What makes a condition a cause?

What you will work out

Set out the four ways something might be thought to arise and explain why Nāgārjuna treats them as the only possibilities.
Reconstruct the argument that nothing can ultimately be a condition, and name the assumptions about time and simplicity it relies on.
Explain the dilemma between an effect already present in its causes and an effect wholly other than them.
Compare the reading on which causal connection is imputed on observed regularity with Westerhoff’s stronger reading, on which a group of causes is assembled by an explaining mind.

A seed in a drawer

A seed can lie in a drawer for years and nothing happens. Put it in warm, damp soil and a shoot appears. We say the seed, together with the soil, moisture and warmth, caused the sprout. Classical Indian philosophers used exactly this example, and, as Siderits and Katsura point out, they usually treated causation as a relation between things, the seed and the sprout, rather than between events.[1]

The last lesson ended with a tension in the Abhidharma view. Its ultimately real elements, the dharmas, were supposed to have natures of their own, and yet most of them were supposed to arise from causes and conditions. Chapter 15 argued that a nature produced by causes could not be intrinsic. The first chapter of the Mūlamadhyamakakārikā approaches the same problem from the other side. It examines the relation of producing itself, and asks whether anything with an intrinsic nature could ever stand in it.

Four ways a thing might arise

The chapter opens with its conclusion.[1]

Not from itself, not from another, not from both, nor without cause: Never in any way is there any existing thing that has arisen.

The four options map the ways an effect might be related to what produces it. It might already exist in its cause and simply come out. It might be something distinct from its cause. It might be both. Or it might arise from nothing at all. Nāgārjuna rejects all four as accounts of how anything ultimately real could arise.

Why think the list is complete? Westerhoff’s reconstruction treats the options as covering every possible relation between two things: the same, different, both or neither.[2] If every option fails, the fault may lie in an assumption the options share. He compares a question about the shape of the color red: circle, triangle and every other shape fail, not because the right shape is unusual, but because a color has no shape to find. The shared assumption here, on his reading, is that cause and effect are independent things, each with its own intrinsic nature.

Two of the options can be set aside quickly, and the opponent in the text agrees. Arising without any cause would mean, as the commentator Bhāviveka put it, that anything could come from anything at any time. The third option inherits the faults of the first two. The real work concerns the first and the second.

Why not from itself?

The first option says the effect was already there in its cause. Indian philosophers seriously defended a version of this view: the Sāṃkhya school held that an effect exists in its cause in an unmanifest form, waiting to appear. Verse 1.2 answers it.

The intrinsic nature of existents does not exist in the conditions, etc. The intrinsic nature not occurring, neither is extrinsic nature found.

According to the Akutobhayā, a commentary whose author is unknown, the first line rejects this option: if the effect were already in its cause, the cause would have the effect’s intrinsic nature, and it plainly does not. The other commentators add that production would then be pointless, and they illustrate the point with fire. We want fire because we want heat. If the heat were already present in the fuel, “in order to feel heat we would only need to touch unignited fuel”, as the translators summarize them.[1] Chapter 20 adds a second point: if the sprout were already in the seed, soil, moisture and warmth, we should be able to find it there, and no amount of looking finds it (20.3).

The opponent’s theory of conditions

The opponent in chapter 1 is a fellow Buddhist who holds the second option. Effects arise from conditions that are distinct from them, and the Abhidharma had classified those conditions.

[The opponent:] There are four conditions: the primary cause, the objective support, the proximate condition, and of course the dominant condition; there is no fifth condition.

The primary cause is what the effect comes from, as the sprout comes from the seed. The other three name further factors: the object a perception is of, the moment immediately before the effect that gives way to it, and whatever the effect could not arise without. Nāgārjuna examines each in turn, but the core argument concerns what makes anything a condition at all.

When is a condition a condition?

They are said to be conditions when something arises dependent on them. When something has not arisen, why then are they not nonconditions?

Something cannot be called a condition whether the object [that is the supposed effect] is not yet existent or already existent. If nonexistent, what is it the condition of? And if existent, what is the point of the condition?

The seed counts as a condition only because a sprout arises from it. Before there is a sprout, there is nothing for the seed to be a condition of. Once the sprout exists, there is nothing left for the seed to do. We might look for a third time, between the two, when the sprout is coming into being and the seed is producing it.

The translators explain why the opponent cannot use that third time. A chariot can be said to come gradually into existence while its parts are being assembled. But during that process the chariot both exists and does not exist, which is exactly why the Abhidharma counted it as a useful fiction and not as ultimately real. An ultimately real dharma was supposed to be simple, with a single nature. It either exists or it does not; it has no intermediate state of coming to be.[1]

This pattern, which the translators call the argument of the three times, recurs throughout the work. Applied to causes, it leads to a conclusion they state this way: the relation of production “is something that we impute upon observing a regular succession of events, but it is not to be found in reality.”

Why not from another?

The second option says the effect is wholly distinct from its conditions. Verses 11 and 12 press it.

That product does not exist in the conditions whether they are taken separately or together. What does not exist in the conditions, how can that come from the conditions?

If that which does not exist [in them] is produced from those conditions, how is it that the product does not also come forth from nonconditions?

The commentators use cloth and threads. The cloth is not in any single thread, and it is not in the threads taken together either, so on this view the cloth is something wholly other than the threads. But then threads are no closer to cloth than they are to curd. Bhāviveka, as the translators report him, draws the conclusion that “we should expect to be able to get curd from threads.”[1] Chapter 20 makes the same point with clay: firing clay yields a pot and never curds, and a theory of wholly distinct effects cannot say why (20.4). Its verdict is blunt: if cause and effect are separate, “there would be equivalence of cause and noncause” (20.20).

Westerhoff explains what being wholly other must mean for this argument to bite.[2] Nobody denies that a spark differs from the fire it starts. The claim under attack is that cause and effect are distinct in the strong sense of each existing independently, with its own intrinsic nature. A world of things like that would sit still. Each would be what it is regardless of the others, and nothing about them would make one the cause and another the effect rather than the reverse, or pick out one pair rather than another.

What the argument does not deny

It would be easy to read chapter 1 as saying that seeds do not grow and that planting is a waste of time. The translators explain why that would be a mistake. A verse in chapter 7 says, “Whatever exists in dependence, that is free of intrinsic nature” (7.16). Commenting on it, they explain that for Nāgārjuna’s school, dependent arising, the Buddha’s teaching that things arise in dependence on conditions (the translators call it “dependent origination”), can be understood in two ways:[1]

Understood as a conventional truth, it applies to such things as the pot and the cloth, which arise in dependence on causes and conditions. Understood as an ultimate truth, however, it is the teaching that no ultimately real things ever arise.

So the argument works at the ultimate level. Ultimately, nothing with an intrinsic nature produces anything. Conventionally, seeds planted in warm soil still sprout, clay still becomes pots, and knowing which conditions bring about which results is still worth having. What the argument denies is that production is a relation holding between independently real things, waiting to be discovered apart from anyone who notices a pattern.

How strong is the conclusion?

Interpreters disagree about how far this goes. On the reading the translators give from the commentaries, causal connection is imputed when we observe a regular succession. The regularity is real enough to guide action; the producing relation, as something over and above it, is our addition.

Westerhoff defends a stronger reading.[2] A cause never works alone: a seed needs soil, water and warmth. The translators call the whole group the “assemblage” of cause and conditions; Westerhoff calls it the “causal field”. Nāgārjuna asks how there can be an assemblage without an effect (20.24). Westerhoff takes this seriously. On his reading, a causal field is not a collection found ready-made in the world. It is “a cognitive artifact”, gathered by a mind that wants to explain why a particular effect came about, and it exists as a causal field only because of that effect and that explanation. Cause and effect then depend on each other and on us.

Westerhoff himself notes a gap in this line of argument. Showing that something is produced by causes shows that it depends on its causes. It does not yet show that it depends on concepts or minds. He answers with a principle he takes from Mark Siderits: if a thing essentially involves a property that is conceptually constructed, the thing is conceptually constructed too. You need not accept that principle, or Westerhoff’s reading of the causal field, to follow the rest of the study; the commentators’ more modest reading is enough for the argument of chapter 1.

Where the opponent can push back

The argument is not free of assumptions, and the opponent has places to resist.

The first concerns time. In The Golden Age of Indian Buddhist Philosophy, Westerhoff argues that the causal argument relies on two assumptions: that only the present exists, and that everything lasts only a moment.[3] Together they guarantee that cause and effect never exist at once, so the relation between them always lacks one of its terms. Not every Abhidharma school accepted both. The Sarvāstivāda school held that past and future things exist too, and in his earlier book Westerhoff concedes that the missing-term problem “does not arise” for them in that form, though he argues that they face a related difficulty.[2] He concludes that the two sides do not mean different things by intrinsic nature: “They do, however, have different concepts of causation.”[3]

The second concerns scope. A verse in chapter 24 concludes that, since there is no dharma that does not arise dependently, there is no dharma that is non-empty (24.19). The translators observe that the verse gives no argument for its premise: it “seems to rely on our having already accepted the premise that everything ultimately real is dependently originated.”[1] An opponent who held that some ultimately real thing is uncaused, as some Buddhists held of space, would not be moved by chapter 1 alone.

Try the dilemma

Imagine Hamid, a baker, explaining his morning to an apprentice. At first he says the loaf was already there in the dough, and the oven only brought it out. When the apprentice asks why he bothers to light the oven, Hamid changes his account: the loaf is something wholly independent of the dough and the heat, which merely come before it.

Reconstruct the dilemma Nāgārjuna would press on Hamid’s two accounts. Then name one premise Hamid could reject to escape it, and say what the exercise shows about the argument. Writing an answer is optional.

A full explanatory answer

Hamid’s two accounts are the first and second of Nāgārjuna’s four options, and each fails for the reason chapter 1 gives.

If the loaf is already in the dough, baking is pointless. Nothing new needs to be produced, and Hamid could skip the oven, just as, on the commentators’ example, one could warm one’s hands on unlit fuel. We should also be able to find the loaf in the dough, and we cannot.

If the loaf is wholly independent of the dough and the heat, then nothing about the dough makes it the source of bread rather than of anything else. Dough would be no more a condition of bread than threads are of curd. The dough and the oven’s heat would merely come first, with nothing to explain why what follows is bread.

The dilemma bites only because Hamid assumes that loaf and dough must be either one and the same thing or two things each existing on its own. That is the assumption that each has an intrinsic nature. A baker who says, in the ordinary way, that the loaf comes from the dough through the oven’s heat has made no claim the dilemma touches. He is speaking conventionally, and Nāgārjuna has no quarrel with that.

Hamid has another way out. He could reject the assumptions about time on which the three-times argument relies. If past and future things also exist, or if a process can have genuine duration, the cause need not have vanished before the effect appears. Westerhoff notes that some Abhidharma thinkers held views of this kind.

The exercise shows where the argument is vulnerable. It does not settle whether it succeeds. That depends on whether the opponent can make one of those escape routes work without giving up intrinsic nature altogether.

Carry this forward

Causes were supposed to bring about change. If production is not a relation between independent things, what becomes of change itself, and of motion, the most visible change of all? The next lesson, “Where does the going happen?”, follows Nāgārjuna’s argument that motion cannot be located anywhere.

Sources and limits

[1] Nāgārjuna, Nāgārjuna’s Middle Way: Mūlamadhyamakakārikā, translated with commentary by Mark Siderits and Shōryū Katsura (Wisdom Publications, 2013), for the verses of chapters 1 and 20 and the translators’ commentary.

Causation as a relation between things; chapter 1’s outline: pp. 17–18.
Chapter 1: all fourteen verses with commentary, pp. 17–29. The verses quoted are 1.1 (p. 18), 1.2 (p. 19), 1.3 (p. 20), 1.5–6 (pp. 22–23) and 1.11–12 (pp. 26–27). Bhāviveka’s remark on uncaused arising is on pp. 18–19, the Akutobhayā’s reading of 1.2 and the unignited fuel on p. 19, the translators’ statement about imputed causation on p. 23 and Bhāviveka’s curd example on p. 27.
Chapter 7: verses 1–34 with commentary, pp. 71–88, were read; the lesson quotes 7.16 and the translators’ distinction between conventional and ultimate dependent arising, pp. 81–82.
Chapter 20: all twenty-four verses with commentary, pp. 212–226, including 20.3–4 (pp. 216–217), 20.20 (p. 224) and 20.24 (p. 226).
24.19, the translators’ observation about its premise and their remark on space: pp. 278–279.

The translators follow Ye Shaoyong’s 2011 edition and reverse the order of verses 1.2 and 1.3 relative to translations based on the La Vallée Poussin edition, so the four conditions appear here at 1.3. Words in square brackets are the translators’, supplied from the commentaries. Bhāviveka, Buddhapālita and Candrakīrti are reported only as the translators report them. The Sāṃkhya theory is described only in the terms the translators and Westerhoff use; no Sāṃkhya text was read.

[2] Jan Westerhoff, Nāgārjuna’s Madhyamaka: A Philosophical Introduction (Oxford University Press, 2009), chapter 5, for his reconstruction of the arguments about causation. Chapter 5 and its endnotes were read in full.

The four options as exhaustive; the color red: §5.1, pp. 92–94.
The causal field and its dependence on the effect and on us: §5.2, pp. 94–99, with “a cognitive artifact” on p. 97.
Causation by another; why independent objects cannot be causally related: §5.3.2, pp. 104–109.
The Sarvāstivāda view of time: §5.4.1, pp. 114–118, with “does not arise” on p. 115.
Summary, the gap between causal and conceptual dependence, and Siderits’s principle: §5.4.5, pp. 122–124.

The lesson treats Westerhoff’s reading as one interpretation among others. Siderits’s work, cited by Westerhoff, was not read.

[3] Jan Westerhoff, The Golden Age of Indian Buddhist Philosophy (Oxford University Press, 2018), chapter 2, section 4a, “Nāgārjuna and the criticism of the Abhidharma project”, for the causal argument’s reliance on presentism and momentariness and the conclusion that the two sides differ about causation. The copy consulted is the publisher’s corrected proof, whose page numbers have not been checked against the published edition, so the section title is given instead of pages.

Hamid’s bakery and the seed in a drawer are original teaching examples; the seed and sprout, fire and fuel, threads and cloth, and clay and pot are the commentators’ examples as the translators report them.