The Logical Impossibility of Intrinsic Nature

The Substantialist's Last Defence

When confronted with the principle that all phenomena depend on conditions, the substantialist often retreats to a sophisticated counterargument. They may concede that dependencies come first in the order of practical explanation, but insist that a phenomenon still possesses an intrinsic nature—specifically, the underlying fact that these particular dependencies obtain rather than others. In this view, intrinsic nature acts as an independent ground that determines the specificity, stability, and character of causal patterns.

However, this defence contains a fatal logical contradiction. For an intrinsic nature to serve as an independent ground, it must be identifiable independently of the dependencies it supposedly grounds. If it can only be identified by pointing back to them, it is not independent; it is merely a redundant re-description of those dependencies.

The Exhaustive Analysis: No Unconditioned Remainder

Take the phone in your hand and look for its intrinsic nature—what it is in itself, apart from everything it depends on. Begin with its structure. It is a phone only while its parts hold together in a certain way: glass, battery, circuit board, case. Take them apart and lay them on the table, and the phone is not among them, nor is it their heap. Its structure is how the parts cohere, and the coherence is not one more part.

Look to its origin. It came from a factory and from designers, from mined lithium and refined silicon, from a long history of telephones before it. None of that is in the phone, yet all of it is what the phone is identified through. Look to its function. It is a phone because it can call, message and connect—and it can do so only where there is a network, a charger, a hand, and someone at the other end. In a world without networks it would be a smooth slab of glass.

Look to its qualities: the cool weight, the bright screen, the black of the case. Each is felt or seen in contrast—cool against a warm palm, bright against a dim room, black against the white table—and by someone who perceives it. Look to its continuity. It is 'the same phone' as yesterday only within a framework of before and after, one that counts the new scratch on its corner as a change to it rather than the arrival of a different thing. And look, finally, to its boundary. Is the case part of the phone? The charging cable? The software, the account, the photographs stored on a server across the world? Where the phone ends is decided by designation, for a purpose.

Each place the analysis looks yields something vivid and something dependent. The same six questions—structure, origin, function, qualities, continuity, boundary—can be put to a cup, a tree, a thought or a person, with the same result.

When all these dependencies are systematically examined, nothing unconditioned remains. The substantialist is left with an inescapable dilemma: either admit that nature is identified through dependencies (so that 'intrinsic nature' adds nothing to the dependencies), or insist that nature is eternally inaccessible (performing zero explanatory work).

A Lived Test: When Does the Phone Stop?

Try this, in imagination or with an old device you no longer need. Remove the case: still a phone. Crack the screen: a phone with a cracked screen. Take out the battery: a phone that will not turn on—still a phone, we say, only a dead one. Remove the circuit board, then the speaker, then the camera. At some point people stop calling it a phone and start calling it parts. But at which point? There is no moment at which a phone-essence leaves, and no part whose removal takes the phone-ness with it.

Yet the designations shift reliably. Nobody tries to make a call on a pile of components, and a repair shop knows exactly what to charge at each stage. 'Phone', 'broken phone' and 'parts' are used correctly, and they track something—not an essence that was present and then departed, but what the assembled things can do and how they are met. This is the chariot of Chapter 3 in a modern form: sought among its parts it is not found, and as designated it rings.

Occam's Razor vs. Logical Impossibility

If an inaccessible intrinsic nature performs no explanatory work, Occam's razor dictates that we should reject it as metaphysical clutter. This does not make dependencies the new explanation of why things are as they are: the conditions we cite are themselves designated—valid within their framework and unfindable when sought—so they replace the essence as a description, not as a new ground.

However, the argument for emptiness goes beyond pragmatic simplicity. The analysis of dependence does not merely suggest that we should not posit an intrinsic nature; it demonstrates that intrinsic nature is logically impossible for dependent phenomena. Because everything that can be posited is identified only in dependence—as the next section argues—intrinsic nature would require a content identified independently of everything through which anything is identified, and no such content can be posited.

The difference is the difference between an unnecessary entity and an incoherent one. Occam's razor tells us not to posit a ghost in the attic when the creaking is explained by the wind, but it leaves open that there might be a ghost after all. Nobody needs a razor to dispose of a square circle. The phrase does not describe a shape we have failed to find; it fails to describe anything. Intrinsic nature, the argument claims, is of the second kind: not an extra posit for which evidence is lacking, but a requirement that nothing could meet.

Why Nothing Escapes Dependence

The argument so far shows that intrinsic nature is impossible for dependent phenomena. Its force depends on a further step: that no phenomenon of any other kind can be coherently posited. Nāgārjuna states it plainly—there is nothing that is not dependently arisen, and therefore nothing that is not empty (MMK 24.19). If this were an empirical generalization, it would be open to the obvious reply that we have not examined everything, and that 'unfindable' may simply mean 'not yet found'. But the step is not inductive.

This is not a definition of what phenomena are, still less a thesis about the nature of being. It is an observation about what anyone who proposes an exception must already be doing. To propose a phenomenon—to perceive it, conceive it, name it, or entertain it as a hypothesis—is to pick it out. Picking out requires contrast, for nothing can be singled out except against what it is not; and it requires a designation under which it is singled out. The proposal of an independent phenomenon is therefore made in dependence on the very conditions it claims to escape. This is the third and subtlest of Tsongkhapa's three senses of dependence, set out in Chapter 4: dependence on designation.

Consider the strongest candidates for an exception: a creator, an absolute, a pure awareness that underlies all experience, a final particle, or the unconditioned itself. Each is proposed by contrast with what it is said not to be. The unconditioned is defined against the conditioned; the absolute against the relative; the ground against what it grounds. As Nāgārjuna observes, where the conditioned is not established, the unconditioned cannot be established either (MMK 7.33).

The only remaining option is a 'phenomenon' that is identified in no way at all—related to nothing, contrasted with nothing, discerned by nothing. But to propose it is already to identify it, and so to make the proposal dependent; and if it is not identified, nothing has been proposed at all. The proposal of an independent phenomenon undermines itself either way. This is why the conclusion is a matter of logical necessity rather than induction, and why it is a negation rather than a thesis—a non-implicative negation (prasajya-pratiṣedha), not an implicative one (paryudāsa) that would leave something in place of what it denies. It does not affirm that everything possesses the property of dependence; it shows only that no independent phenomenon can be coherently posited. Nāgārjuna says the same of his own method: if he had a thesis, the fault would be his; having none, he is without fault (Vigrahavyāvartanī 29).

Three misreadings should be set aside. First, the argument does not make dependence a universal law of being, a new ground that everything obeys; dependence, too, is identified only in contrast and designation, and is empty (Chapter 4). Second, dependence on designation is not production by the mind. Conventionally, a table is not made by thinking of it; and the cognition that designates is itself designated, as empty as what it discerns. Third, the argument makes no claim about things beyond all identification—neither that there are such things nor that there are not, for either would be a thesis about what lies beyond reach. It observes only that talk of them has no content: nothing has been picked out for it to be about.

Does Dependence in Being Known Mean Dependence in Being?

A careful realist can grant everything argued so far and remain unmoved. Of course, she says, picking something out requires contrast and a designation; reference is dependent. But the thing referred to need not be. The mountain was there before anyone named it, and would be there had no one ever done so. That an object can be known only through concepts tells us about knowing, not about the object. Dependence in being known does not entail dependence in being. This is the strongest form of the substantialist's case, and the previous section does not answer it merely by repeating that every proposal is made in dependence. The realist accepts that, and denies that it matters.

Madhyamaka does not reject the ordinary claim the realist begins from. Conventionally, the mountain was there before it was named; it does not wink out when no one looks; nobody's thinking of it makes it. Tsongkhapa insists on exactly this: what valid conventional cognition establishes is not overturned by analysis, and to say that the mountain depends on designation is not to say that a mind manufactures it (the second misreading above). If the realist meant only this, there would be no quarrel.

But mind-independence in this ordinary sense is not what intrinsic nature claims. Svabhāva claims more: that there is a referent determinate from its own side—with its own boundary, its own identity and its own unity—which designation merely tracks, as a label tracks a jar it did not make. The disagreement is not over whether the mountain is there when no one looks. It is over whether the mountain that is there comes with its edges and its oneness already drawn.

Here the argument from identification needs the search, and here the search does its real work. Put to the mountain-in-itself the questions put to the phone. Where does it end: at the tree line, the foothills, the river? Is it the rock, the soil, the snow, their sum, their shape? Is it the same mountain after an earthquake, or after a million years of erosion? Every one of these questions has a usable answer, and every answer is supplied with a purpose and a scale of description (Chapter 13); none is found in the mountain apart from that. The determinacy the realist needs—the very feature that would make the mountain a self-standing 'it' for words to latch onto—turns out, at each point examined, to be designated together with the mountain's identification. What the realist calls the object independent of our knowing is, feature by feature, the object as identified.

The realist may retreat one step further: grant that the boundaries and the identity are ours, but insist that something is still there beneath them, which our designations carve. Locke reached the same place with substance and could describe it only as 'a supposition of he knows not what support' (Essay II.xxiii.2). This is the 'something' of the third misreading above. It has no boundary, no identity and no properties, for each of those has been handed back to designation; it is therefore not a candidate for intrinsic nature at all, only the empty place where one was expected. To affirm it says nothing, and to deny it would be a thesis about what lies beyond identification, which this book does not advance. Either way, nothing has been found that has its nature from its own side.

This is also what the necessity in this book's title amounts to. It is not a claim to have proved, from premises any metaphysician must accept, that intrinsic nature fails in every possible world. It is the necessity of a consequence (prasaṅga): whoever proposes a thing with its nature from its own side must say what that nature is; to say it is to identify it, and every feature offered, when examined, is found to be designated together with its identification; and if nothing is said, nothing has been proposed. The argument needs no premises of its own about being; it works with the opponent's. And its conclusion remains a negation. It does not tell the realist what the mountain is made of instead. It shows that the mountain does not need to be made of anything in order to be there, and to be climbed.

Objections: The Regress and the Search for a Ground

Two objections press hardest on the argument of this chapter. The first is the regress objection. If everything depends on something else, and that on something further, the chain of dependence never ends. Many philosophers regard such a regress as vicious. In recent metaphysics, grounding theorists such as Jonathan Schaffer have argued that chains of dependence must terminate in something fundamental, or nothing would ever be real at all (see Schaffer, 'On What Grounds What', 2009, and 'Monism: The Priority of the Whole', 2010).

The regress is vicious only on a particular assumption: that dependence transmits reality from a ground, so that each item borrows its existence from what lies beneath it. On that assumption an endless chain never pays out, and the objection succeeds. But the argument of this book makes no use of that assumption. Dependence is not offered as a mechanism that supplies existence; it is used to show that no item, when sought, is found to have existence of its own. The regress therefore has nothing to defer. It is the grounding picture, not the Madhyamaka analysis, that needs a first link—and the demand for one is precisely the demand for an intrinsic nature that the preceding sections have shown cannot be met. Jan Westerhoff examines this non-foundational structure in Nāgārjuna's Madhyamaka: A Philosophical Introduction (2009).

The second objection comes from the philosophy of physics: ontic structural realism, which denies that objects have intrinsic natures but makes relational structure what fundamentally exists (James Ladyman and Don Ross, Every Thing Must Go, 2007). Chapter 15 takes it up in detail. In short, it keeps a ground—structure itself—and so repeats the move Chapter 4 declines, putting relation in the place the substance vacated.

A third objection, that 'unfindable' may simply mean 'not yet found', was answered earlier in the chapter, and a fourth, that dependence in being known need not mean dependence in being, in the section just before this one. Taken together, the replies show the shape of the argument. It does not compete with these views for the title of the true foundation. It shows that the search for one cannot succeed.

Dependent Arising and Emptiness: One Appearance, Two Directions

Emptiness is not a separate metaphysical doctrine appended to dependent arising, and dependence is not a ground from which emptiness unfolds. They are one appearance seen from two directions: dependent designation is how phenomena appear and function conventionally, while emptiness—of the dependence as much as of what depends—describes the total unfindability of any independent essence within those same appearances.