Entering and Remaining: A Non-Obvious Chinese Parallel

One of the little mysteries of the Madhyama Āgama to me for the past couple years has been a verbal expression that occurs quite often at the end of sentences (in Indic fashion): 成就遊. It’s peculiar to two Chinese Buddhist texts: The Madhyama Āgama and the older translation of the Abhidharma Vibhāṣā. Its most frequent usage is as the main verbs in the dhyana formulas, but it occurs in other contexts.

成就遊 is actually two verbs. 成就 means to accomplish or achieve something completely. 遊 means to go for a walk, wander, tour, to travel without a destination. As a result, it sometimes refers to the peripatetic lifestyle of Indian ascetics. Indeed, 遊 occurs in the title of the Dīrgha Āgama’s parallel to the Mahāparinibbāna Sutta (DĀ 2 遊行經, lit. “The Journey Sutra”). These two verbs are almost always tacked onto the end of sentences in the Indic way of verbs coming last (which is not the normal Chinese S-V-O sentence structure).

Let’s look at an example. MĀ 2’s description of the first dhyāna reads:

“Furthermore, the noble disciple is secluded from desire and secluded from bad and unskillful things. With thought/feeling and contemplation, that seclusion gives rise to joy and happiness, and they attain the first dhyāna, accomplishing and traveling [it].”

The confusion is caused by the presence of a verb in the usual Chinese word order: 得 (“to attain”) 初禪 (“first dhyāna”). So, what’s 成就遊 doing there? It seems superfluous. Up until recently, it’s been a head scratcher how to best render this expression, especially 遊. Unsure what it’s intended to translate, I’ve thus far passed over it, translating the passage as “attained accomplishment of the first dhyana” and moving on with my life.

The translators of the BDK edition of MĀ 2 didn’t fare any better, rendering it as “… dwells having attained the first absorption.” They’ve chosen to read 成就遊 creatively to mean something like the Pali reading.

Which brings me to the non-obvious parallelism taking place with this Chinese expression that I’ve noticed on a closer look. The problem appears to be that the Chinese translators interpreted these two verbs, which were probably Prakrit versions of P. upasampajja viharati, differently than Pali translators today.

P. upasampajja can mean to reach, attain, enter upon. This is close to the meaning of accomplish or achieve, though there’s a sense of arrival. P. viharati can mean to live or dwell, but also to sojourn in an area. So, it stands to reason that viharati has been read to mean “sojourn” rather than simply “dwell.”

When we look at the Chinese verb clause 得初禪成就遊, I think what we are seeing is a verb translated twice: In the Chinese way and in a literal Indic way. One is active voice, and the other is passive, but actually I think they both translate the same verbs in the original formula.

To the Chinese reader, it ends up meaning, “attains the first dhyāna, accomplishing and travelling [it].” This might have been to make sure it’s understandable to Chinese readers, who would be confused if 得 wasn’t there. Without it, it would read nonsensically as “the first dhyana accomplishes and travels.” We know that many early translation projects had native Chinese literati touch up texts to make them easier to read. 得 may have been inserted in that editing process after the passage was translated literally.


Does this only occur for the first jhana or all four?

1 Like

The first dhyana was just an example. It’s an expression used in all four dhyanas as well as a few other contexts like the four immeasurables and formless samadhis in the Madhyama Agama.

1 Like

It means evaluative thought (Anguttara Nikaya 5.28). This ‘objectivity in reserve’ identifies and rejects the drawbacks of each successive stage, propelling advancement.

" "And furthermore, the monk has his theme of reflection well in hand, well attended to, well-considered, well-tuned[1] by means of discernment.

“Just as if one person were to reflect on another, or a standing person were to reflect on a sitting person, or a sitting person were to reflect on a person lying down; even so, monks, the monk has his theme of reflection well in hand, well attended to, well-pondered, well-tuned by means of discernment. This is the fifth development of the five-factored noble right concentration.”

The analogy of three steps of relaxation means the evaluation applies to each stage.

I’m sorry, Paul. The essay was about what is probably the Chinese translation of P. upasampajja viharati in the Madhyama Agama. Are you talking about vitakka?

1 Like

I didn’t see this the first time around @cdpatton, but I was actually just wrestling with this last week. I’m still working on translating An Shigao’s 「人本欲生經」(T 14) and, let me tell you, his rendering (or “renderings,” because he’s not consistent) of upasampajja viharati is quite interesting, to say the least. I’m on the road at the moment; could I check back in in a few days and show them to you? Maybe you’d have some insight on how to tackle it. (In fact, you’d probably have quite a bit of insight on how to approach the discourse.)


Greetings, @cdpatton and All:

Where the Mahānidāna Sutta has this:

Sabbaso rūpasaññānaṁ samatikkamā paṭighasaññānaṁ atthaṅgamā nānattasaññānaṁ amanasikārā ‘ananto ākāso’ti ākāsānañcāyatanaṁ upasampajja viharati, ayaṁ catuttho vimokkho.

Sabbaso ākāsānañcāyatanaṁ samatikkamma ‘anantaṁ viññāṇan’ti viññāṇañcāyatanaṁ upasampajja viharati, ayaṁ pañcamo vimokkho.

Sabbaso viññāṇañcāyatanaṁ samatikkamma ‘natthi kiñcī’ti ākiñcaññāyatanaṁ upasampajja viharati, ayaṁ chaṭṭho vimokkho.

Sabbaso ākiñcaññāyatanaṁ samatikkamma nevasaññānāsaññāyatanaṁ upasampajja viharati, ayaṁ sattamo vimokkho.

Sabbaso nevasaññānāsaññāyatanaṁ samatikkamma saññāvedayitanirodhaṁ upasampajja viharati, ayaṁ aṭṭhamo vimokkho.

the Ren ben yu sheng jing has this:

一切從色想已度、滅地想、若干想不念,無有量空慧已受竟 ,辟天名為‘空慧’:是名為第四解脫處。



一切從不用慧得度,無有思想、亦不無有思想竟受止,辟天名為‘思想 ’:是名為第七解脫處。

一切從無有思想 竟得度,滅思想、亦覺盡,身已更竟受止 :是為第八解脫處。

The bold-type above show the corresponding phrases in the Chinese and Pāli. AS far word-order, position in the sentence, and the grammar and syntax implied by the rest of the sentence, they tally pretty well. Additionally, Vetter, in A Lexicographical Study of An Shigao’s and his Circle’s Chinese Translations of Buddhist Texts, suggests 受竟 , 受止, 竟受止 (even 受竟行below) to be read as renderings of upasampajjā viharati. (I’m sorry, I can’t recall the page number off-hand, but cf. entry for 空 at Vetter 224.)

Despite knowing pretty much for certain what the Indic source word was, I can’t do what you did @cdpatton with your passage: reconstruct the translation process. I just don’t see what An saw. Obviously, the fact that each of the five phrases translates slightly differently makes it much harder. The difficulty is further compounded by the fact that An uses similar wording (but not exactly the same, though possibly intended to be the same: one can never really tell with An) phrasing for the formless states in the satta viññāṇaṭṭhiti list, where we would not expect to find upasampajja viharati (at least not as per the Pāli).

Any ideas?

1 Like

Well, that level of inconsistency makes me conclude the passage has been damaged by copying errors over time, so I wouldn’t spend too much time trying to make sense of it. I’d instead spend some time trying to reconstruct the original translation.

One thing I noticed while searching for term usage in T14 is this very similar passage:

[0245a17] 「有不色,為令從是,一切從色想度,多想滅,為無有量空空慧受意止,辟天名為空慧行,是名第五識止處。
[0245a20] 「有不色,為令從是,一切從空行竟,過無有量識,從慧受意止,辟天名為識慧,是名為第六識止處。
[0245a23] 「有不色,為令從是,一切從識慧過度,無有量不用從是慧意受止,辟天名為不用從受慧,是為第七識止處。

The last one looks to have suffered a transposition, otherwise I’d say that 受意止 is the actual translation in your mystery passages, which has been almost completely garbled. Someone probably miscopied 意 as 竟, which is a common visual typo, and then the passages got worse as subsequent copyists tried to “correct” it to something that made sense.

意止 was an old translation of S. smṛty-upasthāna, so I’m guessing then that the translation is functional rather than literal. It means something like “gets that abode of mindfulness.” Or something.

The fact that this passage from the abodes of consciousness matches the passages in the eight liberations makes me wonder whether the original actually has that expression upasampajja viharati in it. It seems like it may have been different.

This, BTW, is why I’ve stayed away from the old translations in the Taisho. Some of them are just in really bad shape. It feels like I’m just making stuff up, trying to correct the garbled passages, and I don’t like that feeling as a translator. As a creative writer, it’s great; as a translator I need to feel grounded in what the original author is telling me, not floating off on some speculative reconstruction.


Obviously, that can’t be discounted and was, in fact my first thought, as it would be anyone’s. But there is SO much inconsistency in all of An Shigao’s works that it defies even simple attribution to copyists’ errors. Again, this is not to discount that, it’s just that inconsistency (between texts, within the same text) is actually the only consistent thing with him. In your opinion, how sure can we be that at least some portion of doesn’t go back to him? Or, probably, the better question is "how much can we reasonably attribute to him, and how much to the copyists?

I think I have to answer the question above before I can make decisions on reconstruction. I just don’t feel comfortable presuming consistency in his work when I have no way of knowing.

Yes, passage you give comes from the seven stations of consciousness section which should definitely not contain the upasampajja viharati phrase. So that’s a mystery: why is it there? Or is it? Yes, of course, 意 for 竟 seems logical, but 竟 is attested in lists of meditation/cosmological stages, either denoting transcendence of a requisite stage before moving unto a subsequent one, or as part of a phrase corresponding to upasampajja viharati. That, coupled with the fact that there should be nothing in this seven stations section about upasampajja viharati, gives me pause.

(Just as a side note, the 慧 you see popping up in the passage you highlighted is almost certainly part of the name of the formless attainments. Don’t ask me how!)

I actually got through the vast majority of the discourse relatively unscathed. There are only about two or three sections with lots of technical terminology that are… not even troublesome–just impossible!

But I gather from your comments that this wouldn’t be a “fun and interesting challenge” for you. I understand. This kind of work is more for people who enjoy abuse and self-torture.


No, it’s insight.

“A calm mind supports the development of insight and the
presence of insight in turn facilitates the development of
deeper levels of calmness (Dhammapada 372).”—Analayo

“There was the case where Sariputta — quite secluded from sensuality, secluded from unskillful qualities — entered & remained in the first jhana: rapture & pleasure born of seclusion, accompanied by directed thought & evaluation. Whatever qualities there are in the first jhana — directed thought, evaluation, rapture, pleasure, singleness of mind, contact, feeling, perception, intention, consciousness,[2] desire, decision, persistence, mindfulness, equanimity, & attention — he ferreted them out one after another. Known to him they arose, known to him they remained, known to him they subsided. He discerned, ‘So this is how these qualities, not having been, come into play. Having been, they vanish.’ He remained unattracted & unrepelled with regard to those qualities, independent, detached, released, dissociated, with an awareness rid of barriers. He discerned that ‘There is a further escape,’ and pursuing it there really was for him.”—Majhima Nikaya 111

1 Like

He was a human being, so I doubt his translations were perfect. Inconsistency is natural and requires editing to eliminate. (Believe me, I’ve spent the past couple months trying to eliminate them from my own project.) Beyond that, it would depend on how large the texts were, how much time passed between start and finish of the translation, and so on. For the most part, translators stumble over the same words or expressions, so there are patterns of meaningful vacillation. Copyist errors are more random and often make little or no sense.

But this isn’t just an issue with An Shigao, apart from his texts being very old and obscure. They are on the extreme end of the corruptions I see in Agama translations. As far as the general condition of Chinese Agamas goes, I’d attribute that to their obscurity in Chinese Buddhism. Popular Mahayana sutras don’t have the same level of damage as the Agama translations do. They just didn’t get the same attention and TLC. They may have only existed as a handful of copies in the back of temple libraries for centuries, in some cases.

Something you might look into, if you haven’t already, is whether any better manuscripts have been discovered that aren’t as bad as the Taisho edition. I recall that there was an old Chinese translation discovered in a Japanese temple archive that was much better preserved, but I don’t recall which it was exactly offhand. There are scholars who spend their time tracking down the oldest witness they can find to study how texts change.

Okay. Do you have an example? It’s possible the typo went in the reverse direction, 竟 being changed to 意, but the level of consistency in that other passage makes it look intact to me compared to the section on the liberations.

If I’ve learned anything in the past three years of translating Agamas, it’s that they often vary from Pali passages. It’s a rarity, really, that there’s a complete, word-for-word correspondence. If there’s just a few extra synonyms added in one or the other version, it’s close enough to call them identical. But it’s difficult to tell when the Chinese translators aren’t translating word-for-word.

Yeah, my thought reading it was that he may have been naming the gods rather than the heavens in those passages. I believe that in Pali the gods in the formless heavens are called “cognizers of x” (space, consciousness, etc) and I’ve seen that in Chinese translations, too. Otherwise, I suppose he may be translating āyatana as a faculty rather than a place.

1 Like

I can’t read Chinese, but as to the Pali, I translate upasampajja viharati with ‘has attained’. Because the verb viharati often functions like an axiliary, I feel, meaning nothing much more than ‘to be’, which is also one of the glosses of the PED. Anandajoti has likewise said that “Upasampajja viharati is a periphrastic construction giving durative sense”. Periprhasic, for those who read along, meaning the two verbs don’t convey different ideas but just a single one together, as in “I am gooing to have to walk” is four verbs but the idea is one.

Perhaps the Chinese translators were not sure whether it was such a periphrasis or not, so they translate it as “attain”, and then somehow added the others? I don’t know. As I said, not familiar with Chinese. But perhaps this helps somehow.


Thanks for the Pali perspective. That is the typical way Chinese translators handle it, too. They use “attain” or “enter” for the verb. The Madhyama Agama is a weird case that throws English translators for a loop. It wasn’t until I looked at the Pali closely that I realized what was happening.

1 Like

Ah, that makes sense. Thanks. Kind of like อยู่ in Thai

1 Like

Thanks charles,

Just to be clear, are you saying upasampajja viharati is translated by one verb in Chinese? If so, is it 得?

you don’t think 成就遊 renders upasampajja viharati and it is alien to the Pali?

(To me these just look like squiggles to be honest. But it’s good to have some validity for my translations in the Chinese, in case I ever need to argue for them.)

No, I think the verbs were translated twice. They were translated literally as 成就 (“accomplish”) 遊 (“roam about”) at the end of the sentences, and then more naturally as 得 (“to attain”) in the position Chinese readers (like English readers) expect verbs to be. What confuses us is that 遊 (“roam”) isn’t the normal way of understanding viharati, but it isn’t an impossible reading. That and the fact that a couple verbs are just dangling there at the ends of sentences when there’s already a verb on hand doing its job just fine.

How that happened probably has to do with the translation process. By the time the Madhyama Agama was translated to Chinese, the translators had gotten fairly organized, using a division of labor. It worked like this:

  1. A Central Asian or Indian monk recites the text to be translated from memory.
  2. A translator monk, usually a bilingual monk who is native to Central Asia or India, translates the recited text to Chinese.
  3. A Chinese scribe who knows the written language writes down the translation.
  4. A Chinese literati proofs and edits the draft to make sure it’s understandable to a Chinese reader.

My theory is that the translator translated the jhana formula extremely literally like a word-for-word litany, placing the verbs at the end of the sentences. So, the editor at the end of the process decided to touch up the translation by adding 得 to the sentences. At least then there’s an obvious verb and predicate, even if the bit at the end is still confusing. The oddest thing about it, though, is how consistent it is. These verbs occur in the four jhanas, the four imeasurables, the four formless samadhis. Maybe in a couple other contexts I’m forgetting. It’s just like the Pali parallels.

This didn’t happen in other Chinese translations. They have other issues, like inconsistent translations and typos, but they handled the verbs sensibly. But, if the Madhyama Agama weren’t translated in this awkward way, we wouldn’t have direct evidence that the original had two verbs. So, it’s good for EBT studies, bad for readers, I guess.