03 May 2013

The Simile of the Raft

THERE are a small number of texts which are quoted again and again by Western Buddhists. Perhaps the most common is the so-called Kālāma Sutta and I have already spent several essays trying to demonstrate that it does not support the uses to which it is put (now combined into a booklet called Talking to the Kālāmas). Western Buddhists are simply mistaken about that text.

If the Kālāma Sutta is the most cited text then the Simile of the Raft from the Alagaddūpama Sutta (MN 22; M i.130) would be a good contender for second. This is the text that tells us that the Dharma is a raft to get us to the other side, where it must be abandoned. What follows is an extract from my translation and commentary on the Alagaddūpama Sutta which I hope to publish at some point.

The Simile of the Raft
(M i.134-5)

Bhikkhus I will teach you the simile of ‘the raft for the purpose of getting across’. Pay attention and listen to what I will say. 
“Yes, Bhante,” the bhikkhus replied. 
The Bhagavan said “Suppose a man is following a stretch of road, and he comes to a great flood. The near bank is dangerous and frightening, the far bank is safe and secure. There is no boat or bridge to cross the water. He thinks ‘what if I were to were to gather grass, wood, sticks and leaves and having woven them into a raft, I should swim, and safely cross to the other side?’ So he makes a raft and crosses the flood. Then once he has crossed over to the far bank he thinks: ‘this raft was very helpful to me in crossing the flood, what if I were to pick it up and carry it on my head or shoulders and go on my way?’”. 
“What do you think, bhikkhus, is this man acting sensibly if he takes the raft with him?” 
“No, Bhante.” 
“What would the sensible thing to do be? Here bhikkhus, he has crossed over to the far bank he thinks: ‘this raft was very helpful to me in crossing the flood, now let me haul it up to dry ground, or sink it in the water, and be on my way.’ That, bhikkhus, is the sensible way to act towards the raft. Just so, bhikkhus, I have taught the Dhamma as like a raft for ferrying, for getting across. 
Bhikkhus, through understanding the Dhamma in terms of this parable, you should renounce dhammas, and more-so non-dhammas.”

In this passage the Buddha certainly says that his dhamma is like a raft for crossing a river. And it is clear that having crossed a river, it is foolish to carry the raft along with you. However this is a simile, or really a parable, and the interpretation of what this parable means hinges on how we read the last sentence. The last sentence is the critical part of this passage, and it is also the most difficult to understand.

Now most people take this simile as saying that we don't need the Dhamma when we are enlightened, but this was not the Buddha's view as I will show below. The Buddha never abandoned the Dhamma as a refuge. So we can exclude this meaning. In order to understand the whole passage, to understand what the parable is pointing at with it's comparison of crossing a river we need to understand this last sentence.

dhamma and adhamma

The passage tells us that, having understood the Dhamma in terms of the parable of the raft then we ought to  renounce dhammā and more so adhammā (both in the plural): dhammāpi vo pahātabbā pageva adhammā. The words dhammā and adhammā have evoked a variety of renderings.

Buddhaghosa (MA ii.109) says that ‘dhammā’ here means calm and insight (samatha-vipassanā), specifically craving for calm and insight, but this does not make a great deal of sense, someone on the other shore has no craving to give up and one cannot abandon the raft before getting across. No modern exegetes seem to accept Buddhaghosa’s suggested interpretation. Horner interpreted the phrase as suggesting that we give up morality at the further shore (see Keown 1992: 93). Horner’s (1954) translation is “you should get rid even of (right) mental objects, all the more of wrong ones.” (p.173-4). Gethin (2008) interprets dhammā/adhammā as “good practices and bad practices” (p.161), which echoes Buddhaghosa but is less specific. However ‘practice’ is hardly a usual translation for dhamma (one might even say it is a mistranslation). Also there is plenty of evidence that the Buddha did not give up practice after his awakening.

Ñānamoli and Bodhi (2001) opt for the “teachings and things contrary to the teachings” which is at least a possible translation. I am doubtful about dhammā in the plural being interpreted in the sense of ‘teaching’ (I’ll return to this). Bodhi’s footnote (p. 1209, n.255) acknowledges the ambiguity and justifies their translation with a pious homily. Thanissaro (2010) does not translate the key terms: “you should let go even of Dhammas, to say nothing of non-Dhammas." The capitalisation implies that he understands ‘teachings’, as dhammā as ‘things’ is seldom capitalised and he therefore has the same problem as Ñānamoli and Bodhi. Piya (2003) also avoids committing himself: “you should abandon even the dharmas, how much more that which is not dharmas” [sic]. He refers to MA and Bodhi’s footnote for an explanation, thus seems to be accepting Ñānamoli and Bodhi's reading.

Richard Gombrich (1996) has weighed in with support for translating ‘teachings’ and ‘non-teachings’: “The Buddha concludes that his dhammā, his teachings are to be let go of, let alone adhammā. The occasion for this whole discourse is given by Ariṭṭha, who obstinately declared that he understood the Buddha’s teaching in a certain [wrong] sense.” (p.24). The argument that dhammā in the last sentence is not the dhamma referred to in the earlier parts of the passage Gombrich declares to be “sheer scholastic literalism” (p.24), but I have been unable to locate another passage in which the Buddha uses dhammā in the plural to describe his teaching. Gombrich comments on the irony of taking literally a text preaching against literalism (p.22), with the implication that Ariṭṭha--to whom he emphasises the sutta was directed--is guilty of literalism, or of clinging to the Dhamma. However Ariṭṭha was guilty of stubbornly refusing to relinquish a completely wrong interpretation. He is not a literalist, but simply has a wrong view. His problem is that he does not take the Buddha’s injunction literally enough! That the simile of grasping the snake at the wrong end, which immediately precedes the raft simile, applies to Ariṭṭha we cannot doubt. Ariṭṭha has misunderstood the teaching. The simile of the raft appears to be talking about something entirely different, and unrelated to Ariṭṭha. This is so striking when reading the text that I am inclined to agree with Keown who speculates that the sutta is a composite of originally separate sections (p.96).

Basing his discussion solely on Ñānamoli and Bodhi’s translation, Jonardon Ganeri has attempted to problematise the idea of abandoning the teachings. Firstly he says that if we take dhammā to mean teachings then the teachings only have instrumental value (p.132). Ironically this is not really a problem from a Buddhist point of view as we tend to see the teachings instrumentally (though there are Buddhist fundamentalists). His other argument, which relies on interpreting the Buddha’s word as ‘Truth’ is that for one on the other side “truth ceases altogether to be something of value” (p.132). Again this is not really an issue for Buddhism as truth as expressed in language is always provisional. The ‘Truth’ (if there is such a thing) is experiential, and on experiencing bodhi and vimutti one does not need provisional truth any more. Ganeri seems to misunderstand the pragmatic way Buddhism values truth – truth is whatever is helpful. This is epitomised in two now clichéd passages: in the Kesamutti Sutta (A i.188ff) where the Buddha tells the Kālāma people to trust their own experience in determining right and wrong conduct; and at Vin ii.10 where the Buddha tells his aunt Mahāpajāpatī that the Dhamma is whatever conducive to nibbāna.

If we accept Ñānamoli and Bodhi’s ‘teachings and things contrary to the teachings’ then we must state the standard caveat, which is that one only abandons the teachings after reaching the further shore. Too often this passage is used to attack doctrine being applied on this shore, or in the flood. There is no suggestion but that we absolutely need the raft until we are safely on the other side. 

Thus from various reputable scholars we get the full range of possibilities for translating dhammā: ‘teaching, morality, things, mental objects’.

This parable is also examined in depth by Keown (1992), where he points out that this is the only mention of abandoning the raft (p.95) and that in other texts “it is made perfectly clear that sīla along with samādhi and paññā are part of the further shore and are not left behind on the near side after enlightenment.” (p.95). As Keown points out, in some texts the further shore is morality (e.g. A v.232, and v.252f ). I would add that this idea that one abandons the Dhamma after enlightenment is flatly contradicted in the Gārava Sutta:
Yaṃnūnāhaṃ yvāyaṃ dhammo mayā abhisambuddho tameva dhammaṃ sakkatvā garuṃ katvā upanissāya vihareyyanti. (S i.139) 
“I will reverence, pay my respects, and dwell in subordination to that very dhamma to which I have fully-awakened” 
The Buddha himself does not give up on Dhamma, why should anyone else? This militates against interpreting dhammā as ‘teachings’. Keown’s tentative translation is “…good things (dhammā) must be left behind, much more so evil things (adhammā)” though he affirms the ambiguity. Keown notes that in other places where dhammā and adhammā are contrasted, they seem to mean good things and bad things (p.101). He concludes that the simile has two purposes: 1. to affirm that the dhamma is for the purpose of salvation and no other purpose (this being the main point of the first part of the Alagaddūpama Sutta); and 2. that we must not become emotionally attached to particular doctrines, practices, teachings or philosophical views, and that none should assume a disproportionate status. But that things which are unambiguously evil must certainly be rejected (p.102). Keown is at least thorough and pays attention to the text, and tries to take the text on its own terms, but I still don't find his interpretation satisfying because, again, the Buddha does not give up good things after his awakening. 

Kalupahana (1986: 183) agrees with Keown’s interpretation of adharma in discussing chapter 8 of the Mūlamadhyamakakārikā. “While it is true that the term dharma is used in the Buddhist texts, both in an ontological sense (referring to ‘phenomena’) and in a more ethical sense (meaning ‘good’), there is no evidence at all that the negative term a-dharma was ever used in the former sense.” Thus he treats it as synonymous with akuśala. However we have to take Kalupahana with a grain of salt, because neither the Buddha nor Nāgārjuna thought of dhamma as having an "ontological sense". Indeed both go out of their way to deny this. Dhamma qua phenomena have no ontological status: they are neither existent nor non-existent. It is Kalupahana himself who draws attention to the role of the Kaccānagotta Sutta (S 12.15) in the Mūlamadhyamakārikā. And it is in the Kaccānagotta Sutta where this is plainly stated. Kalupahana himself constantly rejects ontology in his discussion of the texts. His desire to squeeze Buddhism into a Western mould has mixed success. 

Despite this plethora of interpretations by leading interpreters of Buddhism, I can offer yet another. A little later in the Alagaddūpama Sutta one of the bhikkhus asks: “could one be tormented by something externally non-existing (bahiddhā asati)?” The reply is:
“You could, bhikkhu,” replied the Bhagavan. “Suppose one thought like this: ‘it was mine, [now] it is not mine; it might be mine, but I can’t get it.’ They are upset and miserable; distressed and depressed. They are tormented by something externally non-existing.”
By something externally non-existing is meant 'something that they do not possess'. Note here that the thing desired is not non-existent (asati) in the absolute sense, but is merely something lost, or unobtainable. In light of this I suggest that dhammā here could also be ‘things’ (that exist) and adhammā is ‘non-things’ (things that don’t exist in this sense). That is to say we must abandon attachment to what we have, and to what we wish to have. This is not a perfect answer to the problem, but it has the real advantage of not requiring the arahant to give up something that arahants were extremely unlikely to give up!

In the Gārava Sutta (SN 6.2 PTS S i.139) we find the Buddha explicitly turning to the Dharma as his refuge:
Yaṃnūnāhaṃ yvāyaṃ dhammo mayā abhisambuddho tameva dhammaṃ sakkatvā garuṃ katvā upanissāya vihareyyanti. 
I will reverence, pay my respects, and dwell in subordination to that very Dhamma to which I have fully-awakened.
However, no single view of this simile appears to be unproblematic. All we can say with any certainty is that the pop-Buddhism answer that one gives up the Dharma as teaching when one is enlightened is a non-starter. Nor do we give up practising the Dharma. As far as I am aware, none of the enlightened figures of history ever renounced the Dharma. 



Ganeri, Jonardon. 2002. 'Why truth? The Snake Sūtra.' Contemporary Buddhism, 3,2 2002: 127-139.

Gethin, Rupert. 2008. Sayings of the Buddha. Oxford University Press, p.156-167.

Gombrich, Richard. 1996. How Buddhism Began : The Conditioned Genesis of the Early Teachings. London: Athlone.

Horner, I.B. 1954. 'Discourse on the Parable of the Water-Snake.' The Collection of Middle Length Sayings. London: Luzac, p.167-182.

Kalupahana, David J. (1986) Nāgārjuna: The Philosophy of the Middle Way. Mūlamadhyamakakārikā. State University of New York Press.

Keown, Damien. 1992 The Nature of Buddhist Ethics. London: Macmillan, 1992.

Ñānamoli and Bodhi. 2001. The Middle Length Discourses of the Buddha. 2nd ed. Wisdom, p.224-236.

Piya Tan. 2003. Alagaddūpama Sutta: The Discourse on the Parable of the Water-snake [Proper grasp of the Buddha’s Teaching], Majjhima Nikāya (22/1:130-142). Online: http://dharmafarer.org/wordpress/wp-content/uploads/2009/12/3.13-Alagaddupama-S-m22-piya.pdf

Thanissaro. (trans.) 2010a 'Alagaddupama Sutta: The Water-Snake Simile.' Access to Insight. Online: http://www.accesstoinsight.org/tipitaka/mn/mn.022.than.html.


Piotrek said...

About non-existence in the sense of "losing or not being able to obtain something" it's interesting to note that none of the known to me interpretations of SN 44.10 draws on this meaning of atthi/natthi. Yet it seems to be relevant there…

Jayarava Attwood said...


Yes. I think I see what you mean. Indeed if you look at the following passage in M22 (ca. M i.136) you will see that this is precisely what the text says:

“So, bhante, is it possible to be tormented by something internally (ajjhattaṃ) non-existing?”

“It is possible, monk,” replied the Bhagavan. “If one had this view: ‘as the world, so myself: I will exist after death; permanent, constant, eternal; I will remain forever.’ Hearing the teaching of the Tathāgata or one of his disciples for the uprooting of the obstinate, prejudicial bias for speculative views on permanence ; for pacifying all volitions; for the rejection of all foundations; for the destruction of craving; for the cessation of passions; for extinguishing [greed, hatred and delusion] they think: ‘I will definitely be destroyed, I will perish, I will cease to exist!’. They are upset and miserable; distressed and depressed. They are tormented by something internally non-existing.”

Gui Do said...

Yes, some "enlightened figures" have given up the dharma, e.g. Ikkyu who drank alcohol and had whores. As the dharma consists of rules and Ikkyu was an ordained monk, we can say so. It may not mean to give up all, but to be able to give up any attachment to dogma. In Zen it is also interpreted as not to stick on the upaya (means), and depending on the practitioner this may also be meditation itself, as well as other parts of the eightfold path, and it certainly means not to attach to one's own awakening. When this is the case, it is only logic that all means which helped to awaken can be dropped off. Where there was conscious effort, life flows more freely and naturally on "the other shore".

Jayarava Attwood said...

Gui Do

But if he drank and had whores, then was he enlightened? Not from the point of view of the text in question. In Pāli Buddhism such a thing is not possible. So we can cross Ikkyu (and similar figures like Chogyam Trungpa) off the list. The statement stands. Enlightenment is as enlightenment does. It would be better to say that Ikkyu, at one time thought to be enlightened, demonstrated that he was not in fact enlightened through his use of alcohol and whores. As admirable as he might have been otherwise, there's no advantage to Buddhism in bring enlightenment down to that level.

However that is a minor distraction. What we're trying to understand here is a subtle distinction between two closely related words: dhamma and adhamma. As I say, the popular interpretations of this text all seem to overlook the existence of this dilemma - as does your interpretation. No answer which fails to address the question of dhamma/adhamma is interesting since it fails to address the part of the text upon which exegesis ought to turn. Simply ignoring it seems to be a poor way to proceed. Don't you think?

What might be interesting is to learn is how Zen intellectuals understood the terms dharma and adharma as a pair, and especially what Kanji/Hanzi they used to represent them. In what Zen text is this dilemma discussed? What is the history of the idea in Zen exegesis?

Specifics might be interesting. General discussions of Zen (in particular) and popular restatements of Buddhist dogmas (in general) don't interest me at all.

Gui Do said...

(For some reason, before I can send my comment there is always ads popping up, maybe this computer in the internet cafe is infected. I am not sure if you got my first response, in this case please prefer, if any, this one.)

1) When the Buddha spoke to the Kalama in a special context, he did not speak to you or me.

2) If one still tries to get s.th. out of a speech that was not directed at oneself, he or she is very well justified to do this in relation to his time and place, where - contrary to the Kalama - one has the Palicanon and can include it in the scriptures that should not just be believed in, meaning to develop a critical attitude while reading the sutras. This too is exegesis.

3) You e.g. do not only analyze but believe in dhamma and adhamma (thus you think that I misunderstood the text). You also judge Ikkyu within your own system of thinking, based on the words of the Palicanon. If s.o. is stringent here, he would acknowledge that Shakyamuni, the Buddha, was also not enlightened because by his own standards he could not have used abusive speech against Devadatta, to give just one example.

4) In the zen tradition - but this you obviously do not want to discuss - someone is rather considered enlightened when unattached to dogma, rules and words. Good and bad are replaced by shades of gray or just serve as right and wrong in daily dealings.

5) Not being willing to plough a field and instead living on donations and not having sex at all (behaviour that would lead to the extinction of mankind) can of course not be a role model, so your personal opinion that the enlightenment of Ikkyu is at a minor level can probably only be understood by a Theravadin, it is a vicious circle.

Jayarava Attwood said...

1. If the Buddha spoke the Kālāma Sutta he certainly spoke it to and for Buddhists. I am a Buddhist.

2. What? The syntax in this is a bit hard to follow.

3. You have no idea what I believe and clearly do not understand my approach to exegesis.

4. Yes. Right. I don't want to discuss Zen.

5. See 4.

Thanks for your contribution, such as it is, but you apparently have nothing to say on the subject under discussion, and now you're set on criticising me personally.

Please check the rules for commenting. No further warning will be given.

Gui Do said...

Referring to 1) No, the Kalama Sutta is about Buddha speaking to the Kalamas who were not yet Buddhists but followed the Brahmins or were just insecure whom to follow. Thus the Buddha spoke to non-Buddhists.

Jayarava Attwood said...

You won't find much sympathy for your kind of naive literalism from me.

You really ought to read my book on this text. Or any of the several blogs I've written about it.

But basically I think you're reading the wrong blog. This kind of superficial discussion bores me.

So long.

Qianxi said...

Whatever 'dhammas' is, it has to fit in as the raft in the simile. Something that helps you deal with fear, then at a certain point is no longer useful.

I think it works if you see it as 'various teachings' rather than 'The Teachings' as a whole.

I agree the raft simile could have been composed as a separate unit, but the placement behind the snake simile is useful perhaps as an early commentary on the meaning of the text. That context reinforces the idea of using the teachings practically and not for competitive debate.

I had a look for some Chinese Agama parallels and I found Madhyama-āgama 200 and Ekottarika-āgama 43.5. (i formatted an extract from MA 200 and the whole of EA 43.5 here http://pastebin.com/MhBLc043 )

I don't think it changes the picture much, but there are a couple of interesting things:
MA 200 also contains the snake simile and also places the raft simile after the snake simile.
EA 43.5 has the raft simile on its own.
MA 200 is very close to the pali, just slightly expanded.

EA 43.5 is expanded quite a lot. It actually starts saying that if you are kidnapped by bandits you should cultivate equanimity (as in the last of the brahmaviharas), accepting the good and the bad, the clean and dirty with the same attitude, like the earth does. Therefore "you should renounce dhammas of wholesome action, and more so unwholesome dhammas and bad habits."(所以然者,行善之法猶可捨之,何況惡法而可翫習。 translation certainly debatable!)

Only then does it continue with the raft simile. But I don't think this quite works. In this case the raft would be 'wholesome action' or something, but in the description of being kidnapped the emphasis is not on giving up wholesome action, but on not allowing fear and revulsion to arise. So the kidnapping example and the raft simile dont quite add up. And actually, i'm not sure the idea of 'wholesome/unwholesome action' (of the 'dhammas' gloss) is relevant to equanimity (of the kidnapping example), which I thought was a mental state.

After the raft simile in EA 43.5, there is an additional comment. A monk asks: "How should we renounce dhamma(s), and more-so non-dhamma(s)? Don't we practice the path through the dhamma(s)?" 云何當捨於法,而況非法? 我等豈非由法學道乎?'

I think here dhamma(s) 法 could be read as either 'wholesome action' or more generally 'teachings'.

The Buddha's reply is an unhelpful semi-non sequitur, he just talks about the seven kinds of pride, then about his determination to vanquish the prideful Mara on the night of his enlightenment.

The EA is a very confusing text! Lots of layers that don't quite work together.

Does arrival at the far shore necessarily imply awakening? Could it not just be that the emphasis is on the practical use of the boat. Perhaps the metaphor doesn't aim to describe the post-awakening state, it is just an illustration that the Buddha's teachings are practical and not for carrying around on your head.

Although, I must admit, if you only wanted to talk about the practicality of the boat you could have the guy carry it around on his head before he has successfully crossed, and have him complain that putting the boat on his head is not getting him anywhere.

Related Posts with Thumbnails