Some thoughts on the dichotomy of self-empty (rang stong) and other-empty (gzhan stong):
The basis of the dichotomy comes about from considering the role and scope of negation in the apprehension of emptiness.
The dialectic of emptiness that runs through the various tenet systems (from Vaibhashika to Sautantrika to Cittamatra to Madhyamaka) progressively refines the negandum or the object of negation from the coarsest level (of substantial existence) to the subtlest level (of inherent existence). Consequently there is the progressive expansion of the domain or set of all unreal existents and a correlative diminution of the set of all real existents. We go from denying the reality of substantially existent selves and affirming the reality of irreducible atomic phenomenal particulars or “dharmas” on the coarsest level at the base of the dialectic, to denying the reality of all phenomena and affirming the reality of no phenomena on the subtlest level at the culmination of the dialectic.
We can see that the movement operates by a process in which each subsequent moment subsumes and preserves every preceding moment while simultaneously negating the ground upon which those moments stood. This trajectory follows reasonably from Sakyamuni Buddha’s doctrine of anattā or no-self, so his followers continue to excavate various degrees of self-grasping with the overall interests of extricating themselves from those errors. But we can also see how in its movement away from the extreme pole of absolutism, the tradition inevitably moves towards the extreme pole of annihilationism.
It is over concern for this risk that some Tibetan Madhyamikas posited another level that goes even beyond Prasangika-Madhyamaka—the hitherto highest level. The idea is that while Prasangikas basically get it right when they include all conventional phenomena in the set of all unreal existents, this leaves untouched a remainder that cannot be negated by the Prasangika because it is not a phenomenon: it is the ultimate original ground, and consummate climax, of all phenomena. This ground is also empty, but it is empty in a different way from the way phenomena are empty: it is empty of anything other than itself but not empty of itself (other-empty), whereas all phenomena are empty of themselves (self-empty). Without affirmation of the transcendental ground that is the immanent essence of all sentient beings, it is said that the Prasangika’s deconstructive logic will become an obstruction to the fulfillment of its own aim of Buddhahood by getting fixated with self-emptiness. The other-empty or Zhentong/Shentong view, sometimes called the “Great Madhyamaka,” is not supposed to contradict the Prasangika per se, nor all that which precedes it, but aims to subsume and preserves their insights while also going beyond them. It should also be stated that the particular form of Prasangika under critique here is what the Shentongs call Rangtong.
While this move makes sense on a certain intuitive level, there are a number of issues that immediately come up.
Firstly, the introduction of the Shentong view as a corrective to the annihilationist risk of the Rangtong-Prasangika reads a bit ad hoc. It breaks with the continuity of the transition between each moment of the dialectic as a progressive development of the pursuit of negating the negandum. It merely reminds us of the positive state of a permanent, changeless, uncompounded essential core that is intrinsically replete with the fully endowed qualities of Buddhahood from the very beginning.
Secondly, I am confused as to whether or not the Shentong view includes the notion of an intrinsically existing Buddha-nature as a (or even the sole) member of the set of all real existents, or if it introduces an entirely new category marked by its own domain that is distinct from the set of all real and all unreal existents. Whichever one is the case, however, this would be another indication of a massive rupture in the otherwise continuous dialectic. Because in the typical movement, there is a gradual transfer of all the members of the set of reals into the set of unreals, no additional categories other than those,* and no regression where the set of reals grows relative to a previous moment in the movement.
*You may be inclined to ask, what about “both real and unreal” and “neither real nor unreal?” I think whatever fits into “both”, can be members of both the real and unreal sets, and since nothing is “neither” real nor unreal, that category is functionally useless here. The tetralemma is basically reducible to a fundamental binary opposition; the third and fourth limbs are just complex combinations of those basic elements.
If we were to model the Rangtong view in set theoretical terms, a conventional existent can be defined as a set with members that does not include (hence is empty of) itself. This accords with the basic Buddhist premise that a designated object (such as a self) is not identical to, nor different from, nor dwelling amongst the factors of its basis of designation (such as the aggregates). Ultimate reality here can be defined as that universal set of all sets which are empty of themselves.
Now the question is: is the universal set of all sets which are empty of themselves, empty of itself? If it is empty of itself, then by its own definition it should necessarily include itself. But then immediately this would mean that it cannot include itself, otherwise, it would not be empty of itself! So perhaps, it is not empty of itself? That would mean that it includes itself, which contradicts its own definition as including nothing but sets which are empty of themselves. Through this model we can see how the Rangtong runs up against a paradox, risking either the faults of inconsistency, self-contradiction, or both.
If we were to model the Shentong view in set theoretical terms, we would retain the same definition of conventional existents as the Rangtong, as sets which do not include themselves. Note, that just because there is a set that is empty of itself does not mean that it counts as a conventional existent. But, all conventional existents are sets which are empty of themselves since part of the definition of a conventional existent is to be without self.
Now with the Shentong view, ultimate reality can be defined as that particular set which is empty of everything that is not itself. It would basically seem to amount to an empty set: a set with no members. However, another important part of the Shentong view is that while ultimate reality is empty of everything other than itself, it is not empty of itself, meaning, it includes itself as a member. Now we run into a number of issues.
Can this set truly count as empty if it includes itself? It is tautological to just state that the empty set which includes itself as a member is, ipso facto, also empty. To actually make a significant point worth our attention however, Shentong would have to go beyond simple tautologies. Indeed it does, when it claims that the sphere or set of ultimate reality which is empty of everything other than itself is replete with all of a Buddha’s fully formed qualities from the beginning. So it would seem that in fact, the set of ultimate reality in the Shentong view is not truly empty in any meaningful sense of the term, since it includes a myriad qualities. Since it is a non-empty set which includes itself, it also counts as a kind of self. Since ultimate reality here has the character of selfhood, it counts as an ultimate self, paramātmān.
So we seem to be left with two choices: Either we follow Rangtong-Prasangika, stick to the negative dialectic, and deal with paradoxes at the apex of our development, or we follow Shentong and introduce a positive rupture expressing the self-revelation of a truly existing absolute. Either we risk undermining any basis for further practice by annihilating everything we come into contact with, or we surrender to the very thing we wanted to avoid from the beginning in our admission of a mystically intuited absolute. We seem to find ourselves, ironically, with the most sophisticated articulations of the two extremes found within Madhyamaka, the path beyond extremes, itself!
Can we find a middle way beyond these two extremes of the middle way?
A few short, sketchy remarks on this:
When we identify two extremes the first thing we should be doing as Madhyamikas is to find what they hold in common—the unity of these opposites. Once we uncover this unity, we would have found the source of the issue, and then can properly diagnose it in order to prescribe remedies to absolve ourselves of this illness.
One important thing that the two extremes of Gelug “Rangtong” Prasangika and Jonang Shentong seem to share in common is a certain kind of commitment to an ontological project. It is a kind of investment, whether explicitly acknowledged or not, in the endeavor to construct a totalizing conception of reality on the basis of received scripture and sophisticated reasoning. An important character of ontological projects are that they are almost invariably closed and (at least intend to be) self-sufficient systems, requiring no communicative exchange or dialogue with the outside in order to function. So although they seem to be going in opposing directions, these two views are united in their both being committed to ontological projects.
If we diagnose a commitment to ontological projects as the source of the dilemma, what would be our prescription? If the issue is too much ontology, then in the interest of maintaining homeostasis we can suggest more epistemology. A significant reevaluation of the very limit and function of thought may be in order. Perhaps when we say that things are empty, this is not true in any ontological sense in which there is a relationship of objective correspondence between the idea of an emptiness and the reality of this emptiness. If statements on, or cognitions of, emptiness are to be veridical without being ontological, then truth itself needs to be reevaluated and understood differently.
If statements on or cognitions of emptiness are delimited to the sphere of analysis, where analysis itself is understood in terms of being a particular mode of practice, then emptiness has no ontological basis but a practical basis. Emptinesses would be grounded on the basis of, form constituent features of, and become realized upon, the deployment of a practical program deriving value from the concrete effects that result from abiding in emptiness. Considered independently of a practical, soteriological program for collective well-being and personal self-realization, emptiness cannot be said to meaningfully exist in any way. So emptiness is true, not because it truly corresponds to an ultimate state of affairs but because it delivers what it promises when put to use, which is the transition from mere sentience to perfect awakening.
From this perspective, we can see how both the Rangtong and the Shentong views can be useful, but only once their respective views are conditioned by the axiom that all things derive their origin and meaning from a particular mode of creative praxis, and that considered independently of this praxis they can mean nothing that is thought of them and can deliver nothing which is sought from them. Instead of being opposing extremes vying to occupy the same position on a hierarchical scheme of tenets, we can freely make use of both of them in the appropriate contexts as they are needed. When they function as counter-weights of each other, they help us find our own balance as we walk the middle path to complete knowledge and wellness.