25 September 2020

Early Buddhist Heterodoxy: Pudgalavāda

I've been reading the excellent book chapter on pudgalavāda "the person doctrine" by Amber Carpenter (pictured). Carpenter  (2015) summarises a coherent, emergentist philosophy and it's clear to me that Buddhism is much the poorer for the disappearance of such heterodox thinking. There was a time when Buddhists disagreed about fundamentals and this produced both innovative thinking and bracing polemics. Even so, the polemics acknowledge that those who espoused pudgalavāda were Buddhists. The aim was to refute and convert, not to exclude.

It is worth clarifying at the outset that pudgalavāda is not the name of a sectarian lineage like the Theravāda (hence, I do not capitalise it). Rather, it was a view espoused by members of a range of such lineages but somewhat independently of them. Thus, it does not make sense to refer to "Pudgalavādins" in the way that we refer to Theravādins. Pudgalavāda is a philosophical view of certain issues that Carpenter draws out.

If we take a typical introduction to discussion of pudgalavāda we learn that:
The [pudgalavāda] doctrine holds that the person, in a certain sense, is real. To other Buddhists, their view seemed to contradict a fundamental tenet of Buddhism, the doctrine of non-self. — Internet Encyclopedia of Philosophy.
This view subtly takes the orthodox view and treats pudgalavāda as heterodox and it sets up a false dichotomy. This is partly because most accounts of pudgalavāda are drawn from polemics that seem to deliberately take a mistaken view of what is meant by pudgala, i.e. that it is an ātman substitute. It is emphatically not an ātmavāda and it is not in any way connected to the idea of a permanent self. So when a source like this links pudgalavāda to the anātman doctrine this is a strawman. 

Although Carpenter highlights karma in her title, the role of karma in the story of pudgalavāda is down-played in a way that I found a little frustrating, because of my own project. Rather than review Carpenter, I will simply recommend the chapter to anyone interested in early Buddhist thought. It is well written and cogently argued and makes a good deal of sense of what might otherwise be a puzzling development in Buddhist thought. What I aim to do here is reframe Carpenter's observations in terms that better suit my own purposes.

In particular, I want to put karma and problems with the early Buddhist doctrine of karma at the forefront and pursue my thesis that it was problems integrating other doctrines with the karma doctrine rather than, say, problems with anātmavāda, that motivated the development of early Buddhist sectarianism. I have long argued, for example, that there is a conflict between karma and the doctrine of dependent arising which does not allow for continuity over time or for consequences manifesting long after their condition has ceased to exist. And I've also argued that in resolving the conflict, Buddhists always seek to resolve it in favour of karma. Even Nāgārjuna, who deduces, based on unchallenged presuppositions, that karma cannot be ultimately real, ends up accepting that karma is still operative at the level of conventional reality. No one seems to question whether this is a coherent worldview, let alone whether it is a coherent view of karma (in my view the answer is "no" in both cases). 

Framing Buddhist sectarianism as based on philosophical disagreements over the nature of selfhood is commonplace. Everyone seems to assume that disagreements in which (an)ātman is discussed must be based in speculations about the nature of ātman. And this is true even when only the secondary literature discusses ātman. But why, oh why, were Buddhists speculating about the nature of something that they explicitly did not believe in? This is not a question that writing on anātmanvāda seems to address. Whatever the answer, Carpenter points out that scripture was not able to resolve such conflicts since "the best evidence available—namely the discourses of the Buddha...—was both ambiguous and conflicting" (3). And yet, as she also says on the following pages, the conflict over pudgala was not over the anātman doctrine. Rather, proponents of pudgalavāda were at pains to affirm anātmavāda (5). The pudgala is emphatically not an ātman substitute. So if the pudgala was not ātman then what was it?

What is the Pudgala?

The pudgalavāda argues that the skandhas are not randomly distributed or related. Rather, there is a distinct arrangement of the skandhas associated with each individual. If Devadatta experiences a vedanā, it only happens to him. Yajñadatta does not have a cognition based on Devadatta's sense experience. If Devadatta and Yajñādatta are both looking at an orange and Devadatta closes his eyes, Devadatta cannot see the orange, but Yajñadatta still can. A dharma cognized by the manas of Devadatta is not cognized by the manas of Yajñadatta. This is a very obvious fact but one that seldom comes into play in mainstream Buddhism philosophy. Whatever you wish to say about ātman, at some level we are individuals having individual sense experiences that are not public, not shared, but private and subjective. 

Most Buddhist thought is relentlessly reductive. Early Buddhists sought reality through reductive analysis. Such methods result in knowledge about substances (dharma, dravya). The Abhidharma extended this to include substances and the relations (pratyāya) between them, but they still reduced everything to their catalogues of substances and relations. 

The whole discussion was made more complex by the realisation that metaphysical reductionism causes a problem. If only simple substances (dharmas) are real, then what is the nature of the complex world we interact with? The answer was sought in a doctrine that had originally described the arising of experiences in awareness: contact between sense organ and sense object gave rise to sense consciousness. This was now generalised to apply to reality in the broadest sense: everything arises in dependence on conditions.

This led towards a hierarchical construction of reality into ultimately real (paramārtha-sat) and conventionally real (saṃvṛti-sat). Initially, only substance was ultimately real and structures and systems (the macro world we interact with) were only conventionally real. The view is hierarchical because that which is conventionally real derives its existence from that which is ultimately real. Nāgārjuna abandoned the idea that substance is real and asserted that only absence (śūnyatā) is real and then complained when people called him a nihilist.

Carpenter makes the excellent point that by the time of Vasubandhu (upon whose polemics we often rely for information about heterodox Buddhists) paramārtha-sat and saṃvṛti-sat were routinely being conflated with another pair of concepts, i.e. substantially real (drayva-sat) and conceptually real (prajñapti-sat). In this view, conventionally real phenomena are merely conceptual (prajñapti). The world that we interact with as pṛthagjanas is just a bunch of concepts. 

However, it seems to me that the corollary—that what is ultimately real was the substantial (or objective)—was not so widely accepted. Indeed, it is this idea that is in the sights of both Nāgārjuna and Vasubandhu. As I've noted. Nāgārjuna, using deduction from a set of axioms, is forced into the conclusion that only absence is real. As I understand Vasubandhu, he would also reject that idea that ultimate reality can be conflated with substance. His response was to flirt with idealism (though his actual commitment to idealism is the subject of ongoing debates) in the form of the mind-only doctrine (cittamātravāda aka "doctrine of mere conceptions" vijñaptimātravāda). In this sense, Vasubandhu seems to lean toward connecting paramārtha-sat to prajñapti-sat.

This conflation of conventionally real and conceptually real should sound familiar because some version of it is present in almost all modern Buddhist teachings. We are taught that what appears to be real to the unenlightened turns out to be mere conceptions. And "insight" blows away the delusions so that we see reality as it is, without the overlay of conceptions.

Keeping in mind that pudgala is defined as an arrangement of the skandhas, where would it fit into these definitions of what is real according to pudgalavāda

The pudgala is truly real (paramārtha-sat) because it really exists, but it is not a substance (dravya), so that is only conceptually real in the sense that we conceive of the pudgala based an arrangement of the skandhas in a particular individual. Not being a substance, but being only a peculiar but real arrangement of skandhas means that the pudgala does not qualify as an ātman.

This combination of paramārtha-sat and prajñapti-sat was confounding for other Buddhists with their commitment to reductionism. Vasubandhu argued that only two possibilities exist: the pudgala could either be reduced to the really existent skandhas, or it could not, and amounted to an ātmavāda. And he rejected both alternatives. He could not conceive of a non-reductive reality. 

In modern terms, the pudgalavāda is an anti-reductive or emergentist argument. The pudgala is real because of the structure of the skandhas, not because of their substance. I also see parallels with John Searle's distinctions between ontological and epistemic truths and between objective and subjective truths. The pudgala is an ontologically subjective truth, but at the same time epistemically objective. The mode of existence of the pudgala is mental and thus subjective, but for anyone who is aware, they are aware of being someone and having some kind of continuity over time.

Given that Buddhists are bound to stipulate the anātmavāda, this is a considerably better philosophical answer to the problem of our experience of continuity than the two truths doctrine, which to my mind is incoherent. But why is this an important area of speculations and conjectures for early Buddhists? We can see, I hope, what the pudgala is, but what is the point of postulating such a thing? 

What Does the Pudgala Do?

Proponents of pudgalavāda do not seem to have been involved in speculations about self. They themselves denied it. They were dragged into discussion "self" by strawman arguments based on reductionist metaphysics. It was like they could not help but buy into the framing of the criticisms from other Buddhists. 

Rather, they seem to be thinking about continuity. And continuity is important because without it karma is meaningless. On the other hand, ātman is what provides the necessary continuity for karma in the late Vedic religion and since Buddhist arguments about karma and ātman often revolve around explicit rejection of Vedic theology, the concept of "self" always seems to creep into the discussion, whether we like it or not. 

Carpenter makes the point that the Brahmanical ātman is a convenient idea because it performs a number of roles. It is the agent (kārtṛ) of actions (karman) and it is the patient (bhoktṛ) of consequences (vipāka) (11-12). There is no doubt for the ancient Brahmin that the one who acts is the one who suffers the consequences. This is all well and good because without this karma does not make sense, and neither does morality, generally. In fact, "actions have consequences" is a modern Buddhist catchcry. 

By adopting dependent arising as a metaphysics, Buddhists sacrificed the continuity necessary to make karma coherent. But only theoretically. When discussing karma or morality Buddhists still acted as though continuity is a given. Nowhere is this more apparent that the fifth of the "Five Remembrances" from the Upajjhatthana Sutta (AN 5.57):
Kammassakomhi kammadāyādo kammayoni kammabandhū kammapaṭisaraṇo yaṃ kammaṃ karissāmi kalyāṇaṃ vā pāpakaṃ vā tassa dāyādo bhavissāmī 
I am the owner of my actions, the heir of my actions, born from my actions, bound to my actions, and find refuge in my actions. I will be the heir of whatever good or evil actions I do.
There is no doubt in my mind that these lines strongly imply continuity over time: either the kārtṛ and the bhoktṛ are the same person or the words don't make sense. And much the same is true throughout the Jātakas which are used to teach Buddhist morality in more traditional societies. These are stories of how the someone's actions in a past life are the cause of some event in their present life. The protagonists are all given definite identities in the past. Morality is not possible without this kind of continuity. 

At the same time Buddhists have long denied any continuous entity which could provide the continuity necessary for karma to work like this. The "self" (ātman) or "being" (satva) is analysed into components in a typically reductive method, exposing the complex object as insubstantial, but also going the extra mile and declaring that only substances can be real and that therefore complex objects, like a self, are not real. Modern Buddhists recite this without reflection on the applicability of reductionism. 

By breaking the reductionist mould, pudgalavāda found a way out of the dilemma of having morality and metaphysics that are inconsistent. The loss caused by the disappearance of this stream of thought from Buddhism is immeasurable. 

Without ever conceding that they intended any kind of ātman, or that anything like an ātman could be inferred from pudgalavāda, the argument proceeded like this. The skandhas are not randomly distributed. Experience is subjective and thus individualised. We do not have access to each other's experience, stories of extrasensory powers notwithstanding. 

If experience is and remains subjective, then each occurrence of the five skandhas as a group stays together as a group or else subjective experience would be incoherent. There is a persistent structure that is not drvaya-sat, i.e. not a real substance, but which is paramārtha-sat, i.e. truly real. They appear here to be groping towards an emergentist metaphysics by admitting that something other than a substance can be real. Although reductive ontologies are the current mainstream, the argument for only including substance and excluding structure/system from reality is an ideological argument. The argument for a reductive metaphysics is not inescapable or paramārtha-sat.

Each individual is a collection of skandhas. When Devadatta sees something, Yajñadatta doesn't experience a vedanā. Yajñadatta only experiences a vedanā when his own sense organs contact a sense object. The skandhas themselves are associated with each other over time, they are grouped together in this way. No other explanation fits with the subjectivity of experience. It is this association that is real and persistent. It is not a real and persistent substance and thus emphatically not an ātman. Moreover, it is not a "relation" (pratyāya) in the Abhidharma sense. It is this association of skandhas that we name pudgala in this context.

The important thing is that the pudgala provides the kind of continuity that karma demands and that dependent arising makes impossible. And because the association itself is not dealt with the reductive metaphysics of early Buddhists, it is not specifically denied. What's more there is some (admittedly minimal) textual support for this interpretation in the Bhāra Sutta (SN 22:22).
Katamo ca, bhikkhave, bhārahāro. Puggalo Tissa vacanīyaṃ. Yoyaṃ āyasmā evaṃnāmo evaṃgotto; ayaṃ vuccati, bhikkhave, bhārahāro. (SN III.26)
What is the bearer of the burden (bhāra-hāra)? It should be called “person” (puggala). This elder named thus and from a particular clan. Monks, this is called the bearer of the burden
And what is the burden (bhāra)? It is precisely the skandhas, or here the five branches that fuel existence (pañcupādānakkhandhā). This is only one of many texts in the Theravāda canon that contradict Theravāda orthodoxy. 


It seems unhelpful to think of the pudgalavāda as a speculative philosophy born from conjectures about the nature of the self. Rather, if I may say so, it fits better with my account of Buddhists being trapped in an invidious position of embracing the moral doctrine of karma and at the same time adopting a metaphysics which denied the reality of the mechanism by which karma works (i.e. continuity over time). Many ways out of this dilemma were proposed including the "always existence dharma theory" (sarvāstivāda) and the theory of momentary series of dharmas (kṣanavāda). Pudgalavāda is a lost solution to this problem that is far more interesting and promising than popular presentations would suggest. I think Amber Carpenter covered the problems with karma, but to me they deserve special emphasis. 

The pudgala refers to the specific arrangement of skandhas that makes up an individual. It must be individual because sense experience is subjective. This arrangement, this structure, as I would call it, is paramārtha-sat but not dravya-sat. In John Searle's (1995) terms it is ontologically subjective, but epistemically objective. Even though the experience is subjective, we know that it is real as opposed to, for example, an illusion, a hallucination, or a dream. A real experience is distinguishable from an unreal one in a number of ways: the unreal experience does not follow our explicit or implicit physical laws, in an unreal experience our senses do not concur with each other (e.g. something heard but not seen), and other observers do not concur on the existence of unreal experiences. Obviously, this distinction is not infallible. 

If karma visits the consequences of actions on the agent (kārtṛ), then the agent must persist in some way to become the patient (bhoktṛ) or the doctrine is incoherent. Buddhists scriptures themselves link the kārtṛ and bhoktṛ

Alone amongst all of Buddhism, as far as I know, the pudgalavāda is the only Buddhist doctrine to question the supremacy of reductionism. This alone makes it worthy of respect and interest.Since reading Richard Jones Analysis and the Fullness of Reality, I am personally interested in any philosophy which embraces the reality of structure or systems. It is a delight to find such a view in the unrelenting reductionism of both early and modern Buddhism.

I also cannot help but note that the Prajñāpāramitāvāda appears to avoid the metaphysical problems by staying inside the lines of epistemology. Rather than dabbling in speculative metaphysics, Prajñāpāramitā remains focussed on the phenomenology of meditation and in particular the state of absence of sense experience (śūnyatā). Where Nāgārjuna is sucked down the rabbit hole of asserting the ultimate reality of absence, the Prajñāpāramitā sūtras do not make this mistake. They do not follow the white rabbit. They know that what absence offers is not reality, but a deeper glimpse into the nature of subjectivity.

I'm still trying to counteract the disastrous misreadings of history so this is probably an overly generous view of Prajñāpāramitā. Because, of course, they were trying to make sense of Buddhist doctrines as well. Prajñāpāramitā is still at heart a soteriology, a way to escape from rebirth. The religious worldview of karma and rebirth is still the context in which the Prajñāpāramitā discourses take place.



Carpenter, Amber. 2015. "Persons Keeping their Karma Together: The Reasons for the Pudgalavāda in Early Buddhism." In The Moon Points Back. Edited by Koji Tanaka et al, 1-44. Oxford University Press.

Searle, John R. (1995). The Construction of Social reality. Penguin.

Amber Carpenter is an associate professor at Yale University. She is involved in the Integrity Project
Related Posts with Thumbnails