“Theravāda Buddhism” seems to be a transparent and straightfor-ward term. It is taken for granted as an integral feature of the religious landscape not only of South and Southeast Asia, but also of contem-porary Buddhism in the West. The term is regularly used without any attempt at definition, and without asking to what degree “Theravāda Buddhism” is a valid or useful category.2 A chapter entitled “Theravāda Buddhism” in one recent book uses the word “Theravāda” and cog-nate forms forty-one times in about seven pages of text (not counting captions and side-bars).3 Is there anything surprising in this? Perhaps not: but when we consider that the term Theravāda is rare in Pali lit-erature, and that for nearly a millennium it was rarely used in the Pali or vernacular inscriptions, chronicles, or other premodern texts of Southeast Asia, this might give us pause.
This essay is written on the premise that we—historians of Bud-dhism—do not adequately understand, and have not adequately at-tempted to understand, the term “Theravāda.” Nonetheless, we have imposed it in our studies to create, in many cases, artificial and ahis-torical entities. I believe that we need to reexamine the evidence, to see how the complex of historical movements within Theravāda define and refer to themselves, and to see how they define and refer to others, both Buddhist nikāyas and other religions.4 I take it as axiomatic that the history of Theravāda cannot be written on the basis of Pali sources alone. We must exploit the full range of sources, including not only the vernaculars of the cultures in which Theravāda has developed from a
monastic aggregation into distinctive social complexes, but also the classical Buddhist languages—Sanskrit, Tibetan, and Chinese—and other vernaculars used by Buddhist traditions. We must take epi-graphic, literary, archaeological, iconographical, and anthropological evidence into account.
As a working hypothesis, I suggest that “Theravāda Buddhism”
came to be distinguished as a kind of Buddhism or as a “religion”—re-membering that “Buddhism” is a modern term and that “religion” is a vexed concept—only in the late colonial and early globalized periods, that is, in the twentieth century. In the pre-colonial and early colonial periods, Europeans grouped non-Christian religions in several ways, and Buddhism was subsumed with Indian, Mongolian, Chinese, and Japanese religions under the categories of “idolaters” or “heathens.”
Eventually European savants, for the most part working in milieux that were deeply charged by Christian beliefs and presuppositions, realized that the religious life of certain groups or societies was centered on
“Buddha.” Gradually they saw that this religion, which they eventu-ally came to call Buddhism, had different forms or schools. Exactly when “Theravāda Buddhism” gained currency as a discrete category, and how this category in due course entered the consciousness of
“Theravādin societies” themselves, is a good question, but it seems that it was rather late in the game—that is, only by the middle of the twentieth century.5 We need to bear these points point in mind if we are to understand the history of Theravāda.
Theravāda has now become a standard and authoritative term, de-fined (or mis-dede-fined) even in computer dictionaries. It is a ready-made label that gives us “Theravādin meditation,” “Theravādin philosophy,”
“Theravādin psychology,” “Theravādin art,” “Theravādin iconogra-phy,” and so on. As a type of Buddhism, the very idea of Theravāda is a by-product of globalization. With increased international migra-tion in the second half of the twentieth century, Sinhalese, Burmese, Khmer, Lao, and Thai communities have had to construct identities in a multicultural world, and so have become “Theravāda Buddhists.”
Beyond this, we live in an age in which packaging and labeling are es-sential to the social constitution of both the individual and the group.
Ambiguity is not tolerated. Today Theravāda is a self-conscious iden-tity for many, although not necessarily in the societies that have nur-tured it for centuries. For most Thai, for example, the primary marker of identity remains to be “Buddhist” (pen phut, pen chao phut, naptü
phra phutthasāsanā). To say “to believe in Theravāda” (naptü therawāt) is unnatural.
The evolution of Theravāda as a modern religion is not the concern of this paper. My concern is the use of Theravāda as a historical cat-egory. The problem is this: the word “Theravāda” and cognate forms or near-synonyms “Theriya” and “Theravaṃsa” are infrequent in Pali texts. Their use is limited to some of the commentaries and to the historical (vaṃsa) literature—literature specifically concerned with school formation and legitimation. But in the pre-modern period, what we call the Theravāda Buddhists of Southeast Asia did not seem to use the term at all. It was neither a marker of identity nor a standard of authority in the inscriptions or chronicles of the region. Nor does the term occur in the early European accounts of the religion and society of the region, whether Portuguese, Dutch, French, or English. Simply put, the term was not part of the self-consciousness of the Buddhists of the region.
The overuse of the term Theravāda in historical studies has led to several misconceptions. One is the idea that there existed some sort of monolithic religion, or institutional entity, called “Theravāda,”
that spread throughout the region. This obscures the fact that the Southeast Asian sanghas that renewed their ordination lineages in Sri Lanka were, as soon as they returned to their own lands, autonomous or rather independent entities. They invoked their Lankan credentials as a claim to ritual purity, but they did not maintain binding institu-tional links with Lanka. The new lineages established their own identi-ties; more often than not within one or two generations they fell into dispute and split into further independent lines.
This state of affairs arises from the nature of ordination, of the independent system of self-reproduction of Buddhist monastic com-munities. Higher ordination (upasampadā) can be performed without reference to any outside authority (except, depending on circum-stances, temporal authority, which made efforts to control sanghas for political and economic reasons). The only conditions were the pres-ence of a monk qualified to act as preceptor (upajjhāya, who must have a minimum ten years’ ordination) and a quorum of monks to perform higher ordination. It is, precisely, our task to understand how ordina-tion lineages spread, how they defined themselves, how they related to other lineages in the region, and how they contended with each other and with temporal powers for recognition and patronage.
The history of “Theravāda” is, then, a history of ordination lin-eages. It is not a history of “sects” in the sense of broad-based lay groups, as in Reformation Europe. Monks, rulers, and lay supporters were concerned with establishing or restoring pure ordination lin-eages in order to sustain the life of the sāsana by activating pure “fields of merit” and ensuring the continuity of ritual. The records show little concern for ideas or philosophy. It is this spread of monastic ideals and lineages that we must try to understand, usually through indirect evidence, since despite the importance of claimed descent, lineage re-cords were not maintained or constructed.
Our study must maintain an awareness of the monastic/lay dis-tinction. How should we understand relations between laity and lin-eage? Even if the monastics were Theravādin, or better (see below) belonged to a Sīhalavaṃsa or some other lineage, can we say this of the laity? To what degree did the laity participate in the distinctions and contentions of monastic lineages? What range of ideas or activi-ties do monastic lineages embrace? How far are they relevant to the social and religious lives of the laity? Traditionally, could categories like Theravāda or Sīhalapakkha apply to laypeople at all? What are the boundaries in terms of individual or society?
The preservation, transmission, and study of the Pali canon and the use of Pāli as a liturgical language—by monastics and laity—is one distinctive and unifying feature of the Theravādin lineages.6 But the use of Pali should not overshadow vernacular literature and practice.
If Pali was a resource, a database, that offered stability and continuity to a congeries of constantly evolving traditions, it was the vernacular transformations of the Pali—through sermon, gloss, bilingual recita-tion, and the plastic arts—that enabled what Steven Collins has called
“the Pali imaginaire” to function as a vital agent in the religious life of mainland Southeast Asia.7
Defining Theravāda
What is Theravāda? If we describe it as a system, what do we mean?
A system of thought? A system of ethics? A monastic infrastructure, an economic institution, or a soteriological framework? If it is several or all of these, how does it differ from other Buddhist systems? Buddhism:
The Illustrated Guide defines Theravāda as follows: “Of the many distinc-tive schools of Buddhism that formed in the first centuries after the death of the Buddha, only one has survived into the present day—the
Theravāda or ‘Doctrine of the Elders.’ The followers of this tradition trace its origins back to Gautama Buddha himself. They maintain that the Buddha’s teaching has been handed down in an unbroken succes-sion within the Sangha or monastic community, hence the reference to
‘elders’ or venerable members of the Sangha who have protected the tradition’s integrity.”8 Unfortunately, the definition misses the point.
The term “thera” does not refer to a lineal succession of “elders,” but to a specific “historical” or foundational group: the five hundred arhats who recited and collected the teachings of the Buddha at Rājagṛha after the first rains-retreat following the death of the Buddha.9 This is stated, for example, in the Dīpavaṃsa: “because the collection was made by the theras, therefore it is called the Theravāda.”10 For a more accurate definition, we may turn to Ven. Payutto (Prayudh Payutto, Phra Brahmaguṇābharaṇa, 1938–). After describing the events of the First Rehearsal, he writes: “The teachings thus agreed upon that have been handed down to us are called Theravāda or ‘the teachings laid down as principles of the Elders.’ The word Elders in this context refers to those 500 Arahant elders participating in this First Rehearsal. The Buddhism that is based on the First Rehearsal mentioned above is called Theravāda Buddhism. In other words, the Buddha’s teachings, namely the Doctrine and Discipline, both in letter and in spirit, that were thus rehearsed were to be remembered as such and strictly adhered to.”11 One problem with this traditional definition is that all of the eighteen Buddhist schools trace their origins back to the First Council, which is their common heritage. Each of the surviving Vinayas and other re-cords preserved in several languages presents its own version of the
“Council of the Five-Hundred.” How then, do the Ceylon theras differ?12 Indo-Tibetan Perspectives
There are several Pali terms for what we call “Theravāda,” includ-ing “Theriya” and “Theravaṃsa.”13 Before examining them, however, we will turn to India, to see how the later north Indian Buddhists de-scribed the “Theravāda.”
What did the Indian Buddhist schools call what we call
“Theravāda,” and how did they present it? To start with, they did not call it “Sthaviravāda.” “Sthaviravāda” and “Sthaviravādin” are ghost words. They are Sanskrit neologisms coined on the analogy of Pali
“Theravāda,” and they have not been found in any Sanskrit text.14 The word “sthavira” does occur, but as a technical Vinaya term for a senior
monk, defined as one who has been ordained as a bhikṣu for ten years or more (in Pali thera, in Hybrid Sanskrit also sthera). The term is used by all Buddhist schools, and is not a marker of “Theravāda.” It occurs in compounds like saṃgha-sthavira and rāja-sthavira, or as a prefix to a proper name, as—taking examples from Indian inscriptions—in Sthavira Pūrṇadāsa, Sthavira Mahānāma, or Sthavira Acala.15 The com-pound “Siṃghala-sthavira,” found in an inscription from Bodh-Gayā, is ambiguous, although my own interpretation is that it refers to a
“Sinhalese senior monk.”16
The Sthāvira lineage is regularly referred to in Indian doxographic works that were translated into Tibetan and Chinese, but no Sanskrit versions survive. Therefore we cannot say with certainty what Sanskrit terms lie behind the translations. The few available references in Sanskrit suggest a vṛddhi form: Sthāvira,17 Sthāviriya,18 or Sthāvarī.19 These are Sanskrit counterparts of “Theriya,” a common Pali and Prakrit counterpart to “Theravāda.” In this essay, I tentatively adopt the form Sthāvira.
In his History of the Dharma, Bu-ston Rin-chen-grub (1290–1364) de-fines the Sthāviras as follows:
gnas brtan ’phags pa’i rigs yin par smra bas gnas brtan pa.20
Because they assert that they belong to the noble lineage (āryavaṃśa) of the Sthaviras they are *Sthāviras.
I am not certain this tells us much—the ārya-vaṃśa seems to be, again, a shared quality of Buddhist monastic heritage.21
The Four Vinaya Schools and the Four Philosophical Schools
A persistent problem in the modern historiography of Buddhism in India is a denial of tradition—a refusal to try to understand how Indian Buddhism looked at itself. This strikes me as odd. To believe in or to accept tradition is one thing; but to achieve historical understand-ing, one must first try to see how Buddhism presented itself, whether or not one believes it or accepts it as historically accurate or viable. A priori rejection because the information goes against received opinions or because a source is preserved only in Tibetan amounts to ideologi-cally motivated blindness.
In India, classifications of Buddhism depended on context. In terms of Vinaya, there were the “four nikāyas”: Sarvāstivāda, Sthāvira, Sāṃmitīya, and Mahāsāṃghika. This classification subsumed the
traditional lists of the (conventionally enumerated) eighteen nikāyas.
The fourfold classification was widespread in north India by the sev-enth century, if not earlier, and is reported by Yijing (635–713) in his
“Record of the Inner Law Sent Home from the South Seas” and in other sources.22 The four schools are mentioned in tantras and tantra com-mentaries such as the Hevajra-tantra and the Yogaratnamālā on the Hevajra-tantra.23
When the four schools are explained in detail, three branches are listed under Sthāvira.24 These are given in Indian sources from the eighth century on preserved in Tibetan, such as Vinītadeva’s Nikāyabheda, Subhūtighoṣa’s Sarvayānālokakāra-vaibhāṣya, the Śrāmaṇera-pṛcchā, and the Bhikṣuvarṣāgra-pṛcchā,25 as well as in the great Sanskrit-Tibetan lex-icon of the late eighth century, the Mahāvyutpatti. The names of the three schools derive from the monasteries in Anurādhapura, Sri Lanka, at which their sanghas were based:
Mahāvihāravāsin Residents of the Great Monastery Abhayagirivāsin Residents of the Abhayagiri Monastery Jetavanīya Residents of the Jetavana Monastery.
North Indian accounts agree with the indigenous tradition of Sri Lanka.
The Mahāvaṃsa refers to the three schools in several places as “the three nikāyas” (nikāyattaya); in the twelfth century they were unified by King Parākramabāhu.26 Even if the nature of the unification and the role of the Mahāvihāra vis-à-vis the other schools are debated, we can safely say that before the twelfth century, the Mahāvihāra was not the sole representative of “Theravāda,” and that after the twelfth century the surviving “Theravāda” was only one branch of the earlier Lankan school. The idea of “Theravāda” as an unchanging and perennial lin-eage contradicts the school’s own history (as well as common sense).
North Indian tradition as preserved in Tibet defines each of the four schools in terms of a fixed set of categories. Gorampa (1429–1489) explains that “among the four root Śrāvaka schools there are four different Vinayas, four different languages, four different precep-tors, four different numbers of panels of the monk’s robe, and four different insignia on the edge of the robe.”27 In his “Sun of the Land of Samantabhadra,” published in 1699, ’Jam-dbyaṅs Bźad-pa’i-rdo-rje (Ṅag-dbaṅ-brtson-’grus, 1648–1721/2) describes the Sthāviras as follows:
gnas brtan pa ni | skad bi śa tsi śa za’i skad dam ’briṅ du ’don pa | mkhan po rje rigs smyig ma mkhan mtha’ ’khob ’dul ba’i mchog tu luṅ bstan pa katya ya na nas | snam phran sṅa ma ltar daṅ | grva rtags ’khor lo miṅ lha daṅ |
’byuṅ gnas daṅ | go cha ste | rin chen ’byuṅ gnas daṅ | śes rab go cha lta bu yin te | ṅes brjod las |
gnas brtan pa la go cha lha ||
sde daṅ ’tsho daṅ skyoṅ ba’o | źes so ||28
The language of the Sthāviras is “Paiśāci,” the “language of the flesh-eaters” (piśāca), or the “intermediate recitation.” Their [first]
preceptor is the Vaiśya bamboo-maker (veṇukāra) Kātyāyana, who was declared [by the Blessed One] to be “foremost among those who convert the border regions.” The robe [snam = snam sbyar = saṃghāṭī]
has [from twenty-one to twenty-five panels (khaṇḍa)] like that of the preceding [Sāṃmatīyas].29 The insignia on the edge of the robe is the wheel (cakra). The names [of monastics end in] -deva, -ākara, or -varman, as for example Ratnākara or Prajñāvarman. According to the ṅes brjod:
The [names of] Sthāviras [end in] varman, deva sena, rakṣita, and pāla.
In the Pali tradition, Kaccāna or Mahākaccāyana, a direct disciple of the Buddha, is a brahman, as his gotra name suggests. It is possible that the reference here is to the enigmatic author of the Nettippakaraṇa, but the gotra name raises the same problem. None of the other information can be confirmed. If there is any substance to it, the passage—transmitted in various Tibetan sources, but not yet located in an Indian text—may describe the Sthāviras (and other schools) according to north India ste-reotypes of the Pāla period. Sakya Paṇḍita (1182–1251) summarizes the tradition in his A Clear Differentiation of the Three Codes:
The four fundamental communities of the Disciples had four distinct codes of discipline,
and their canonical languages, too, were four:
Sanskrit, Prākrit, Apabhraṃśa, and Paiśācī.
The eighteen schools that developed therefrom had eighteen distinct codes of discipline, because all these schools differed in their procedures—for accepting vows in the beginning; for observing them, repairing them, and reciting the Prātimokṣa meanwhile; and, finally, for renouncing them.
What one prohibited is permitted for another.30
The Four Philosophical Schools
North Indian tradition grouped Buddhist philosophy under four main schools: Vaibhāṣika, Sautrāntika, Cittamātra/Vijñānavādin, and Mādhyamika. This classification was sufficiently current to be adopted in the classical Indian doxography, the Sarvadarśanasaṃgraha,31 and in other non-Buddhist texts,32 and to be discussed in tantric and Kālacakra literature.33 Mimaki suggests that the classification was developed by the end of the eighth century;34 it was widely adopted in Tibet where it became standard in any number of doxographic manuals.35 ’ Jam-dbyaṅs Bźad-pa’i-rdo-rje is somewhat dogmatic about the number of philosophical schools:
raṅ sde bye smra mdo sems dbu ma ba, bźir ṅes ’dir ni grub mtha‘ lṅar min gsuṅs.
Regarding the tenets of our own [that is, the Buddhist] schools, our own schools are limited to the four: Vaibhāṣika, Sautrāntika, Cittamātra, and Mādhyamika, for it is said that here there are not five systems of tenets.36
This statement reflects Tibetan scholastic politics rather than the situ-ation in India. Earlier Tibetan works in the genre listed the schools in several ways, one of which names five schools, adding the Sāṃmitīyas.
Despite the fact that it is equally valid in terms of the historical evolu-tion of Buddhist thought, the fivefold classificaevolu-tion did not find favor in Tibetan scholasticism.
The association of the four philosophical schools with the four Vinaya lineages is problematic. The Vaibhāṣikas and Sautrāntikas are usually considered to be philosophical movements within the Sarvāstivādin lineage.37 The fundamental Yogācāra works rely on Sarvāstivādin texts, while Mādhyamika is not associated with any par-ticular Vinaya tradition. The tenets laid down in the doxographic lit-erature, particularly those of the Vaibhāṣikas and Sautrāntikas, may well have existed largely in the textbooks by the time the later manu-als were composed. In any case, the Sthāvira tradition is conspicuous by its absence. How do we explain this? Why did north Indian tradi-tion not recognize the Sthāviras as a philosophical school, even when it knew it as a Vinaya lineage?
In the absence of evidence, no certain answer can be given. The four-school classification appears to be a curriculum list and may reflect the interests of the professors of Nālandā or other monastic universities