No Buddhist school has been more vilified by its Buddhist peers or misunderstood by modern scholars than the so-called pudgalavāda1 school. Other Buddhists accused them of violating the fundamental Buddhist tenet of no-self (anātman) by holding the view that a real ontological self exists that, their accusers argued, Pudgalavādins try to camouflage by calling it pudgala (person) rather than ātman (self). Modern scholars, forming opinions largely based on or influenced by the hostile polemical literature of the Pudgalavādins’ opponents, reiterate that accusation.2 In addition, until recently scholars considered Pudgalavādins to be a marginal sect, of minor historical and doctrinal influence, significant only for playing the role of reviled heretics. Even the term pudgalavāda, which scholars continue to use, appears to be a disparaging label foisted on them by their opponents, not a term they used to characterize themselves. However, both accusations—of promoting the idea of an ontological self and of being marginal—are directly contradicted by the surviving examples of the Pudgalavādins’ own literature and by a more judicious examination of the historical record.
Starting with the charge of marginality, it turns out that the Vātsīputrīyas (their actual name, taken from their founder) were one of the most popular mainstream Buddhist sects in India for more than a thousand years. Some traditional sources claim their origins go back to the time of the Buddha, though most scholars think that other sources assigning their beginning to the third century B.C.E. are more accurate. By the second century C.E. at the latest, they had subdivided into four distinct subschools, the most prominent and successful being the Saṃmitīyas (see Vasumitra’s Tenets, translated here). Two Chinese pilgrims who traveled to India in the seventh century, Xuanzang and Yijing, inform us in their travelogues that the Saṃmitīyas were to be found throughout India and even in Southeast Asia and the South Sea Islands. They were especially prominent in Western India, a region that also served as a travel route through which Buddhism flowed to the north out of India and into which Central and East Asian Buddhists came to study in Buddhism’s homeland.3
While their opponents—notably Vasubandhu’s Abhidharmakośa-bhāsṃya (chap. 9, Duerlinger 2003), Candrakīrti’s Madhyamakāvatara (chap. 6), and Śāntaraksṃita and Kamalaśīla’s Tattvasamgraha-pañjikā—accused them of promoting the idea of a “real” self, the handful of surviving Vātsīputrīya texts strenuously deny this, instead insisting that the pudgala is a prajñapti (only a nominal existent) that is neither identical to nor different from the skandhas. Since accusing a Buddhist opponent of harboring an ātmavāda view (view of eternal selfhood) is one of the most virulent accusations a Buddhist can lodge against another, and we find this trump card played in several other questionable situations—such as against the Yogācāras’ theory of an eighth consciousness, the ālaya-vijñāna, when clearly the Yogācāras have taken great pains to define the ālaya-vijñāna in ways that fully avoid that charge—we should be cautious about accepting such accusations on their face.
Prajñapti is a multivalent term that many Buddhist schools deployed in a variety of ways. Literally prajñapti means “leading to knowledge.” It can mean a teaching device, a designation, an instruction, a heuristic, a name or label for a complex of conditions, and so on.4 For example, Buddhists argued about whether things such as “karmic accrual” (prāpti) or “aging” are actual real things (dravya) or only nominalist labels (prajñapti) for complexes of causal processes. The causal processes would be real (dravya), the labels only conceptual shorthand (prajñapti). That the earliest Buddhist texts associated pudgala with prajñapti is clear not only from the prooftexts cited in Vātsīputrīya texts (which correspond to passages still found in the Theravāda Tipitṃaka), but from the title of the fourth text in the Pali Abhidhamma canon, Puggala-paññati (Skt. Pudgala-prajñapti), heuristically translated into English as Human Types (Law 1979).
The Vātsīputrīya argument is that the pudgala is a necessary prajñapti since any theory of karma, or any theory that posits that individuals can make spiritual progress for themselves or can assist other individuals to do likewise, is incoherent without it. Karma means that an action done at one time has subsequent consequences for the same individual at a later time, or even a later life. If the positive and negative consequences of an action don’t accrue to the self-same individual, then it would make no sense to speak of things like progress (who is progressing?), and Buddhist practice itself becomes incoherent. If there are no persons, then there is no one who suffers, no one who performs and reaps the consequences of his or her own karma, no Buddha, no Buddhists, and no Buddhism. Obviously, those are not acceptable consequences for a Buddhist.
Buddhists speak of skandha-upādāna,“the aggregates of appropriation,” which raises the obvious question: Who/what appropriates the five skand-has, collecting them into a single living entity? If the appropriator is something different from the skandhas themselves, then there is a sixth skandha, which is doctrinally impermissible. If the skandhas appropriate themselves, that leads to a vicious cycle of infinite regress. Hence, the Vātsīputrīyas argue, the nominal person (pudgala) is neither the same as nor different from the skandhas. It is a heuristic fiction that avoids these unwarranted consequences and lends coherence by also corresponding to how actual persons experience themselves—that is, as distinct individuals continuous with, but not absolutely identical to or reducible to, their own pasts and futures. Similarly, Buddhists speak of past and future lives. But what remains constant or continuous between such lives? If it is a self-same invariant identity, then this would indeed be a case of ātmavāda, a view the Vātsīputrīyas, like all Buddhists, reject. In what sense would someone be the same or different from the person in one’s previous life? If completely different, then to posit a continuity between them is incoherent. If the same, then their real discontinuities are ignored, leading to a form of eternalism, another impermissible view for Buddhists. Hence, they are neither the same nor different, but linked by a fictional pudgala. Finally, Buddhist practice leads to nirvana; but who attains this? If there is an integral individual that ceases on attaining nirvana, then this would entail the unwarranted view of annihilationalism. If there is no cessation of the karmic individual, then there is no nirvana. Both extremes, though implicit in standard Buddhist formulations, render Buddhism itself incoherent, a problem only solved, the Vātsīputrīyas argue, if one admits the fictional pudgala implicit in standard Buddhist doctrine.
A “fiction,” in this sense, does not simply mean something unreal, but rather, like any good work of fiction, something that does explain, in a non-literal way, something real, and that can move, inspire, elicit, and evoke meaningful thoughts and actions. The pudgala is that type of “fiction.” The self as permanent selfhood is unreal, but the experience of individual per-sonhood is a fiction everyone experiences.5 While for the Pudgalavādins there is no ontological “self” or permanent, substantial person, there is a fictitious “person” that is neither the same as nor different from the actual ontological processes accepted by all other Buddhists as “real” constituents of a being, namely, the skandhas, āyatanas, and so on. The three prajñaptis discussed in the passages are unavoidable fictions that not only provide doctrinal coherence; they also serve as refutations and correctives for insidious false views, such as eternalism and annihilationalism.
Though only a tiny portion of the Pudgalavādins’ vast literature has survived, we are fortunate to have two Chinese translations of what, at its core, was a single text. These are Si ahanmu chao jie (Commentary on the Four Āgamas), authored by *Vasubhadra,6 and translated by Kumārabuddhi in 382, and Sanfa du lun (Treatise on Liberation by the Threefold Teachings), also attributed to *Vasubhadra, translated by Gautama Sanṃghadeva in 391.7 For convenience, I will refer to these as the Longer Version and Shorter Version, respectively.8 Both Chinese renderings—as is unfortunately the case for many pre-fifth-century Chinese Buddhist translations—are difficult texts, with many problematic passages. These two versions also greatly differ from each other in wording, phrasing, semantic implications, the ordering of parts, and so on, with one or the other expanding at certain points through extended passages entirely absent from the other version. This will be obvious when comparing the two sections translated here, which are based on a single core passage.
Both versions describe three types of prajñapti. The Shorter Version appears first here, even though it was translated later, because it offers a more concise version of the passage. The Longer Version expands on several things, most notably the second type of prajñapti, which the Shorter Version describes only as “prajñapti of the past,” while the Longer Version renames this “prajñapti of metaphorical devices” and applies it to the “three times,” that is, past, present and future. Whether the differences represent different redactions of a root text, sectarian distinctions among the Vātsīputrīyas, or liberties taken by the translators is unclear. The key to understanding both versions is to see that all three prajñaptis have no other purpose than to avoid the hidden, “unsaid” presuppositions lurking in the doctrines held by other Buddhists; that while other Buddhists might leave the word “pudgala” unsaid, the narratives presupposed in their doctrines require it.
The issue the passage raises is not the affirmation of something that exists ineffably—as some modern scholars have assumed—but rather that Buddhists who talk about such things as skandhas-of-appropriation (skandha-upādāna), previous (and future) lives, and nirvana as entailing the cessation of the appropriation of skandhas—as all mainstream Buddhists do—are dabbling in “unsayables,” but they are not aware of that, and consider such discussion taboo. One of the two likely Sanskrit candidates for the term being translated into Chinese as “not-said” or “unsayable” is avācya, which means something “not to be addressed,” “improper to be uttered,” or “not distinctly expressed.” The other candidate is avaktavya, which also means something that should not to be said, but may also indicate something indescribable. The Vātsīputrīyas are using it in that double sense: the pudgala is a taboo subject for other Buddhists, even though the metaphysical narratives they employ presuppose it; and what is indicated by the prajñapti “pudgala” cannot be explained more precisely, since appropriation without an appropriator,
linkage across lives without an invariant identity, and the cessation of a nonself are intrinsically incoherent notions. The Vātsīputrīyas are offering a clever polemic that accuses other Buddhists of ignoring the “unsayables” in their own heuristic expressions. The “unsayable” of primary interest to the Vātsīputrīyas is the pudgala (person), which is a necessary nominal construction required to perform Buddhist analysis, progress on the Buddhist path, and make sense of the most basic Buddhist concepts, especially karma.
The pudgala is on the one hand merely a linguistic construction. On the other hand, it involves something in everyone’s experience about which we can say nothing definitively coherent. It is unreal (merely nominal), but experientially, even soterically, effective. As the passage makes clear, the Vātsīputrīyas are attempting to forge a middle way between extremes of eternalism and annihilationalism, existence and nonexistence, while affirming that, nevertheless, the pudgala as a prajñapti is an effective, if imprecise, way of talking about requisites for cultivating the Buddhist path.
The third selection here is the section on the Vātsīputrīyas from Vasumitra’s Tenets of the Different Schools based on Xuanzang’s Chinese translation.9 The Indic original is not extant but is available in one Tibetan and three Chinese translations. This text was probably composed around the second century C.E. Vasumitra gives lists of main tenets for eighteen Buddhist schools—several varieties of Mahāsānṃghikas, Sarvāstivāda, Prajñaptivāda, Haimavatas, Mahīśāsakas, and so on—presenting them in roughly chronological order with explanations of which schools splintered from which.10
Q: What is unsayable?
A: The unsayable is [what remains implicit in] the Figurative Expressions (prajñaptis) concerning appropriation, the past, and cessation11.
[These are] the figurative expressions (prajñapti) concerning appropriation, the figurative expressions concerning the past, and figurative expressions concerning cessation. If someone doesn’t know [what “appropriation,” “past (lives),” and “(nirvanic) cessation” entail,] then they don’t know the unsayable.12
Figurative expression concerning appropriation is the analysis of whether sentient beings are the same or different from the skandhas, dhātus, and āyatanas they have already appropriated.13
As to figurative expressions concerning the past, saying “[this person is now so-and-so] because of past skandhas, dhātus, and āyatanas” is just like saying “At that time [in a previous life,] I was named Kuśendra.”14
As to figurative expressions concerning cessation, saying “it is because appropriation has already ceased” is just like saying “The Bhagavat’s Parinirvānṃa.”
Moreover, [the purpose of these figurative expressions is to dispel false views]. The figurative expressions concerning the past dispel [the idea] that sentient beings are annihilated. Figurative expressions concerning cessation dispel [the idea] that they exist permanently. Figurative expressions concerning the appropriation [of skandhas, etc.] dispel [the idea that sentient beings are] nonexistent. Figurative expressions concerning nonappropriation dispel [the idea that sentient beings qua an eternal self] exist.
Q: What is not said (* avaktavya or *avācya)?
A: The not-said: [This refers to what is] not said [or left implicit by other Buddhists] in the heuristics (prajñapti) for appropriation, metaphorical devices, and cessation.
Those are the heuristics for appropriation, heuristics by metaphorical device,15 and the heuristics for cessation. This means that whoever is stupid concerning these “not saids” lacks insight (*ajñana).
The heuristic for appropriation [involves] using the term “a living-one” (jīva).16 [The idea] that the presently appropriated skandhas, dhātus, and āyatanas are appropriated by an inner living-one is a heuristic. This means that [when one talks about the] present appropriation by an inner living-one who appropriates dharmas17 due to karmic conditioning (saṃskāras) and the fetters (saṃyojana), these are heuristics for [discussing] appropriation.
The dharmas that the living-one heuristically appropriates are not the same as the living-one. It’s not as if one seeks to get the jīva and the body to combine [into a single thing. To do so would entail opposing extremist absurdities]. If they are the same, then [the jīva would be] impermanent and [prone to] suffering; if they are different, then a permanent [jīva] would be prone to suffering.
If it is permanent, one wouldn’t [need to] practice brahmacarya (a religious life). If it is not permanent, one would be unsuited for the brahmacarya fruit.18 For that which is impermanent, receiving and giving (i.e., meritorious activities) would be meaningless [since an impermanent being would perish before such activities could mature into fruition]. Meaninglessness is tantamount to nihilism; in these two metaphorical devices [of permanence and annihilation] there is no dharma[conducive to either] suffering or the favorable.
The heuristic by metaphorical device is naming.
[To speak of a person as being the same person in the] “past, future, and present” is to practice the heuristics by metaphorical device. This is the heuristic metaphorical device of naming [i.e., giving a single name to conditions that vary over time] which [linguistically posits] a relationship across the three times (past, present, and future). For example: “In the past, I was King *Kuśa” or “In the future you will have the name Ajita,” [or] “At present I am a prominent merchant,” and [other] such activities as were assumed [in the past] or have not yet been assumed [in the future]. Such conventional roles are numerous, hence they are heuristically adopted [by assuming the person undergoes] annihilation and permanence. [The extremist assumptions embedded in this are exposed by such questions as] If Kuśa has ceased, in what sense am I he? If he has not ceased, in what sense can one say he is I? It is by means of conventionalisms (vyavahāra) that one says so; it is a heuristic metaphorical device.
Q: What are the heuristics for cessation?
A: The heuristics for cessation [are statements such as] “appropriation is exhausted,” or “no [further] appropriation [will occur],” or “coming to rest.”
Appropriation is as explained above. Once that has been exhausted, [one says] “no [more] appropriation,” “no obtaining another [life],” “coming to rest,” “[nirvana] with no remainder,” “passing from this shore to the other shore”—these are heuristics for cessation.
The way a [being is usually thought of by other Buddhists,] as cycling through samsara, [implicitly presupposes the extremist views of] annihilation and permanence. If one seeks to stop such [samsaric] activities, one turns to the heuristics of appropriation and the heuristics of Parinirvānṃa. This (i.e., Parinirvānṃa), too, is a not-said.
If [the pudgala] is [intrinsically] different [from Parinirvānṃa], then one doesn’t [obtain] Parinirvānṃa. If [the pudgala] is not [inherently] different from [Parinirvānṃa], then one doesn’t [obtain] Parinirvānṃa.19
These kinds of views have given rise to suffering and have not been explained (“said”) [adequately by other Buddhists].20 [Such Buddhists] would [say] “Parinirvānṃa is like the ceasing of an internal lamp.” The same [applies to] appropriation. If one seeks [to understand] suffering and yet doesn’t clarify it with the heuristics of appropriation and metaphorical devices, [such as] past skandhas, dhātus, and āyatanas, basically this is like saying “I am named King Kuśa.” In such a way the heuristic of future cessation means that the cessation of appropriation is the main point to be explained.
These are tenets that the Vātsīputrīya schools hold in common:
The pudgala is neither the same [as] nor different from the skandhas. It is a
prajñapti dependent on the skandhas, āyatanas, and dhātus.
Saṃskāras (conditioned dharmas) have a temporary duration, while some cease in an instant (ksṃanṃika).
Dharmas, if apart from the pudgala, cannot move on from a previous lifetime to a subsequent lifetime. On the basis of the pudgala, one can say there is transference (saṃkrānti).
Moreover, even non-Buddhists can attain the five rṃddhis (superpowers).21
The five consciousnesses have no kleśas, and are not apart from kleśas.22
If the bonds (saṃyojana) of the Desire Realm (kāma-dhātu) are eliminated during the Cultivation Stage (bhāvanā-mārga), one is called “free from desire.” But not if eliminated during the Seeing Stage (darśana-mārga).23
It is by [the four wholesome roots, namely,] ksṃānti (forbearance), nāma (name), nimitta (image), and laukikā agra-dharmāhṃ (the highest meditative insight) that one can enter into the correct nature in which no mental defilements (kleśas) arise (niyāmāvākrānti or samyaktva-niyāma).24
If entering niyāmāvākrānti during the twelve mental moments, this is called “Going toward.”25 If during the thirteenth mental [moment], this is called “abiding in the fruit.”
In ways such as this, there are many different opinions.
Because of holding different interpretations of a single verse, this school branched off into four schools, which are Dharmottarīya (Higher Dharma), Bhadrayānṃīyas (Inheritance from the Honorables), Saṃmatīyas (Correct Measure), and the Sānṃdṃagirikas (Hidden in Forests and Mountains).
That verse says:
Already liberated, again one backslides
backsliding due to desire. Again returning
recovering peaceful joy and the place of happiness.
From the enjoyable (postepiphany life) to perfect happiness.26
Cousins, L. S. (1995) “Person and Self.” In Buddhism into the Year 2000. Bangkok: Dhammakaya Foundation, pp. 15–31. Reprinted in Paul Williams, ed., Buddhism: Critical Concepts in Religious Studies.2 London: Routledge, 2005, 2:84–101.
Duerlinger, James. (2003) Indian Buddhist Theories of Persons: Vasubandhu’s “Refutation of the Theory of a Self.” London: RoutledgeCurzon.
Hurvitz, Leon. (1967) “The Road to Buddhist Salvation as described by Vasubhadra.” Journal of the American Oriental Society 87/4: 434–486.
Lamotte, Étienne. (1988) History of Indian Buddhism. Louvain-la-Neuve: Peeters Press.
Law, B. C. (1979) Designation of Human Types (Puggala-paññatti). London: Pali Text Society.
Maclean, Derryl. (1989) Religion and Society in Arab Sind. Leiden: Brill.
Masuda, Jiryo (1925) “Origins and Doctrines of Early Indian Buddhist Schools: A Translation of the Hsüan-chwang Version of Vasumitra’s Treatise.” Asia Major 2: 1–78.
Priestley, Leonard. (1999) Pudgalavāda Buddhism: The Reality of the Indeterminate Self. Toronto: Center for South Asian Studies, University of Toronto.
Thich, Thien Chau. (1999) The Literature of the Personalists of Early Buddhism. Translated by Sara Boin-Webb. Delhi: Motilal Banarsidass.
Tsukamoto, Keishō (2004) The Cycle of the Formation of the Schismatic Doctrines. Berkeley: Numata.