Is emptiness empty too?

What you will work out

• Explain the opponent’s argument that emptiness needs a bearer that is not empty, and Nāgārjuna’s reply in 13.7.
• Explain why emptiness is compared to a medicine that must leave the body with the illness, and why treating emptiness as a real thing is called a mistake.
• Use Candrakīrti’s pot that is “free of hairs” to explain how emptiness can be ascribed to things without being a property they have on their own.
• Explain how “neither empty nor non-empty” can be read without contradiction, and name the rival reading that accepts a contradiction.

Soap that stays in the shirt

Think of washing a stained shirt. The soap lifts the stain, but if the soap is never rinsed out, the shirt is not clean: it is merely dirty in a new way. Soap does its work only if it leaves along with what it removes.

The study has now reached a point where the same question arises about emptiness itself. Lesson 6 ended with a temptation. If every analysis finds that things lack intrinsic nature, then perhaps emptiness is what they really are underneath: the one thing that survives every argument, the true nature of reality at last. This lesson follows Nāgārjuna’s reasons for refusing that step, and asks how his refusal can make sense.

Must emptiness have a bearer?

The objection is old, and it is a good one. In chapter 13 the opponent grants that composite things such as persons are empty: they change, so they must have parts, and whatever has parts lacks a nature of its own. But the opponent insists that not everything can be empty. Candrakīrti, as the translators report him, puts the argument this way: emptiness is supposed to be a feature of things, and a feature needs something to bear it. “Indeed the son of a barren woman being nonexistent, black color cannot be attributed to him.”[1] What does not exist cannot have a color. In the same way, if anything is empty, there must be ultimately real things for emptiness to belong to, and those, the opponent argues, cannot themselves be empty. So at least the dharmas must have natures of their own.

Nāgārjuna agrees with part of this, and uses it:

7. If something that is non-empty existed, then something that is empty might also exist. Nothing whatsoever exists that is non-empty; then how will the empty come to exist?

The translators explain what he is conceding. He agrees that emptiness could not ultimately occur without ultimately real things for it to characterize. But he does not withdraw his claim that nothing has intrinsic nature. Instead he denies that emptiness is ultimately real either. In their words: “To say of things that they are empty is just to say that they are not ultimately real, and their not being ultimately real is not itself ultimately real.” The opponent’s argument shows that emptiness, taken as an ultimately real feature, would need an ultimately real bearer. Nāgārjuna’s answer is that emptiness is not an ultimately real feature.

A medicine that must leave the body

The next verse gives an image of what emptiness is meant to do:

8. Emptiness is taught by the conquerors as the expedient to get rid of all [metaphysical] views. But those for whom emptiness is a [metaphysical] view have been called incurable.

The “views” in question are, in the translators’ gloss, theories about the ultimate nature of reality. The word they translate as “expedient” literally means something that expels or purges. Candrakīrti illustrates the verse with a passage from a sūtra in which the Buddha speaks with Kāśyapa:

“It is as if, Kāśyapa, there were a sick person, and a doctor were to give that person a physic, and that physic having gone to the gut, having eliminated all the person’s bad humors, was not itself expelled. What do you think, Kāśyapa, would that person then be free of disease?”

Kāśyapa answers that the illness would be more intense than before. Emptiness is the medicine. It clears away theories that credit things with intrinsic nature, and it cures only if it does not then settle in as the final theory itself. Another classical commentator, Buddhapālita, as the translators report him, says of those who understand this: “They see that emptiness is also empty.” (The translators note in passing that the ancient Greek skeptics used the same image of a purge that purges itself.)

Two ways to grasp a snake

Chapter 24 names the danger in another image:

11. Emptiness misunderstood destroys the slow-witted, like a serpent wrongly held or a spell wrongly executed.

A snake is safe to handle only if held the right way. Candrakīrti, as the translators report him, describes two wrong grips. One takes emptiness to mean that no conditioned thing exists at all. The other takes emptiness to be “a really existing thing with a real locus”, something real that sits somewhere. The first is the slide into nothingness that chapter 15 refused. The second is the temptation this lesson began with. Both, the translators add, come from failing to understand the two truths.

A few verses later comes the verse the translators call the most celebrated in the work:

18. Dependent origination we declare to be emptiness. It [emptiness] is a dependent concept; just that is the middle path.

“Dependent origination” is the translators’ term for dependent arising, the Buddha’s teaching that things arise in dependence on conditions. The translators explain the verse through Candrakīrti. Whatever arises dependently arises without intrinsic nature, and so is empty. On this reading, emptiness is not simply the same thing as dependent arising; it follows from it. And to call emptiness “a dependent concept” is to say that it is like the chariot: a useful concept, not the name of something with a nature of its own. “That is,” the translators conclude, “emptiness is itself empty.”

A pot free of hairs

How can something be true of everything and yet not be a real property of anything? Westerhoff finds the clearest answer in an example he draws from Candrakīrti.[2]

Some people suffer from an eye condition that makes them see hairs floating everywhere. Nobody with ordinary vision would describe an empty pot as “free of hairs”. It is empty of countless things, and that is not one worth mentioning. But to someone who sees hairs in every pot, you might say exactly that, in order to correct the mistake. “Free of hairs” names no feature the pot has in its own right. It is ascribed to correct a mistaken ascription, and it would have no point if no one made the mistake.

Emptiness, on Westerhoff’s reading, works the same way. People ordinarily take things to have natures of their own; calling things empty corrects that mistake. Emptiness is not a further property sitting in things, and not a hidden substance underneath them. A mind that did not credit things with intrinsic nature would have no need to call them empty. Westerhoff adds a related image, also Candrakīrti’s, from his commentary on 13.8. A shopkeeper tells a customer that he has nothing to sell, and the customer asks him to sell that nothing. Taking emptiness as a view is like taking “nothing” as the name of something for sale.

The same reading answers a worry that Westerhoff states sharply. Candrakīrti, as Westerhoff reports him, sometimes describes emptiness in words that sound like intrinsic nature: unchanging, uncreated, not dependent on anything else. Does emptiness then have exactly the marks of the thing it denies? Westerhoff’s answer draws on the distinction from lesson 1. Emptiness is a feature every object has and could not lose, as heat belongs to fire. It is an essential feature in that everyday sense, not something that exists independently of everything else. Being always true of things is not the same as being a thing in its own right.

Neither empty nor non-empty

A verse in chapter 22 goes further than any we have read:

11. “It is empty” is not to be said, nor “It is non-empty,” nor that it is both, nor that it is neither; [“empty”] is said only for the sake of instruction.

This is the tetralemma, the pattern of four alternatives: that something is so, that it is not, that it both is and is not, and that it neither is nor is not. Here Nāgārjuna rejects all four. It can look like a rejection of logic itself. How can a thing be neither empty nor non-empty, and not even neither?

Westerhoff argues that the pattern makes sense within ordinary logic once we notice that there are two kinds of negation.[2] Indian philosophers distinguished them as paryudāsa and prasajya negation, and Westerhoff interprets the difference as follows. A presupposition-preserving negation keeps the assumptions of the claim it denies: “this coat is not blue” suggests that the coat has some other color. A presupposition-cancelling negation rejects those assumptions. Asked whether next Tuesday is heavy or light, you should reject both answers. Your rejection does not mean that Tuesday is of middling weight; it means that the question assumes Tuesdays have weights, and they do not. The same holds if the questioner adds two more options, “both heavy and light” and “neither, but somewhere in between”: each still assumes that Tuesday has a weight. Rejecting all four rejects the assumption they share, not logic. Westerhoff adds a refinement. So that these rejections cannot cancel each other out into an assertion, he reads Nāgārjuna as declining to assert each alternative (“I do not assert that…”) rather than asserting its opposite.

What assumption do the four alternatives of 22.11 share? On Westerhoff’s reading of this verse, it is that there are objects with natures of their own, existing independently of us, which words simply pick out, so that “empty” either applies to them or does not. Asserted on that assumption, even “everything is empty” would be misunderstood; asserted without it, it is what Nāgārjuna wants to say. The translators put the verse’s last line in similar terms: saying that things are empty is not a statement of the ultimate truth about reality, “just a useful pedagogical device”. That is why emptiness is “said only for the sake of instruction”.

Not everyone reads the tetralemma this way. Westerhoff reports that some modern interpreters, notably the philosophers Graham Priest and Jay Garfield, have suggested that Nāgārjuna may have used a logic that tolerates contradictions, and that the emptiness of emptiness may be a genuine contradiction. The view that some contradictions are true is called dialetheism. Westerhoff judges that such readings develop Madhyamaka ideas further rather than reconstructing what Nāgārjuna had in mind, and he finds no clear evidence of a logic of this kind in Nāgārjuna or his Indian commentators.[3] Even so, he allows that some questions, the emptiness of emptiness among them, might be fruitfully read in a dialetheist way.[2] The study does not settle this dispute.

Giving up all views

The Mūlamadhyamakakārikā ends by saluting the Buddha, “who, based on compassion, taught the true Dharma for the abandonment of all views” (27.30). The translators read this as the model for the whole work. When the Buddha refused every answer to certain questions, he did so by rejecting the presupposition the answers shared, and the same strategy, they suggest, “may help us avoid turning emptiness into yet another metaphysical theory.”[1]

That strategy invites an obvious charge. If Nāgārjuna rejects every view about the ultimate nature of reality, isn’t that rejection itself a view about it? The commentator Bhāviveka, as the translators report him, answered with an analogy: it is no more inconsistent than saying “Quiet!” to stop the noise. The translators suggest two ways to take his reply. Perhaps some purely negative statements, which remove false descriptions, come closer to reality than positive ones. Or perhaps statements should be judged not by how well they describe the ultimate nature of reality, since there is none, but by how well they achieve what the speaker is trying to do. The next lesson follows Nāgārjuna’s own answer to the charge, in a different work.

Try the correction

Imagine Mateo, who has followed this study so far and draws a conclusion: “So emptiness is the real nature behind everything. It is the one thing that exists in its own right.”

Using 13.8 and the pot free of hairs, explain Mateo’s mistake. Then say what your answer does not yet settle. Writing an answer is optional.

A full explanatory answer

Mateo has turned the medicine into a new illness. His conclusion treats emptiness as the thing that really exists behind appearances, and something that exists in its own right is exactly a thing with intrinsic nature. The analysis was meant to show that there are no such things. If it ends by finding one, called emptiness, it has recreated what it set out to remove. That is the patient in the sūtra Candrakīrti quotes on 13.8, whose purge was never expelled, and the second wrong grip on the snake of 24.11.

The pot free of hairs shows the better reading. Emptiness is ascribed to things to correct the mistake of crediting them with natures of their own, as “free of hairs” is said of a pot only to someone who sees hairs in it. It names no further property that things possess on their own. And 24.18 calls emptiness itself “a dependent concept”, like the chariot: useful, and without a nature of its own.

Mateo is right about one thing. On Westerhoff’s reading, emptiness is true of everything and cannot be lost, as heat belongs to fire, so there is a sense in which it is the nature of all things. His error is the step from “always true of things” to “exists in its own right”, which confuses lesson 1’s two senses of intrinsic nature.

The answer stops short of one question. It shows that the claim “everything is empty” must apply to emptiness too. It does not yet show that a claim which is itself empty can refute anything. That is the objection of the next lesson.

Carry this forward

If the claim that everything is empty is itself empty, what gives it any force against someone who disagrees? The next lesson, “Does the argument undermine itself?”, follows the objection in its strongest form and Nāgārjuna’s own reply, and ends the study.

Sources and limits

[1] Nāgārjuna, Nāgārjuna’s Middle Way: Mūlamadhyamakakārikā, translated with commentary by Mark Siderits and Shōryū Katsura (Wisdom Publications, 2013), for verses of chapters 13, 22, 24 and 27 and the translators’ commentary.

• 13.3 and Candrakīrti’s statement of the opponent: p. 140.
• 13.7–8 with commentary: pp. 144–146, with the translators’ sentence on p. 145, the Kāśyapaparivarta passage (§65, as Candrakīrti quotes it) on p. 145, Buddhapālita on p. 146 and the footnote on the Greek skeptics on p. 146.
• 24.11 with commentary: pp. 274–275.
• 24.18 with commentary: pp. 277–278.
• 22.11 with commentary: pp. 247–249, with “just a useful pedagogical device” on p. 247 and Bhāviveka’s reply and its two interpretations on pp. 248–249.
• 27.29–30 with the translators’ closing remarks: pp. 334–335.

Words in square brackets are the translators’. Candrakīrti, Buddhapālita, Bhāviveka and the sūtra are reported only as the translators report them. The skeptics’ parallel is the translators’, in a footnote; no Greek source was read.

[2] Jan Westerhoff, Nāgārjuna’s Madhyamaka: A Philosophical Introduction (Oxford University Press, 2009), for his account of emptiness as a correction, the tetralemma and the two kinds of negation. These sections and their endnotes were read in full.

• The apparent contradiction in the emptiness of emptiness; the pot free of hairs; emptiness as an essential feature: §2.1.3, pp. 40–46, with the pot on p. 45.
• Two kinds of negation and the rejection of two alternatives: §§4.1–4.2, pp. 68–73.
• The rejection of four alternatives; 22.11 as a semantic point: §4.3, pp. 73–89, with the illocutionary reading on pp. 78–80 and 22.11 on pp. 84–85.
• Where a dialetheist reading might help: the end of chapter 4, p. 90, with §2.1.3’s note 101 (p. 42) on Priest’s view of the emptiness of emptiness.
• The customer who asks to buy nothing: §9.4, p. 197.

Westerhoff reports the pot from Candrakīrti’s Madhyamakāvatāra and the customer from Candrakīrti’s commentary on 13.8; neither work was read. His own translation of 22.11 ends “only said for the purpose of designation”; the lesson quotes Siderits and Katsura’s. The Tuesday and the coat are original examples; Westerhoff’s own examples are numbers and colors.

[3] Jan Westerhoff, The Golden Age of Indian Buddhist Philosophy (Oxford University Press, 2018), chapter 2, section 4c, “Contradictions and Nāgārjuna’s thought”, for the dialetheist reading and his assessment of it. The section and its footnotes were read in full; the copy is a corrected proof, cited by section title. The works of Priest and Garfield that Westerhoff cites were not read.

The shirt and soap and Mateo’s conclusion are original teaching examples.