hence, like the land-finding crow which always returns to the ship when land is not close by (see Vism 657; Ppn 21:65), I had to fall back on “formations,” which is colourless enough to take on the meaning being imparted by the context. Sometimes I prefixed this with the adjective “volitional” to bring out the meaning more clearly.

Saṅkhārā is derived from the prefix saṃ (= con), “together,” and the verb karoti, “to make.” The noun straddles both sides of the active-passive divide. Thus saṅkhāras are both things which put together, construct, and compound other things, and the things that are put together, constructed, and compounded.

In SN the word occurs in five major doctrinal contexts:

(1) As the second factor in the formula of dependent origination, saṅkhāras are the kammically active volitions responsible, in conjunction with ignorance and craving, for generating rebirth and sustaining the forward movement of saṃsāra from one life to the next. Saṅkhārā is synonymous with kamma, to which it is etymologically related, both being derived from karoti. These saṅkhāras are distinguished as threefold by their channel of expression, as bodily, verbal, and mental (II 4,8–10, etc.); they are also divided by ethical quality into the meritorious, demeritorious, and imperturbable (II 82,9–13). To convey the relevant sense of saṅkhārā here I render the term “volitional formations.” The word might also have been translated “activities,” which makes explicit the connection with kamma, but this rendering would sever the connection with saṅkhārā in contexts other than dependent origination, which it seems desirable to preserve.

(2) As the fourth of the five aggregates, saṅkhārā is defined as the six classes of volitions (cha cetanākāyā, III 60,25–28), that is, volition regarding the six types of sense objects. Hence again I render it volitional formations. But the saṅkhārakkhandha has a wider compass than the saṅkhārā of the dependent origination series, comprising all instances of volition and not only those that are kammically active. In the Abhidhamma Piṭaka and the commentaries the saṅkhārakkhandha further serves as an umbrella category for classifying all mental concomitants of consciousness apart from feeling and perception. It thus comes to include all wholesome, unwholesome, and variable mental factors mentioned but not formally classified among the aggregates in the Sutta Piṭaka.

(3) In the widest sense, saṅkhārā comprises all conditioned things, everything arisen from a combination of conditions. In this sense all five aggregates, not just the fourth, are saṅkhāras (see III 132,22–27), as are all external objects and situations (II 191,11–17). The term here is taken to be of passive derivation—denoting what is conditioned, constructed, compounded—hence I render it simply “formations,” without the qualifying adjective. This notion of saṅkhārā serves as the cornerstone of a philosophical vision which sees the entire universe as constituted of conditioned phenomena. What is particularly emphasized about saṅkhāras in this sense is their impermanence. Recognition of their impermanence brings insight into the unreliable nature of all mundane felicity and inspires a sense of urgency directed towards liberation from saṃsāra (see 15:20; 22:96).

(4) A triad of saṅkhāras is mentioned in connection with the attainment of the cessation of perception and feeling: the bodily formation, the verbal formation, and the mental formation (IV 293,7–28). The first is in-and-out breathing (because breath is bound up with the body); the second, thought and examination (because by thinking one formulates the ideas one expresses by speech); the third, perception and feeling (because these things are bound up with the mind). Two of these terms—the bodily formation and the mental formation—are also included in the expanded instructions on mindfulness of breathing (V 311,21–22; 312,4–5).

(5) The expression padhānasaṅkhārā occurs in the formula for the four iddhipādas, the bases for spiritual power. The text explains it as the four right kinds of striving (V 268,8–19). I render it “volitional formations of striving.” Though, strictly speaking, the expression signifies energy (viriya) and not volition (cetanā), the qualifier shows that these formations occur in an active rather than a passive mode.

Apart from these main contexts, the word saṅkhāra occurs in several compounds—āyusaṅkhāra (II 266,19; V 262,22–23), jıvitasaṅkhāra (V 152,29–153,2) bhavasaṅkhāra (V 263,2)—which can be understood as different aspects of the life force.

The past participle connected with saṅkhārā is saṅkhata, which I translate “conditioned.” Unfortunately I could not render the two Pāli words into English in a way that preserves the vital connection between them: “formed” is too specific for saṅkhata, and “conditions” too wide for saṅkhārā (and it also encroaches on the domain of paccaya). If “constructions” had been used for saṅkhārā, saṅkhata would have become “constructed,” which preserves the connection, though at the cost of too stilted a translation. Regrettably, owing to the use of different English words for the pair, a critically important dimension of meaning in the suttas is lost to view. In the Pāli we can clearly see the connection: the saṅkhāras, the active constructive forces instigated by volition, create and shape conditioned reality, especially the conditioned factors classified into the five aggregates and the six internal sense bases; and this conditioned reality itself consists of saṅkhāras in the passive sense, called in the commentaries saṅkhata-saṅkhārā.

Further, it is not only this connection that is lost to view, but also the connection with Nibbāna. For Nibbāna is the asaṅkhata, the unconditioned, which is called thus precisely because it is neither made by saṅkhāras nor itself a saṅkhāra in either the active or passive sense. So, when the texts are taken up in the Pāli, we arrive at a clear picture in fine focus: the active saṅkhāras generated by volition perpetually create passive saṅkhāras, the saṅkhata dhammas or conditioned phenomena of the five aggregates (and, indirectly, of the objective world); and then, through the practice of the Buddha’s path, the practitioner arrives at the true knowledge of conditioned phenomena, which disables the generation of active saṅkhāras, putting an end to the constructing of conditioned reality and opening up the door to the Deathless, the asaṅkhata, the unconditioned, which is Nibbāna, final liberation from impermanence and suffering.


In MLDB, I also had changed Ven. Ñāṇamoli’s “name-and-form” back to his earlier rendering, “mentality-materiality.” In some respects the latter is doctrinally more accurate, but it is also unwieldly, particularly when translating verse, and thus here I return to “name-and-form.” The compound was of pre-Buddhistic origins and is used in the Upani˝ads to denote the differentiated manifestation of brahman, the nondual reality. For the sages of the Upani˝ads, nāmarÒpa is the manifestation of brahman as multiplicity, apprehended by the senses as diversified appearances or forms, and by thought as diversified names or concepts (the assignment of names and concepts being understood as grounded in objective reality rather than as the end-product of a purely subjective process). The Buddha adopted this expression and invested it with a meaning consonant with his own system. Here it becomes the physical and cognitive sides of individual existence. In the expression bahiddhā nāmarÒpa, “external name-and-form” (at II 24,2), we seem to find a vestige of the original meaning—the world as distinguished according to its appearances and names—but divested of the monistic implications.

In the Buddha’s system, rÒpa is defined as the four great elements and the form derived from them. Form is both internal to the person (as the body with its senses) and external (as the physical world). The Nikāyas do not explain derived form (upādāya rÒpaṃ), but the Abhidhamma analyses it into some twenty-four kinds of secondary material phenomena which include the sensitive substances of the sense faculties and four of the five sense objects (the tactile object is identified with three of the great elements—earth, heat, and air—which each exhibit tangible properties). Though I render nāma as name, this should not be taken too literally. Nāma is the assemblage of mental factors involved in cognition: feeling, perception, volition, contact, and attention (vedanā, saññā, cetanā, phassa, manasikāra; II 3,34–35). These are called “name” because they contribute to the process of cognition by which objects are subsumed under conceptual designations.

It should be noted that in the Nikāyas, nāmarÒpa does not include consciousness (viññāṇa). Consciousness is its condition, and the two are mutually dependent, like two sheaves of reeds leaning one against the other (II 114,17–19). Consciousness can operate only in dependence on a physical body (rÒpa) and in conjunction with its constellation of concomitants (nāma); conversely, only when consciousness is present can a compound of material elements function as a sentient body and the mental concomitants participate in cognition. Occasionally the texts speak of the “descent of consciousness” (viññāṇassa avakkanti) serving as a condition for name-and-form (II 91,14–15); this means that the arrival of the current of consciousness from the past existence into the new one is the necessary condition for the arising of a new psychophysical organism at conception. Sometimes too the texts speak of the descent of name-and-form (nāmarÒpassa avakkanti, II 66,12, 90,19, 101,13); this denotes the beginning of sentient life when the current of consciousness, arriving from the previous existence, becomes established under the fresh conditions.

Nibbāna, Parinibbāna

As is well known, nibbāna literally means the extinction of a fire. In popular works on Buddhism, nibbāna plain and simple is often taken to signify Nibbāna as experienced in life, parinibbāna Nibbāna attained at death. This is a misinterpretation. Long ago E.J. Thomas pointed out (possibly on the basis of a suggestion by E. Kuhn) that the prefix pari- converts a verb from the expression of a state into the expression of the achievement of an action, so that the corresponding noun nibbāna becomes the state of release, parinibbāna the attaining of that state. The distinction does not really work very well for the verb, as we find both parinibbāyati and nibbāyati used to designate the act of attaining release, but it appears to be fairly tenable in regard to the nouns. (In verse, however, we do sometimes find nibbāna used to denote the event, for example in the line pajjotass’ eva nibbānaṃ at v. 612c.) Words related to both nibbāna and parinibbāna designate both the attaining of release during life through the experience of full enlightenment, and the attaining of final release from conditioned existence through the breakup of the physical body of death. Thus, for instance, the verb parinibbāyati is commonly used to describe how a bhikkhu achieves release while alive (e.g., at II 82,20; III 54,3; IV 23,8–9, etc.) and also to indicate the passing away of the Buddha or an arahant (e.g., at I 158,23; V 161,25).

The past participle forms, nibbuta and parinibbuta, are from a different verbal root than the nouns nibbāna and parinibbāna. The former is from nir + , the latter from nir + . The noun appropriate to the participles is nibbuti, which occasionally occurs in the texts as a synonym for nibbāna but with a function that is more evocative (of tranquillity, complete rest, utter peace) than systematic. (It seems no prefixed noun parinibbuti is attested to in Pāli.) At an early time the two verb forms were conflated, so that the participle parinibbuta became the standard adjective used to denote one who has undergone parinibbāna. Like the verb, the participle is used in apposition to both the living Buddha or arahant (I 1,21, 187,8) and the deceased one (I 122,13, 158,24). Possibly, however, parinibbuta is used in relation to the living arahant only in verse, while in prose its technical use is confined to one who has expired. In sutta usage, even when the noun parinibbāna denotes the passing away of an arahant (particularly of the Buddha), it does not mean “Nibbāna after death.” It is, rather, the event of passing away undergone by one who has already attained Nibbāna during life.

The suttas distinguish between two elements of Nibbāna: the Nibbāna element with residue (sa-upādisesa-nibbānadhātu) and the Nibbāna element without residue (anupādisesa-nibbānadhātu)—the residue (upādisesa) being the compound of the five aggregates produced by prior craving and kamma (It 38–39). The former is the extinction of lust, hatred, and delusion attained by the arahant while alive; the latter is the remainderless cessation of all conditioned existence that occurs with the arahant’s death. In the commentaries the two elements of Nibbāna are respectively called kilesaparinibbāna, the quenching of defilements at the attainment of arahantship, and khandhaparinibbāna, the quenching of the continuum of aggregates with the arahant’s demise. Though the commentaries treat the two Nibbāna elements and the two kinds of parinibbāna as interchangeable and synonymous, in sutta usage it may be preferable to see the two kinds of parinibbāna as the events which give access to the two corresponding Nibbāna elements. Parinibbāna, then, is the act of quenching; nibbāna, the state of quenchedness.

To explain the philology of a term is not to settle the question of its interpretation. What exactly is to be made of the various explanations of Nibbāna given in the Nikāyas has been a subject of debate since the early days of Buddhism, with the ground divided between those who regard it as the mere extinction of defilements and cessation of existence and those who understand it as a transcendental (lokuttara) ontological reality. In SN some suttas explain Nibbāna as the destruction of lust, hatred, and delusion, which emphasizes the experiential psychological dimension; elsewhere it is called the unconditioned, which seems to place the stress on ontological transcendence. The Theravāda commentators regard Nibbāna as an unconditioned element. They hold that when Nibbāna is called the destruction of the defilements (of lust, hatred, and delusion, etc.) and the cessation of the five aggregates, this requires interpretation. Nibbāna itself, as an existent, is unborn, unmade, unbecome, unconditioned (see Ud 80–81). It is in dependence on this element (taṃ āgamma), by arriving at it, that there takes place the destruction of the defilements and release from conditioned existence. Nibbāna itself, however, is not reducible to these two events, which are, in their actual occurrence, conditioned events happening in time. On this interpretation, the two Nibbāna elements are seen as stages in the full actualization of the unconditioned Nibbāna, not simply as two discrete events.

In the present work I leave nibbāna untranslated, for the term is too rich in evocative meaning and too defiant of conceptual specification to be satisfactorily captured by any proposed English equivalent. I translate parinibbāna as “final Nibbāna,” since the noun form usually means the passing away of an arahant (or the Buddha), final release from conditioned existence; sometimes, however, its meaning is ambiguous, as in the statement “the Dhamma [is] taught by the Blessed One for the sake of final Nibbāna without clinging (anupādāparinibbānatthaṃ)” (IV 48,78), which can mean either Nibbāna during life or the full cessation of existence.

The verb parinibbāyati perhaps could have been incorporated into English with “nibbanize,” which would be truest to the Pāli, but this would be too much at variance with current conventions. Thus when the verb refers to the demise of the Buddha or an arahant, I render it “attains final Nibbāna,” but when it designates the extinguishing of defilements by one who attains enlightenment, I render it simply “attains Nibbāna.” We also find a personal noun form, parinibbāyı, which I render “an attainer of Nibbāna,” as it can be construed in either sense. In prose the past participle parinibbuta, used as a doctrinal term, always occurs with reference to a deceased arahant and so it is translated “has attained final Nibbāna.” In verse, it can take on either meaning; when it describes a living arahant (or the Buddha) I translate it more freely as “fully quenched.” The unprefixed form nibbuta does not always carry the same technical implications as parinibbuta, but can mean simply “peaceful, satisfied, at ease,” without necessarily establishing that the one so described has attained Nibbāna. At I 24,11 and II 279,8 it has this implication; at I 236,21 it seems to mean simply peaceful; at III 43, in the compound tadaṅganibbuta, it definitely does not imply Nibbāna, for the point there is that the monk has only approximated to the real attainment of the goal. Cognates of parinibbāna appear in colloquial speech with a nondoctrinal sense; for example, both parinibbāyati and parinibbuta are used to describe the taming of a horse (at MN I 446,8–10). But even here they seem to be used with a “loaded meaning,” since the horse simile is introduced to draw a comparison with a monk who attains arahantship.

Other Changes

In MLDB I rendered vitakka and vicāra respectively as “applied thought” and “sustained thought.” In this translation they become “thought” and “examination.” The latter is surely closer to the actual meaning of vicāra. When vitakka is translated as “thought,” however, a word of caution is necessary. In common usage, vitakka corresponds so closely to our “thought” that no other rendering seems feasible; for example, in kāmavitakka, sensual thought, or its opposite, nekkhammavitakka, thought of renunciation. When, however, vitakka and vicāra occur as constituents of the first jhāna, they do not exercise the function of discursive thinking characteristic of ordinary consciousness. Here, rather, vitakka is the mental factor with the function of applying the mind to the object, and vicāra the factor with the function of examining the object nondiscursively in order to anchor the mind in the object.

Bhava, in MLDB, was translated “being.” In seeking an alternative, I had first experimented with “becoming,” but when the shortcomings in this choice were pointed out to me I decided to return to “existence,” used in my earlier translations. Bhava, however, is not “existence” in the sense of the most universal ontological category, that which is shared by everything from the dishes in the kitchen sink to the numbers in a mathematical equation. Existence in the latter sense is covered by the verb atthi and the abstract noun atthitā. Bhava is concrete sentient existence in one of the three realms of existence posited by Buddhist cosmology, a span of life beginning with conception and ending in death. In the formula of dependent origination it is understood to mean both (i) the active side of life that produces rebirth into a particular mode of sentient existence, in other words rebirth-producing kamma; and (ii) the mode of sentient existence that results from such activity.

Sakkāya is a term for the five aggregates as a collective whole (III 159,10–13). The word is derived from sat + kāya, and literally means “the existing body,” the assemblage of existent phenomena that serve as the objective basis of clinging. Most translators render it “personality,” a practice I followed in MLDB (departing from Ven. Ñāṇamoli, who rendered it, too literally in my view, “embodiment”). But since, under the influence of modern psychology, the word “personality” has taken on connotations quite foreign to what is implied by sakkāya, I now translate it as “identity” (a suggestion made to me by Ven. Thanissaro Bhikkhu). Sakkāya-diṭṭhi accordingly becomes “identity view,” the view of a self existing either behind or among the five aggregates.

Nibbidā, in MLDB, was translated “disenchantment.” However, the word or its cognates is sometimes used in ways which suggest that something stronger is intended. Hence I now translate the noun as “revulsion” and the corresponding verb nibbindati as “to experience revulsion.” What is intended by this is not a reaction of emotional disgust, accompanied by horror and aversion, but a calm inward turning away from all conditioned existence as comprised in the five aggregates, the six sense bases, and the first noble truth. Revulsion arises from knowledge and vision of things as they really are (yathābhÒtañāṇadassana), and naturally leads to dispassion (virāga) and liberation (vimutti; on the sequence, see 12:23).

How to cite this document:
© Bhikkhu Bodhi, The Connected Discourses of the Buddha (Wisdom Publications, 2000)

Creative Commons License
This selection from The Connected Discourses of the Buddha by Bhikkhu Bodhi is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.
Based on a work at
Permissions beyond the scope of this license may be available at