Where does the going happen?

What you will work out

• Reconstruct the argument of chapter 2 that going cannot be located on the path already traversed, the path not yet traversed or a path being traversed.
• Explain the doubling argument against “the goer goes” and apply it to a new sentence.
• Compare the reading of chapter 2 through momentary existence with Westerhoff’s reading of it as a pattern about individuals and properties, and say what each needs.
• Explain why “ultimately, there is no going” does not say that nothing moves, and apply the same pattern to change.

Where is the walking?

Imagine someone walking up a hill path, halfway to the top. She is plainly walking. But where, exactly, is her walking taking place?

Not on the stretch behind her: she has already walked that, and whatever walking happened there is over. Not on the stretch ahead: she has not reached it yet. So it must be happening on the stretch she is walking now. But how long is that stretch? Any stretch we pick has a part already behind her and a part still ahead.

The second chapter of the Mūlamadhyamakakārikā opens with this puzzle.[1]

1. Just as the path traversed is not being traversed, neither is the path not yet traversed being traversed. The path presently being traversed that is distinct from the portions of path traversed and not yet traversed is not being traversed.

If motion is real in the strong sense the opponent means, it should be possible to say where it happens. Nāgārjuna argues that it cannot be found in any of the three places.

The opponent’s answer, and a second going

The opponent has an obvious reply.

2. [The opponent:] Where there is movement, there is the act of going. And since movement occurs in the path presently being traversed, not in the traversed nor the not yet traversed, the act of going occurs in the path presently being traversed.

Nāgārjuna answers by asking what makes a stretch of path “the path being traversed”. It is being traversed only because someone is going along it. So the going is already built into the description of the place. If we then say that the going happens in that place, we have counted the going twice.

5. If the act of going is in the path presently being traversed, then two acts of going would follow: that by which the path presently being traversed [is said to be such], and moreover that which supposedly exists in the act of going.

6. If two acts of going are supplied, then it would follow that there are two goers, for it does not hold that there is an act of going without a goer.

Two goings would need two goers, which is absurd. The translators add that nothing stops the regress at two.[1]

The same doubling appears when we turn from the place to the one who moves. “The goer goes” sounds harmless, even necessary. But someone counts as a goer only by going. If we then say that this goer goes, we have one going that makes her a goer and another that she performs (2.11). Nāgārjuna presses the point further. Going cannot begin in any of the three stretches, and it cannot stop in any of them (2.12–17). The goer and the going can be neither the same thing nor two different things (2.18–21). The chapter ends:

25. […] Thus there is no going, no goer, and no destination.

The translators read that conclusion as a claim about ultimate reality: “there can ultimately be no such thing as going, a goer, and a destination.”[1] That word matters, and we will return to it. First, why should the argument work at all?

Two readings of the argument

Interpreters disagree sharply about what chapter 2 is about.[2] The disagreement changes what the argument needs, so it is worth following both sides.

Reading one: a series of moments. Siderits and Katsura read the chapter against the Abhidharma theory that nothing lasts longer than an instant. Their commentary uses an image from the Abhidharma literature: a moving lamp is really a series of flames, each lasting a single moment and each in a single place. None of them moves; the appearance of one moving light comes from running together past, present and future flames. Since only the momentary flames are real, “strictly speaking there is no motion”: ultimately nothing moves, though we still speak, usefully, of carrying a lamp. “It is important to keep this example in mind throughout the rest of the chapter,” the translators write. “Many of the arguments depend on the assumption that nothing lasts longer than an instant.”[1]

Their commentary does not describe the assumptions about time in only one way, though. Of the opening verse (2.1) they say that it makes the same argument as Zeno’s paradox of the arrow and, like that paradox, “relies on the assumption that space and time are both infinitely divisible.” Westerhoff reports a related “spatio-temporal” reading, developed by Mark Siderits and J. Dervin O’Brien in 1976 and by others since, on which the chapter refutes a theory of divisible space and indivisible atoms of time by showing that motion would be impossible under it.[2] The versions differ in detail, but they share an approach: chapter 2 is an argument about how space and time are divided.

Reading two: a pattern about things and their features. Westerhoff thinks the chapter is not mainly about motion at all.[2] Consider “the apple is red”. The apple and the redness are separable: there are green apples and red things that are not apples. It makes sense to think of an individual, the apple, that has a property, redness. But “the goer goes” is not like that. There is no goer without going, and no going without a goer. When we call something “the goer”, we have turned one of its features into a name for it, and then we predicate that same feature of it again.

Westerhoff’s own illustration is thunder. We say “the thunder roars”, but there was never a silent thunder waiting to begin roaring; the roar is what the thunder is. A singer, by contrast, exists before the singing begins. The trouble comes only when we treat every case like the singer: as an independent thing with an independent property, which meet in a fact. For cases like the thunder, that analysis splits one feature into two and yields exactly the doubling Nāgārjuna points out.

Westerhoff finds support for this reading elsewhere in the work. In the chapter on fire and fuel, Nāgārjuna says the analysis of burning follows the pattern already given for “the presently being traversed, the traversed, and the not yet traversed” (10.13), and Candrakīrti, as Westerhoff reports him, substitutes burning for going throughout chapter 2. An argument that transfers so readily to burning, Westerhoff suggests, does not seem to depend on a theory of how space and time are divided. The evidence is less one-sided than it first looks, though. The translators read Candrakīrti’s substituted verse in temporal terms too: there is “no third time between past and future” in which the burning can take place.[1]

What each reading needs. The first reading needs a particular theory of space and time, and its versions disagree about which: that nothing real lasts longer than an instant, that space and time can be divided without end or that space can be divided but time cannot. Whichever theory a version needs, an opponent who rejects it can reject the argument. The second reading needs no particular theory of space and time. It needs only the claim that dividing a situation into an individual and its property is our way of describing it, not a division already present in the world. That is also a substantial claim, and an opponent can dispute it. Westerhoff is candid about the limits of his own approach. On one part of the chapter, the argument about where going begins, he sets out three reconstructions and concludes that the verses do not decide between them: “We cannot tell what the argumentum ipsissimum of Nāgārjuna might have been”, that is, which argument was his own.[2]

A youth does not age

The same pattern applies to change of any kind. In chapter 13, an opponent argues that change shows that ordinary things lack intrinsic nature, but that the ultimate elements must have it. Nāgārjuna replies that if anything had an intrinsic nature, there could be no change at all:[1]

5. It is not correct to say that alteration pertains to the thing itself that is said to alter or to what is distinct. For a youth does not age, nor does the aged one age.

The translators explain. If a youth were ultimately real, youthfulness would be his intrinsic nature, and aging would destroy it; so the youth, as such, cannot be what ages. An old person already lacks youthfulness, so cannot be what ages either. If we say instead that the person ages, first young and then old, we have made the person something with at least two natures, and therefore something composite and not ultimately real. Milk and curds work the same way (13.6): if milk itself became curds, the solidity of curds would already be in the milk; if something other than milk became curds, we could make curds from water. The translators note that this is the dilemma of chapter 1 applied again.

An opponent can try to save intrinsic nature by keeping it away from change. Westerhoff describes two such attempts.[2] On one, the elements never change: they simply cease, and new ones arise in their place. On the other, they never even cease: change is only their rearrangement, as when fire-elements move out of cooling water. He argues that each has a cost. The first must say what makes elements with fixed natures arise and cease; if causes do, he argues, the elements depend on something after all. The second pushes the truly real further and further from anything we experience.

What “no going” does not mean

Does this chapter say that nobody ever walks anywhere? The text does not say that, and neither do its interpreters.

The conclusion of 2.25 concerns ultimate reality. It denies that going, goer and destination are independently real items that meet at a place and a time. It does not deny that the walker reached the top of the hill. Conventionally, people walk, rivers flow and milk becomes curds, and we can describe all of it truly and usefully.

Westerhoff draws the contrast sharply. Mover and motion are, in a sense, illusory, because they do not exist the way they appear to, as independent items. But their lack of reality “is quite different from that of other non-existent objects, such as hares’ horns and present kings of France, which do not exist even at the level of conventional reality.”[2] A hare’s horn is nothing at all. A walk is a real walk, conventionally, even if it is not ultimately real.

He also offers a positive account of the beginning of motion. Since no point on the path qualifies as the beginning of a motion in and of itself, the beginning is “a boundary drawn by the mind in accordance with one’s particular interests and needs.” We cannot draw it just anywhere: the beginning must come before the present stretch, not after it. Within such limits, though, where it falls is our choice, made so that one stretch of moving counts as a single walk. That is Westerhoff’s reconstruction; the verses themselves do not spell it out.

Try the doubling

Imagine Birgit paddling a kayak down a river. Where the river narrows, the current picks up, and someone on the bank calls out, “The current is flowing fast!”

Apply the doubling argument to the claim that the current flows. Then explain which of the two readings the argument needs in order to work, and why its conclusion, “ultimately, there is no flowing”, does not say that Birgit’s kayak failed to move downstream. Writing an answer is optional.

A full explanatory answer

“The current flows” doubles the flowing in just the way “the goer goes” doubles the going. What makes some water a current, rather than a still pool, is its flowing. So when we say that the current flows, one flowing picks out the current and a second is said to occur in it. If current and flowing were two independently real items, the current would have to be something that could exist without flowing, a current at rest, and that is exactly what it cannot be. Compare “the river flows”: a frozen river is still a river, so there the doubling is much less clear. The argument bites where the thing is picked out by the very feature we then say it has.

The argument works most naturally on Westerhoff’s reading. It needs only the claim that dividing this situation into an individual, the current, and a property, flowing, is our way of describing it, and that the division does not track two separate things in the world. It needs no theory about whether time is made of moments. The first reading could also be applied, but it would add something: that the water consists of momentary elements, none of which flows. That extra premise is exactly what an opponent could reject.

Neither reading says that the kayak stayed where it was. The conclusion is about ultimate reality, in the opponent’s sense: there is no flowing that exists independently of the water and of how we describe it. Conventionally, the current flows and carries Birgit to the next bend, and anyone who denied it would be wrong. The distinction between conventional and ultimate truth, introduced in the first lesson, is what keeps the argument from becoming absurd.

The answer has a limit. Showing that one sentence has this doubling structure does not show that every description of motion does, and it does not settle whether the individual–property division is only a way of describing.

End of the first stage

This stage began with a question: could anything be what it is entirely on its own? The Abhidharma thinkers had a serious answer. Composite things like chariots and persons borrow their natures from their parts, but analysis ends in simple elements that have natures of their own. Nāgārjuna accepted their account of what such an element would be and denied that there are any.

Three arguments have supported that denial. Chapter 15 argued that a nature produced by causes cannot be intrinsic, and that giving up intrinsic nature is a middle path, not a slide into nothingness. Chapter 1 argued that nothing with an intrinsic nature could be produced, whether from itself or from something wholly other. Chapter 2, with chapter 13, argued that going and change cannot be located in things with intrinsic natures. Each argument had premises an opponent could question: a caused property might still be a thing’s own; the causal argument relies on assumptions about time; and the argument about motion relies either on a theory of how space and time are divided or on the claim that the individual–property division is our construction.

At every step the conclusions have had two levels. Ultimately, there is no intrinsic nature, no producing relation between independent things and no independently real going. Conventionally, fires still warm, seeds still sprout and walkers still reach the top. One question remains open. Could an opponent keep intrinsic nature by rejecting momentariness and holding that real things endure? Nāgārjuna’s arguments, as we have read them, press hardest on a world of moments, and it is not settled how they fare against a world of lasting things.

Remember, too, what kind of evidence we have been reading: short verses meant to be explained, reached here through a translation guided by classical commentaries and through one modern philosopher’s reconstruction. Other readings exist, and some have been noted along the way.

You can stop here. When you return, the next lesson, “Is the self its parts, or their owner?”, turns to the things that move and change most intimately: ourselves.

Sources and limits

[1] Nāgārjuna, Nāgārjuna’s Middle Way: Mūlamadhyamakakārikā, translated with commentary by Mark Siderits and Shōryū Katsura (Wisdom Publications, 2013), for chapters 2 and 13 and the translators’ commentary.

• Chapter 2: all twenty-five verses with commentary, pp. 31–42. The verses quoted are 2.1 (p. 32), 2.2 (p. 33), 2.5–6 (pp. 34–35) and 2.25 (p. 41); the moving lamp, the remark that many arguments assume nothing lasts longer than an instant and the comparison with Zeno, with its remark about infinite divisibility, are on pp. 32–33, the regress remark on p. 35 and the translators’ reading of 2.25 on p. 42.
• Chapter 13: verses 1–7 with commentary, pp. 137–144, with 13.5 on p. 143 and 13.6 on pp. 143–144.
• Chapter 10: verse 13 with commentary, pp. 116–118, applying chapter 2’s pattern to fire and fuel, with Candrakīrti’s substituted verse and the translators’ remark about the third time on pp. 117–118.

Words in square brackets are the translators’, supplied from the commentaries; the ellipsis in 2.25 marks the omission of its first half (25ab). The moving-lamp image comes from the Abhidharmakośabhāṣya as the translators cite it; that work was not read.

[2] Jan Westerhoff, Nāgārjuna’s Madhyamaka: A Philosophical Introduction (Oxford University Press, 2009), for the interpretation of chapter 2 and the argument from change. Chapter 6 and its endnotes were read in full, together with §2.1.2 and its notes.

• The two readings of chapter 2, thin individuals and the thunder example: §6.1, pp. 130–141. His note 13 (p. 133) names Siderits and O’Brien (1976) and others as proponents of the “spatio-temporal” reading, which he describes on pp. 133–134 (infinitely divisible space, time made of indivisible atoms); those works were not read.
• Three reconstructions of the argument about the beginning of motion, and the beginning as a boundary we draw: §6.2, pp. 142–147, with the quotation on p. 144 and “a boundary drawn by the mind” on p. 147.
• Mover and motion contrasted with hares’ horns: §6.3, p. 150.
• The argument from change and two replies to it: §2.1.2, pp. 38–40.

Candrakīrti’s substitution of burning for going is reported as Westerhoff reports it (p. 134). The lesson follows Westerhoff’s reading as one interpretation and presents the translators’ reading beside it; it takes no position on which Nāgārjuna intended, and it reports the different descriptions of the assumptions about time without reconciling them.

Birgit’s kayak and the walker on the hill are original teaching examples.