A Brief Geneology of the "Three Realms/Three States"

[kelamuni]

The teaching of the "three states" of consciousness makes its first appearance in the Brhadaranyaka Upanishad. This work is considered to be the authoritative basis of the teaching. However, in its second appearance in the Chandogya Upanishad the teaching has undergone a change. Chan Up 8.11.1 marks the change and is an interesting passage. There, Indra, who is being instructed by Prajapati on the three states, notes a fault in a particular teaching, a teaching that can only be that of the Brhadaranyaka. The Brhad's teaching had been that in deep dreamless sleep one "becomes" brahman, or, to put it another more precise way, in deep sleep we revert back to our primordial nature as brahman. But Indra notes that in deep dreamless sleep there is no knowledge, since there is no object of consciousness in deep dreamless sleep. The implication is that if there can be no knowledge in such a state, there can be no awareness that one has become brahman. Thus, in the Chandogya Up, we find the beginnings of an appreciation of the problem of reflexivity, of the question of how it is that one can be aware that one is brahman when brahman is by nature non-dual. For the Brhad, such reflexivity, or awareness that one is brahman, is out of the question, for another teaching of the Brahd is that the Self is not an object of consciousness, but the pure subject, the "seer" (drashtr), or "witness" (sakshin), and "the eye that sees cannot see itself."

For Shankara, the inner Self (pratyag-atman) is also the pure or transcendent Subject (vishayin). He clearly sides with the Brhad Up on the matter of reflexivity: consciousness is no more capable of full reflexivity than a juggler can stand on his own shoulders or a knife cut itself. He does, however, admit a kind of reflexivity by saying that in enlightenment, or brahman-jnana, there is a "fruit" or effect of release, and that is the reflexive knowledge that one is released once brahman-jnana occurs.

Among the the Buddhists, the Madhyamikas agree with Shankara. The Yogacharins, however, hold that consciousness is capable of being aware of itself, and Mandana Mishra, Shankara's great Advaitin contemporary, agrees with them. Thus, we find a fault line running through both traditions, with some Buddhists holding one position and others another, and some Advaita Vedantins holding one position and others the counterposition. I think this fault line can ultmately be traced back to the difference between the teachings of the Brhadaranyaka and Chandogya Upanishads, and to their respective soteriological orientations: one toward transcendence and the other toward immanence.

With the notion that the third formless state is somehow "inadequate", the door is opened for a "fourth" state. While no such "fourth" is mentioned in the Chandoya Upanishad, at least explicitly, it does mention a "pure self" or Purusha that in some sense stands beyond the third state. Later commentators, including Shankara, will take this as referring to the "Fourth" state of the self, to Turiya. We can say then that the teaching of this "Purusha" is the teaching of Turiya in some sort of nascient form.

The "fourth", or Turiya, is explicitly mentioned for the first time in the Mandukya Upanishad, a very late Upanishad that is far removed from the Brhadaranyaka and Chandogya Up. Indeed, it is so late that some commentators do not even take it as an Upanishad. While the teaching of the Chandogya Up prefigures that of Mandukya Up, there is a historical and hermeneutic context that I believe must be supplied in order to understand what the Mandukya is actually saying with respect to this "fourth"; that context is the Buddhist tradition.

In a parallel prefiguring of the Mandukya Upanishad, the Potthapada Sutta --- an early Buddhist text --- refers to, and rejects, three "selves" that are related to three "states." The first self is "with form" (rupa) and "made up" of the four elements, and of "food." This latter formulation is similar to the formulation in the Taittiriya Upanishad, which refers to the first kosha or sheath of the self as being "made up of food". The Potthapada Sutra then describes a second self made up of form (rupa) but "consisting of mind," (mano-maya) which is precisely the same terminology used in the Taittiriya Upanishad. Last, it describes, and rejects, a third "formless" (arupa) self made up of consciousness (samjna). The grounds for the rejection of these three selves is the fact that they "come and go", that is, that each is transitory. Interestingly, transitoriness is also the reason given by the Advaitin Gaudapada, in his commentary upon the Mandukya, for the rejection of the first three selves as not "ultimate." In other words, the three states are not ultimate because they come and go (ie, we go into sleep then come out of it).

These features of commonality, I think, show two things: one, that the Buddhists were not only aware of the Upanishadic teachings, but that they were consciously providing a critique of them; and two, that the Advaitins were not only aware of the Buddhist critique, but that they were in turn consciously responding to it. That the author of the Mandukya Upanishad, and not just Gaudapada who is several centuries later, is himself aware of Buddhist doctrine is apparent in the fact, noted by Nakamura, that it refers to the "non-dual" fourth (Turiya) as "prapanca-upashama" (quieting of discursive proliferation) which is a technical term of Mahayana Buddhism used specifically by Nagarjuna.

There is a related idea that might be of interest to us here, and that is the Buddhist teaching of the "three realms." The three realms doctrine is a quasi-cosmographic idea of later Buddhism. It derives for the most part from the teachings concerning the jhanas and attainments (samapatti). Now, the earliest Buddhist works had only mentioned the four jhanas. But they had also mentioned other "attainments" (samapatti) that Gautama had procured during his life as an ascetic and yogin. What later Abhidharma works like the Visuddhimagga do is integrate these "attainments" with the jhanas that are referred to in the Pali canon. The usual resulting scheme depicts the four jhanas with the four "attainments" (samapatti) piled on top of them. This is a somewhat haphazard textual pastiche; however, it "works" as a kind of pseudo-phenomenology of states of consciousness. What it in fact shows is a process of textual redaction. This is to say, that the two were not at all related together until some commentator decided that they should be related, and in this way, two different teachings about "meditative states" were harmonized.

Now, when we take the four jhanas as one category, and add the four attainments as another, we get two "realms." Obviously these are not "ordinary" states of consciousness or realms of being. So, if we take the various "ordinary" states of being and add them to the other two, we happily get three "realms" or "worlds" (lokas)... and the Indian tradition rubs its hands together whenever it encounters the prospect of deriving a group of three. The "ordinary" states came to be associated with the six modes of being --- human, hell-being, hungry ghosts, animals, gods, ashuras --- and the other two "realms" came to be populated with their own respective "beings." In effect, a cosmography was created: kama-loka, rupa-loka (the realm of pure form), and arupa-loka (the formless realm). In time, the higher two realms were populated by all manner of gods and creatures drawn from the Buddhist creative imagination, with Maitreya, the coming Lord and Buddha to be, "living" in the highest realm, which by now has become a kind of jhana, the jhana of "neither perception nor nonperception."

To return to my initial line of thought, what I would like to suggest is that the Potthapada Sutta, with its description of the "three states," is the context for this later doctrine of the three lokas, as well as background against which was written the Mandukya Upanishad. The Mandukya, like later Buddhism and the Potthapada Sutta, also gives us a description of three "structures" of consciousness. It also describes a "fourth" that somehow transcends the three. As we noted, the textual precedent for the "fourth" can be found in the Chandogya Up, with its critique of the "third" state as somehow "inadequate." But it is the Buddhist tradition, I would suggest, that is the most important context for the composition of the Mandukya Up. The most important feature of this context, I think, is the idea that the three states are somehow "imperfect." And as we noted, the conception of this inadequacy, viz., the transitoriness of the three states, is taken over in toto by Gaudapada. 

What the Vedantins required was a "response" to the Buddhists. They needed an account that "transcended and included" that of the Buddhists. That account was provided by the Mandukya Upanishad. In effect, the Vedantins answered the critique of the Potthapada Sutta by saying the following: The three selves are indeed transitory. But beyond the three states is the true Self that never comes and goes, that is transcendent of the three and yet always already present as their "truth" or immanent basis. In other words they said, "Yes we have indeed described three states and three selves, but we also have a Fourth, which is beyond the three."

And in the process of giving this response they had started the ball rolling in the direction of a new game: the game called  "let's ratchet up the terminology," a game that will lead to the doctrine of Turiyatita, that which is "beyond the fourth," which is found in Kashmiri Shaivism and Tantricized Advaita, as well as to Da's collapsing of certain features of the fourth into the third, the "causal."