Yes, please.
The list is here. Please let me know if this is sufficient. I think Rod is in retreat right now so not sure when to ask him.
Thanks so much. Rod’s out of retreat now; which of course doesn’t mean he has free time, but we’ll see.
Just to be clear, what we need him to do is to go through the items that differ in these sets and determine if they are full or partial. Eg is AN 3.37-38 full or partial parallel with AN 3.70, and so on.
In some sets there’s a ‘1’. What does this mean?
I had a look at the file, and as these things usually go, i thought I’d check out one or two cases. I was writing to Rod to explain what the job was, and I realized I didn’t understand it well enough to be able to describe it, and then, well did the lot. Or those involving Indic texts, anyway, I didn’t touch the Chinese. To be honest, I still don’t really understand the whole 'data" thing, and I appreciate taking some time to work with it so I can grok it a little better.
Have a look at the uploaded file. I went through and checked each case where the Indic text was doubled up, and determined if it was full, partial, or not a parallel. Obviously this only refers to if they are parallels with other Indic texts, not the Chinese. I’m not entirely sure this is what is needed, but anyway, let me know.
inferred_parallels.txt.zip (1.9 KB)
I think it will be good to discuss the data set as it is in more detail. I have the feeling we have different interpretations of the dataset.
Basically, the data set we have at present does not contain all the data about a parallel we would like to have. For instance the set:
[‘an10.83’, ‘sa-2.200’, ‘’, ‘’]
[‘sa-2.200’, ‘sf78’, ‘1’, ‘’]
This means that sa-2.200 and sf78 are partial parallels (the “1” stands for partial). But is sa-2.200 a partial of sf78 or the other way around? We simply don’t know and at the moment have no way of notation for this. What I’m doing at present is just to convert what we have to JSON (and add some more like the DHP parallels). And then hopefully in the future we can refine it.
Now the problem with the above set is that sa-2.200 is mentioned as a full parallel of an10.83. So that would infer that an10.83 and sf78 are partial parallels of each other too. Or not?
I looked at the work you have done but I’m also not sure I understand what you mean. So just to get this right:
[‘sn35.160’, ‘sa207’, ‘’, ‘’]=
[‘sn35.99’, ‘sa207’, ‘’, ‘’]
You mention that the ‘=’ is a full parallel, but does that mean that they are all 3 full parallels of each other: sn35.160 = sn207 = sn35.99? Or is sn35.160 a full parallels of sn35.99 and the pair might be a partial or full parallel of sa207 but you don’t know?
[‘sn36.1’, ‘sa473’, ‘’, ‘’]X
[‘sn36.11’, ‘sa473’, ‘’, ‘’]
Does this mean that the set sn36.1, sa473 should be deleted because it is not a parallel relationship? So only the parallel sn36.11 = sa473 remains? Or do you mean that sn36.1 is not a parallel of sn36.11 but either one might still be a partial parallel of sa473 (actually implying that both sets are partial).
[‘an3.37-38’, ‘ma202’, ‘’, ‘’]
[‘an3.70’, ‘ma202’, ‘’, ‘’]
[‘an8.42’, ‘ma202’, ‘1’, ‘’]
[‘an8.43’, ‘ma202’, ‘’, ‘’]
an3.70, an8.42 and an8.43 are partial parallels with each other.
You mention that the 3 AN texts are partial parallels of each other, but the problem here is ma202. Logic would have it that you if the 3 AN texts are partial parallels of each other, they are also all partial parallels of ma202. Correct?
[‘an2.42’, ‘sa-2.363’, ‘’, ‘’]
[‘sn2.25’, ‘sa-2.363’, ‘’, ‘’]
[‘sn9.13’, ‘sa-2.363’, ‘’, ‘’]
sn2.42=sn9.13. an2.42 is not a parallel.
The relationship sn2.42 = sn9.13 is a new parallel to add but is not related to this question. The question is about the relationship to sa-2.363.
This is actually an interesting case, possibly unique. sa-2.200 and sa 996 are full parallels. They tell of an occasion when ven Purnika went to see some wanderers, who asked whether it is true that the Buddhs teaches the complete ending of rebirth (or something like that). Punniya confirms this. then he goes to the Buddha and askes whether he answered correctly, and the Buddha confirmed that he did.
Now, sf78 begins with Purnika leaving the wanderers, and then speaking with the Buddha. Remember, this is a “fragment”, so presumably the first part of the sutta is lost. In any case it should be considered a partial parallel of the other two.
AN 10.83, on the other hand, has nothing in common with these apart from the fact that it features a “Punniya”. So it should not be parallel with this set at all. But here’s where things get weird.
AN 10.83 is also listed as full parallel with SN54.10, AN3.91, and AN3.84. However it has nothing in common with any of them, nor do they have anything in common with each other. This is bad!
AN 10.83 is also listed partial parallel of AN 8.82, and this is correct.
Something’s gone badly wrong with this set, and it is good that we are going through this process so we can pick up such cases. I am guessing that our "inference engine’ is to blame. (I haven’t checked the data, I am reading this from the site)
Generally speaking the notion of partial as we use it that two things may be partial parallels of each other. In other words its a mutual relation. This includes cases where one is “part” of the other, and also cases where two things are “partially” parallel in the sense that they are similar to each other. Once you get into the nitty-gritty it’s just not easy to specify things more precisely.
But as you say the point for the time being is simply A) we don’t lose any info, and B) we have the potential to enrich it as time goes on.
All my remarks in the file pertain only to the Indic texts (and in a couple of cases the Tibetan). So I haven’t checked sa207 at all.
In this case, however, the original data specifies that sn35.160 = sa207, and sa207 = sn35.99; and I have confirmed that sn35.160 = sn35.99, so we can indeed abbreviate this set as sn35.160 = sa207 = sn35.99
No, like I said I’m only referring to Indic texts.
Yes.
Well, they are both specified as full parallels of sa 473. However, as you imply, this seems somewhat unlikely, since they share nothing in common with each other, apart from a short generic statement on vedana. I have just checked the Chinese, and in fact they should both be listed as partial parallels of sa 473. sn36.1 and sa473 share a set of verses, while sn36.11 and sa473 share an introduction and basic teaching, but sn36.11 is much more elaborated so shouldn’t count as full parallel. (If you compare the two, the Buddha’s answer is just one sentence in Chinese, but a long paragraph in Pali)
If only it were that simple! Again, checking the Chinese, it turns out that AN 3.70 and MA 202 are both teachings to Visakha on the three kinds of uposatha. Thus they are full parallels with each other.
Now, AN8.42 and AN 8.43 are in fact almost identical, differing only in that AN 8.43 is set with Visakha again, while AN 8.42 has no setting. Thus strictly speaking AN 8.42 is “part of” AN 8.43. However, given that that “part” is in fact about 95% of the text, they could well be counted as full parallels.
I didn’t do this since my understanding is that underneath the notion of full parallel is not merely that two texts are similar, but that they appear to stem from the same source “thing”. In this case we can’t say that, as they may be merely the same teaching on different occasions. Nevertheless, this is far from ironclad logic, and you could just as well, in fact probably better, argue that they most likely do stem from the same thing and one is merely extracted from the other, as happens quite a few times in the Anguttara. So let me revise this and say that they are full parallels of each other. That means we have:
The set [‘an8.42’,‘an8.43’] is a partial parallel with the set [‘ma202’,‘an3.70’]
Sorry, this was a typo, it should read
sn2.25=sn9.13. an2.42 is not a parallel.
Again, I have just checked the Chinese, and it turns out that sa-2.363 is also a full parallel, so [sn2.25=sn9.13=sa-2.363].
an3.70 merely shares a stock phrase in common, so should not be considered a parallel at all, and should be deleted.
Some Therigatha parallels:
SN5.2 Soma/Thig 3.8 Somātherīgāthā
SN5.5 Uppalavaṇṇa/Thig11.1 Uppalavaṇṇātherīgāthā
SN 5.6 Cālāsutta/Thig 7.2 Cālātherīgāthā
SN5.7 Upacālāsutta/ Thig 8.1 Sīsūpacālātherīgāthā
2 Likes
Thanks, we’ll add these when we get a chance. When the Dhammapada is done, we can look to the remaining verse collections.
OK. This is my fault for taking random examples from the dataset rather than making up a fake examples. So let me try again. Consider the dataset:
[‘A’, ‘B’, ‘’, ‘’]
[‘B’, ‘C’, ‘1’, ‘’]
So B and C are partials. Is B partial to C or C partial to B? We don’t know.
But if B and C are partials and A and B are full parallels, A and C should also be partials. So question to Rod is if that is correct.
In your example:
[‘A’, ‘B’, ‘’, ‘’]X
[‘C’, ‘B’, ‘’, ‘’]
You mention (with the “X”) that A and C have nothing in common. That implies that both A and C are both partial parallels of B … always! So I guess those I can simply mark as partial and remove from the list for Rod.
In this case however:
[‘A’, ‘B’, ‘’, ‘’]*
[‘C’, ‘B’, ‘’, ‘’]
You indicate with the “*” that A and C are partial parallels. This then means that at least one of them, A or C, or both, is a partial parallel to B because they cannot be both full parallels to B. So This is something Rod has to find out.
I will put the info you have given above about he specific sets into the data.
Thank you very much for this. There are indeed similarities in the first two and last one you mention and I have added those to our list.
I fail however to see the parallels in sn5.6 and thig7.2 apart from the title. thig7.2#5 is parallel with dhp191.
Well, yes, it’s true that we don’t know that. But that’s not what “partial parallel” means in our system. It’s a mutual relation: they are partial parallels of each other. That’s all the data specifies. To say “one is partial of the other” is to introduce a direction, an asymmetrical relation. We may well want to enrich our data at some point by adding this information, but we should not think of our current data as implying this.
In a perfect world, yes. Sadly—and I hope this is not too much of a shock for you—this is not a perfect world! In reality, “full parallel” almost always means something rather less than “absolutely identical”. In other words, parallel is not equivalent. That means that this might not always work out.
This kind of relation would be an “inferred partial” parallel. In such cases, while it will usually be correct, it is probably best to check.
Yes, that’s correct.
Generally, yes, i would agree. It’s possible, given the looseness of the categories, that this might not always apply, but on the whole that would be the case. So yes, they should be checked.
I really like to have my world working with clearly defined mathematical rules!
1 Like
Oh, don’t we all!
I’ve updated the list for Rod. I’m off to Hamburg for the Vesakh now … back on Tuesday.
Sorry, my mistake. SN5.6 Cālāsutta verse no 1 and 2 of the gatha (Kiṃ nu jātiṃ na rocesi…tasmā jātiṃ na rocaye) should be paired with Thig7.3 (Upacālātherīgāthā) verse no 2 and 3 (Kiṃ nu jātiṃ na rocesi…jāto dukkhaṃ nigacchati).
1 Like
A few parallels, these are kind of found in our cross-ref data, but this should be clearer:
SN 14.6=Iti78=sa446=sa448
Thag 265-266=Thag147-148=SN14.16#9-11=Iti78#6-8=sa446=sa448=mil7.7.4#6*
To explain, the first set are all full sutta parallels. These consist of a samyutta sutta with verses.
The second set is just the verses.
Note that in the Chinese texts the doctrinal passages are completely elided, so they can be considered as sutta parallels (if the elided parts are restored) or as verse parallels (if considered just as the extant Chinese text).
The Milinda reference has only one of the verses. In addition, it is a “mentioned” parallel; the text specifies that it is a quote from the Samyuttanikaya.
Thank your for expanding my English vocabulary (“elided”). 
Done that. Just next time, can you also mention in which Thag the specific verse is mentioned? The numbers have changed in Thag with the new numbering system so I take it that the numbers you mention are the pts verse numbers that are on SC at the moment. (i.e. Thag 265 = thag3.16#265, etc.)
I have also uploaded a new version of the list for Rod which also included double LZH Vinaya parallels.
I was using the SC verse numbers, but I think its the same.
Also SN10.1.9 and Thig 3.6 Sukkātherīgāthā
1 Like
They are overlapping for a large degree. The SC numbers mentioned on the pages have changed to starting with 1 and there is a new category called “SC verse number”. Most of the time it is the same number indeed.