Bonus article · The Transition from Anatta to Emptiness
Two Modes, One Seeing
On dependent arising, spontaneous presence, and the recognition that frees the mind from its own fixing.
1. Two modes of mind
Cognition seems to work in two ways. One divides. It takes a whole, separates it into parts, names the parts, and asks which part carries the identity of the whole. A cup becomes a handle, a rim, and a body. Motion becomes a before and an after. The other mode does not divide. It registers what presents without the names, categories, and part-whole structure that thought adds.
Buddhist traditions give these modes different names. The first is conceptual construction (vikalpa, kalpanā). The second is non-conceptual awareness (avikalpa). Dharmakīrti defined direct perception as free of conceptual construction, registering the particular without the name that would make it a thing. Much of Buddhist practice rests on the claim that the first can be quieted and the second is what remains.
The question this essay asks is whether these two modes point to different results or to the same one.
2. Analysis breaks, and the breaking becomes the object
Nāgārjuna's Mūlamadhyamakakārikā (MMK) works largely by dissection. It takes motion, fire and fuel, agent and action, and the conditioned, divides each into parts, relations, and origins, and shows that none of the divisions yields what was claimed. The chariot is not in the wheels, the axle, or the pole. Motion is not in the mover or the moved.
Analysis thus has two faces. Conceptually, it separates an object into components and searches for the one that carries identity. Perceptually, the same division is already built into experience. We see a cup as parts because we have learned to. The MMK's claim is that this partitioning is not a neutral description of what is there. It is a mode of cognition that treats the parts as the real basis of appearance and the whole as an aggregate that needs explaining.
The regress makes the mode visible. Each part can be divided again, and the search for an indivisible basis either never ends or ends in a "part" that cannot be found when examined. Nāgārjuna does not conclude that conventional objects are unreal. He concludes that the analysis cannot find the intrinsic nature it presupposes. The parts, too, are designated.
This is why the MMK uses analysis to expose analysis. It does not abandon the dividing mind. It shows that the dividing mind is a formation and not a discovery.
3. The middle way is dependent designation
MMK 24.18 is the verse that names the middle way. In this site's translation it reads: dependent arising is what we call emptiness; that is a designation in dependence, and that itself is the middle way.
The verse holds three things together. Dependent arising is the fact that things come from causes and conditions. Emptiness is the lack of intrinsic nature that follows from that fact. Designation is the way such things are named and known, and the naming is itself dependent. Dependent designation is therefore not a compromise between existence and non-existence. It is what the middle path looks like when it is understood.
The verse blocks two errors at once. Eternalism treats things as having a fixed, self-standing nature that persists regardless of conditions. Nihilism concludes from the absence of intrinsic nature that nothing exists. Designation avoids both. Things are not nothing, because they function and can be named. They are not self-existent, because their designation depends on causes, parts, and the mind that names them.
4. Dependent arising and emptiness are one movement
The logic of MMK 24.18 runs in one direction. Whatever arises dependently arises from causes and conditions. Something with intrinsic nature would not need to depend on anything, because its nature would be self-sufficient. So whatever arises dependently lacks intrinsic nature. That lack is what "empty" means.
Emptiness is therefore not a separate reality behind phenomena. It is what dependent arising amounts to when its ontological status is examined. MMK 24.19 makes the point explicit: since there is no phenomenon that has not arisen in dependence, there is no phenomenon that is not empty.
The relation also runs the other way. If things had intrinsic nature, they could not change, be caused, or function. Emptiness is therefore not the opposite of dependent arising. Nor is it a precondition that makes dependent arising possible. It is what dependent arising amounts to. Seen this way, the two are not separate doctrines that happen to be related. They describe one appearing from two sides.
5. Openness as non-obstruction
If emptiness is the lack of intrinsic boundary, then openness can be understood precisely. Things obstruct each other only when each has a fixed shape that another runs into. A solid object resists because it is fixed in its own boundaries. Remove the intrinsic nature and the boundary has nothing to be fixed by, so appearances can meet without collision.
This is not blankness. Nothing is removed. Things appear fully, and their appearing does not block anything else. The Tibetan term thogs med (unobstructedness) and the Huayan image of mutually interpenetrating phenomena both express this. Unfixed, each thing is met as open, and that is why each can include the others.
The formulation needs care. "Openness is due to the lack of essence" risks making the absence of essence a ground that produces openness. The safer phrasing is that openness is what the absence of essence looks like when it is seen. It describes how appearances meet. It does not explain why they do.
6. Liberation is the ending of the act of fixing
Why does seeing this matter? Nāgārjuna's answer begins with the chain of suffering. Afflictions (kleśa: craving, aversion, and delusion) drive karma and the cycle of birth and death. Those afflictions arise from conceptual constructions, which arise from conceptual proliferation (prapañca). Proliferation ceases in emptiness (MMK 18.5).
Consider fear. It presupposes that the threat has its own nature and will harm me whatever I do with it. Seeing that the threat is a dependently arising process weakens its power, because it has no fixed nature to be threatening with. The same holds for the self. The ātman that the Buddha rejected is a fixed center that owns experiences and must be protected. Once the self is seen as designated in dependence on the aggregates, the mind that names them, and the conditions of the moment, the urgency of defending it loses its object.
Liberation is therefore not a change in what exists. It is a change in how what exists is held. MMK 25 makes the point that nirvāṇa is not another reality beyond saṃsāra. If the two were separable, liberation would be a place one travels to. Because they are not two, liberation is the cessation of proliferation within the same experience.
7. Two traditions, the same risk
The Dzogchen formulation of the nature of mind sharpens the question. Dzogchen speaks of rang rig, self-cognizing awareness, and lhun grub, spontaneous presence. Both point to a natural state that does not need to be made. The difficulty is in how this is heard. A nature that is not produced and stands as the ground of appearances would be the svabhāva that Madhyamaka refuses. Read with Mipham, Dzogchen does not mean that: "not made by causes and conditions" is Nāgārjuna's "unborn by nature", and spontaneous presence is dependent arising met without fixing.
Candrakīrti's sixth chapter of the Madhyamakāvatāra criticises the self-awareness (svasaṃvedana) of cognition, the Yogācāra view of self-knowing mind. The standard counterexample in these debates is the blade that cannot cut itself: a cognition cannot know itself without splitting into a knower and a known. The Buddhist logicians' reply uses the lamp. Dharmakīrti takes a lamp as the model of self-awareness, and Prajñākaragupta argues that a cognition arises with the essence of shining, so that the blade comparison does not apply to it. On this reading self-knowing is not a second act performed on the flow of awareness. It is that flow's luminosity, and like the flow it is empty, not an independent awareness. Even so, I do not posit self-knowing awareness, not even as convention. What the word points to, I keep only as a denial: knowing is not lit by another, and nothing lights it, not even itself. There is no second act and no knower behind; this clarity is as empty as what appears.
Mipham's Beacon of Certainty, in John Whitney Pettit's translation and study, defends Dzogchen in the terms of Madhyamaka logic. Gorampa (1429–1489) read the Madhyamaka critique of intrinsic nature more absolutely than Mipham (1846–1912) did. The two never debated: Mipham wrote four centuries later and drew on Gorampa's work. The contrast between their readings is useful because both agree on the danger of reification.
Dzogchen teachers' distinction between suppressing thoughts and their self-release (rang grol) makes the same point in practice. Suppression pushes thoughts down, which can create a stillness that becomes one more fixed state. Self-release lets thoughts dissolve when they are not followed or fixed.
8. Two paths in relation
Two paths to liberation are common in Buddhist practice. One is suppression of conceptualities (rnam rtog zhi ba), which aims to quiet thought so that perception is seen without being divided. Śamatha belongs to this family. Mahāmudrā's resting, Chan's no-thought and Dzogchen's non-meditation do not. They are not the stopping of thought but recognition: thought is seen as already empty, and it releases itself. They can be mistaken for suppression, but that is a misreading of them, not what they are. The other is analysis that exposes the conceptual mind's habit of fixing objects, which is the Madhyamaka route.
These two approaches are not equal. Suppression treats conceptuality as an enemy to be removed. Madhyamaka treats it as a formation to be seen through. Candrakīrti accepted śamatha as a conventional means that calms the mind, but he held that the decisive step is analysis, which shows that the parts searched for in the object cannot be found. Stillness forced by effort can become a new object of holding, and the practitioner ends up grasping the absence of grasping.
The Dzogchen path, read through Madhyamaka, points to something else. The natural state is not reached by stopping thought. Thought is recognized as already empty, already not fixed, and in that recognition it stops fixing. This is what it means for the path to be "not by effort but by recognition."
9. Recognition, not achievement
Recognition is the key term. "Spontaneous" (lhun grub) means already complete. It is not produced by effort, and so it is not a state one gets more of by trying harder. The natural state is recognized, not produced. Effortlessness is the same kind of thing. The moment one treats effortlessness as an accomplishment, one has made it another fixed object.
The same caution applies to negation. MMK 13.8 warns that those who take emptiness as a view are incurable. A way out that works by negating appearances until nothing is left is still a fixer, now working on a different object. Recognition differs in kind. It does not negate. It sees that appearances are already not fixed, so there is nothing further to do to them.
This corresponds to a distinction Madhyamaka draws between non-implicative negation (prasajya-pratiṣedha), which denies without installing a replacement, and implicative negation (paryudāsa), which denies and then installs a positive ground in its place. Recognition has to remain in the first category. If "the nature of each appearance" becomes a positive ground beneath appearances, the disease of reification has returned in a new form.
10. Spontaneous presence and dependent arising
The question then becomes how spontaneous presence relates to dependent arising. If both are understood rightly, they point to the same result.
Dependent arising (pratītya-samutpāda) is the analytic description. Things arise through conditions and so have no intrinsic nature. Spontaneous presence (lhun grub) is the experiential description of the same appearing. What presents is already complete and does not wait for construction to become what it is. Both deny production by an essence. Both leave appearance intact. Read this way, they are two descriptions of one event.
The apparent conflict is in the word "spontaneous." It can slide into causelessness, and an uncaused source is exactly the svabhāva Madhyamaka rejects. It would be a hidden ground: the demand for a producer, granted one final exemption. So "spontaneous" cannot mean "uncaused source."
The way to keep the two aligned is to separate two questions. The first asks how an appearance comes about. Every appearance, mental or material, has an answer: it arises from causes and conditions, and the MMK's analysis works at that level. The second asks about the nature of what appears. That nature is emptiness, and it is not confined to mind. There is not one phenomenon that is not empty (MMK 24.19). Emptiness is what dependent arising amounts to when it is examined, so it is the nature of mind and phenomena alike. It is not a further product of conditions, and so the question "why?" does not apply to it. "No why or because" does not contradict dependent arising. It marks the point where the question of cause stops applying, because emptiness is not something produced.
Two safeguards keep this from turning into a new ground. Emptiness is not a separate reality behind phenomena. And it must not become a hidden substrate of emptiness either, because emptiness is itself empty.
11. The why and the no-why
Here the two truths enter. MMK 24.8 through 24.10 distinguish conventional truth (saṃvṛti-satya) from ultimate truth (paramārtha-satya). Causal explanation is conventional. Conditions and effects are real for the purposes of functioning, and speaking of "why" and "because" is how we navigate and act. Candrakīrti is clear that these are valid at the conventional level.
But the ultimate analysis does not find an intrinsic causal link to be the ground of anything. So "no why and because" is the ultimate condition, expressed only conventionally to facilitate functioning. MMK 1.1 shows how this works. The chapter examines production from self, other, both, and no cause, and rejects each. The no-why conclusion is reached by working through the why, not by bypassing it.
MMK 24.10 is the verse that holds the two together. Without relying on conventional truth, the ultimate cannot be taught. Without reaching the ultimate, nirvāṇa is not attained. Neither truth is a lower stage to pass through. The ultimate is not a hidden level behind the conventional.
This is why the two truths cannot be understood apart. Taken separately, each becomes a fixed object. The conventional becomes solid things that really cause one another. The ultimate becomes an empty ground standing behind them. Each reification keeps the other in place. Joined, they dissolve each other's solidity. The conventional is seen to lack intrinsic cause, and the ultimate is seen not as an extra reality but as the lack of fixity in what appears.
12. Seeing vividness and emptiness together
What is it like to see this? Perhaps like this: presentations form one after another, seamlessly, and each is seen as vivid and as empty at once. Seeing its vividness and its emptiness is not seeing two features of one thing. Neither explains the other: the vividness is not produced by the emptiness, and the emptiness is not a property behind the vividness. Nothing is added to the appearance to make it empty.
The seamlessness matters. When presentations form without a gap, there is no moment at which a fixed object is produced and then labeled. The arising, the knowing, and the release happen as one event. The designation is not added after the appearance has formed. The designation is the forming.
The word "effortless" needs watching, as I've said. It is recognized, and the moment one tries to keep it, one has left it. The witnessing is likewise best understood as the flow's own clarity: nothing lights it up, not another and not itself, and no separate knower stands behind the stream. If the witness is a separate subject, it becomes another svabhāva. If it is the vivid, unobstructed quality of the flow, nobody is performing the tasting and seeing.
13. Conclusion
The two modes of cognition, dividing and non-dividing, can seem to pull toward different results: one toward analysis, the other toward stillness. Understood rightly, they point to the same result. Analysis shows that the parts searched for in the object are not there, and that the dividing mind is a formation. Non-conceptual seeing shows the same appearing without the division. Neither needs to be forced, and neither is the final truth of what is seen.
Dependent arising and spontaneous presence are two descriptions of one event. Dependent arising describes how appearances come without intrinsic nature. Spontaneous presence describes how the same appearances are already complete and need no fixing. Neither is established by emptiness, because emptiness is not a further thing that could establish anything. It is what dependent arising amounts to when it is seen.
The way out, then, is not by effort. It is by recognizing the nature of each appearance as already open, already unobstructed, already not fixed. The reification that causes suffering is the act of fixing, and when the fixing stops, nothing needs to be produced to take its place. That is the liberation the MMK points to, and that is the openness Dzogchen names.