This essay examines the Madhyamaka (dbu ma) tradition as it developed from Nāgārjuna’s second-century root text, the Mūlamadhyamakakārikā, through its systematic elaboration in Tibet. Particular attention is given to the doctrine of emptiness (stong pa nyid), the two truths (bden pa gnyis), and the doxographical dispute between Prāsaṅgika (thal ‘gyur pa) and Svātantrika (rang rgyud pa) readings of Madhyamaka logic. I argue that Tsongkhapa’s (tsong kha pa) synthesis, which insists that emptiness and dependent origination (rten ‘brel) are not merely compatible but identical in meaning, represents the most philosophically rigorous attempt in the Tibetan canon to close the gap between ontology and soteriology. A final section renders these ideas in plain language for the non-specialist reader.
Keywords: Madhyamaka, śūnyatā, dependent origination, two truths, Nāgārjuna, Tsongkhapa, Prāsaṅgika, Tibetan Buddhist philosophy
Few doctrines in the history of philosophy have been as persistently misread as emptiness. To the untrained ear, śūnyatā sounds like nihilism dressed in saffron robes. It is not. Madhyamaka, the “Middle Way” school founded by Nāgārjuna (klu sgrub), does not claim that nothing exists. It claims something more precise, and honestly more difficult: that nothing exists the way we instinctively think it does.
This essay is not a survey. Surveys already exist, and good ones. Instead, I want to walk through the internal logic of the tradition the way a Tibetan monastic debate courtyard would — slowly, with an eye for where an opponent’s position collapses under its own weight, and always circling back to the question of why any of this matters for a being trying to end suffering.
Nāgārjuna wrote at a moment when Abhidharma scholasticism had, in a sense, become a victim of its own precision. Early Buddhist analysis had broken the person down into aggregates (phung po), the aggregates into momentary events (skad cig ma), hoping that enough decomposition would finally locate something irreducible — a dharma with “own-nature” (svabhāva, rang bzhin) standing at the bottom of the reduction.
Nāgārjuna’s move in the Mūlamadhyamakakārikā was to turn this reductive method against itself. If causation is real, he asks, does an effect arise from itself, from another, from both, or from neither? Each option, examined closely, fails. Things do not produce themselves — that would be redundant. They cannot arise from something wholly other, because “other” only makes sense relative to what is being produced. Not both, for the same reasons doubled. Not neither, because then anything could arise from anything, and nothing would explain anything. This fourfold negation (mu bzhi, the tetralemma) is not a rhetorical trick. It is a method for showing that when we look for the essence that supposedly makes causation “work,” it is not there, in any of the four logically exhaustive places it could be.
The conclusion Nāgārjuna draws is startling only until you sit with it: things function perfectly well without possessing any fixed essence. In fact, he argues, it is precisely because things lack essence that they can function at all. A permanently fixed, self-contained entity could never change, interact, or produce effects. Only what is empty of inherent existence can participate in the flux of causes and conditions.
When Madhyamaka crossed the Himalayas, it did not arrive as a finished product. Tibetan scholastics — working across Sakya, Kagyu, Nyingma, and later Gelug monastic colleges — spent centuries arguing over how literally to take Nāgārjuna’s negations, and over which Indian commentator (Buddhapālita, Bhāviveka, Candrakīrti) had understood him correctly.
The decisive fault line concerns method. Bhāviveka (legs ldan ‘byed) held that Mādhyamikas should establish their positions using independent syllogistic reasoning (rang rgyud, svatantra) accepted by both parties in a debate. Candrakīrti (zla ba grags pa) rejected this, arguing that a Mādhyamika has no thesis of their own to prove through independent inference, because any such inference would smuggle back in the very essentialism under critique. Instead, the Prāsaṅgika (thal ‘gyur pa) method works by consequence — by drawing out the absurd implications (thal ‘gyur) already hidden in the opponent’s own assumptions, without asserting a competing positive claim.
Tsongkhapa (1357–1419), founder of the Gelug tradition, elevated Candrakīrti’s Prāsaṅgika to the summit of Buddhist philosophical systems in Tibet. His innovation, however, was not merely siding with one Indian authority over another. It was insisting — against readers who took emptiness to mean a kind of blank negation, an absence sitting behind appearances — that emptiness is dependent origination, seen from a different angle. Nothing is added by the analysis. Nothing is subtracted from the conventional world. What changes is only the mistaken belief that things possess a findable core independent of causes, parts, and conceptual labeling.
This is worth pausing on, because it is where Tsongkhapa’s reading is most often misunderstood by later interpreters, including some within his own tradition. He is not saying conventional reality is an illusion to be seen through and discarded. He is saying conventional reality — cause, effect, persons, suffering, liberation — is only possible because things are empty of inherent existence. Emptiness is not the enemy of appearance. It is appearance’s precondition.
A separate current, associated with certain Jonang and some Kagyu and Nyingma masters, developed what is called the zhentong (gzhan stong, “other-emptiness”) position, distinguishing it from rangtong (rang stong, “self-emptiness”). Zhentong proponents held that while conventional, conditioned phenomena are indeed empty of themselves, the ultimate nature of mind — buddha-nature (bde gshegs snying po) — is empty only of what is extraneous to it, not of its own intrinsically luminous qualities. This debate, still alive in comparative Tibetan philosophy today, turns on how to reconcile Madhyamaka’s universal negation with tathāgatagarbha literature’s more affirmative language about buddha-nature.
Everything that exists, exists in dependence on something else — on causes and conditions, on its own parts, or on the conceptual mind that designates it as “this” rather than “that.” A chariot is not identical to its wheels, nor separate from them, nor does it exist independently of the concept “chariot” superimposed on an assembly of parts. Pull on any thread of this dependency and you find, not a hidden essence, but more dependency.
Madhyamaka does not deny the everyday world. It distinguishes conventional truth (kun rdzob bden pa) — the level at which tables, persons, and causal relationships function reliably — from ultimate truth (don dam bden pa), the fact that none of these things withstand analysis for inherent, independent existence. Crucially, the two truths are not two separate realities layered on top of each other. They are two ways of describing a single reality: how things appear to conventional, unanalyzed cognition, and what is found (or rather, not found) when that same reality is subjected to ultimate analysis.
Madhyamaka reasoning systematically refuses four positions regarding any phenomenon: that it exists, that it does not exist, that it both exists and does not exist, and that it neither exists nor does not exist. This is not evasive fence-sitting. It is a precise rejection of the assumption, buried in the question itself, that “existence” must mean “existing by way of its own independent essence.” Once that assumption is removed, phenomena can be said to exist conventionally without falling into any of the four extremes, because the extremes were never really about existence in the ordinary sense — they were about essence.
The Prāsaṅgika method’s refusal to assert an independent positive thesis is often mistaken for philosophical timidity. It is the opposite. It reflects the recognition that any positive metaphysical claim about how things ultimately exist would itself require the very inherent existence Madhyamaka is dismantling. The Prāsaṅgika debater wins not by building a fortress but by showing that the opponent’s fortress was built on nothing.
It would be a serious error to treat this as a dry logical exercise. In the Tibetan curriculum, the study of emptiness is inseparable from the cultivation of bodhicitta (byang chub kyi sems), the aspiration to liberate all beings from suffering. The connection is not sentimental; it is structural. Suffering, in the Buddhist analysis, is generated substantially by grasping — at a fixed self that must be protected, defended, and gratified. If that self cannot be found under analysis, the entire architecture of grasping loses its foundation, not through suppression, but through the collapse of its object.
Compassion, correspondingly, becomes more coherent rather than less once the rigid boundary between self and other is understood as conventionally useful but ultimately unfindable. Śāntideva’s argument in the Bodhicaryāvatāra, that the suffering of another is worth addressing on exactly the same grounds as one’s own suffering, depends on this loosening of the self’s supposed metaphysical priority. Emptiness, in other words, is not opposed to ethical life. It is what makes universal compassion logically defensible rather than merely aspirational.
Madhyamaka’s enduring difficulty is also its enduring achievement: it refuses to give the mind a resting place, whether in naive realism or in nihilistic despair. Tibetan scholasticism, at its best, preserved this discomfort rather than resolving it prematurely into a comfortable slogan. To understand emptiness correctly is not to acquire a new belief. It is to lose an old one — the belief that things must have essences to be real — and to discover that letting go of that belief is precisely what allows the conventional world, causation, ethics, and liberation itself, to make sense.
If you strip away the technical vocabulary, the core idea is something like this.
Imagine a wave in the ocean. It looks like a distinct thing — it has a shape, a beginning, an end, you can point at it and say “there’s a wave.” But if you go looking for the wave’s own separate substance, apart from the water, the wind, and the motion, you won’t find one. The wave isn’t fake. It’s just not a self-contained object; it’s an event made entirely of conditions coming together.
Buddhist philosophy says people, tables, nations, feelings — everything — work the same way. They are real in the sense that they function, cause effects, and can be talked about usefully. But they are “empty” in the sense that none of them have some solid, independent core sitting underneath, separate from their parts and causes and from our own minds labeling them.
This matters practically because a lot of human suffering comes from treating things — especially ourselves — as fixed and solid when they are not. We defend a rigid idea of “who I am” as if it were a fortress under siege. Emptiness, in this tradition, is less a scary void and more like good news: nothing is locked in place, which means change, growth, and healing are always structurally possible. And because the wall between “self” and “other” is looser than it feels, caring about other people’s wellbeing stops being an extra moral effort and starts looking like simple honesty about how things actually are.
Garfield, Jay L. The Fundamental Wisdom of the Middle Way: Nāgārjuna’s Mūlamadhyamakakārikā. Oxford University Press, 1995.
Dreyfus, Georges B. J., and Sara L. McClintock, eds. The Svātantrika-Prāsaṅgika Distinction: What Difference Does a Difference Make? Wisdom Publications, 2003.
Hopkins, Jeffrey. Meditation on Emptiness. Wisdom Publications, 1983.
Ruegg, David Seyfort. The Literature of the Madhyamaka School of Philosophy in India. Otto Harrassowitz, 1981.
Thakchoe, Sonam. The Two Truths Debate: Tsongkhapa and Gorampa on the Middle Way. Wisdom Publications, 2007.
GIAO LONG MONASTERY
GIAO LONG MONASTERY
GIAO LONG MONASTERY
GIAO LONG MONASTERY
