I went through the article you mention @thomaslaw and I don’t find it either “very specific” or containing very many “arguments”. In particular I have no idea form the article what note 21 is supposed to tell me. what does
as a term, in sanskrit, tibetan or chinese have to do with the idea that the S collection is either earlier than D and M or that it has a “3-anga” structure?
Here are my notes on the article, forgive the hyperbolic language, it is for my own consumption, I respect the work Choong is doing in the comparative space and agree that his work and that of Yin Shun should be more widely read and engaged with, so take my invective in the spirit of friendly adversary!:
"Essentially, I argue that the three-aṅga structure of the Saṃyukta-āgama/Saṃyutta- nikāya proposed by him should be regarded as more logical and ac- ceptable than Ācāriya Buddhaghosa’s traditional interpretation. In order to present the entire structure and content of the Sarvāstivāda tradition of the Saṃyukta-āgama preserved in the Yogācārabhūmi (Yujia shi de [lun] 瑜伽師地[論]), I first discuss the vyākaraṇa/ veyyā- karaṇa-aṅga, and then the geya/geyya-aṅga and sūtra/sutta-aṅga.
Ācāriya Buddhaghosa explains the veyyākaraṇa-aṅga thus:
sakalam pi Abhidhammapiṭakaṃ, niggāthakaṃ suttañ ca,
yañ ca aññam pi aṭṭhahi aṅgehi asaṅgahitaṃ buddhava-
canaṃ, taṃ veyyākaraṇaṃ ti veditabbaṃ.
The whole of the Abhidhamma-piṭaka, suttas which contain
no verses and any other word of the Buddha not included
in the other eight aṅgas should be known as veyyākaraṇa."
so here Choong sets up an obvious straw man, no one thinks that the abhidhamma are an early strata of buddhist texts, the entire argument of Yin Shun, and the Japanese scholars who precede him, is that the 3-anga (or 9-anga, or 12-anga) structure precedes the 4 prose agamas as an organising principle of the buddha-vacana. So Buddhaghosa’s position is already moot here, it bears on nothing at all, any position is more “logical” than one that is obviously false.
"According to Master Yinshun (1983: I, 3, 8–9 and 24–29), the col-
lections representing vyākaraṇa-aṅga are two sections in the Saṃyukta-
āgama: the ‘Section Spoken by Śrāvakas’ (Dizi suoshuo song 弟子
所說誦, Sanskrit Śrāvaka-bhāṣita) and the ‘Section Spoken by the
Tathāgata/Buddha’ (Fo/Rulai suoshuo song 佛/如來所說誦, Sanskrit
Buddha-bhāṣita).
His reasons are mainly the following:
(a) The Bahubhūmika (Ben difen 本分地) of the Yogācārabhūmi
(in its explanation of the twelve aṅgas) clarifies that vyākaraṇa has
two meanings: 1. exposition of unclear teachings, and 2. declaration
about rebirth, the future destination after death of a disciple
(Yin- shun 1971: 520 and 1983: I 26, and Nakasaki 2004: 53).
(b) In the Taishō edition of the Chinese Tripiṭaka the vyākaraṇa
portion of the Saṃyukta-āgama is marked off by the editors with the
heading Dizi suoshuo song 弟子所說誦 (‘Section Spoken by Śrāvakas’).
(c) The extant Saṃyukta-āgama belongs to the Sarvāstivāda tra-
dition, so the treatment of the Saṃyukta-āgama should also closely
follow that tradition. The names of these two collections, i.e., ‘Sec-
tion Spoken by Śrāvakas’ and ‘Section Spoken by the Buddha’, are
relevant to the Sarvāstivāda tradition of the Saṃyukta-āgama pre-
served in the Yogācārabhūmi (Yinshun 1983: I 3 and 9), as the texts
discussed below show.
(d) The topics (事, vastu) grouped together into saṃyuktas, the
connected units for the content of the Saṃyukta-āgama, indicated in
the Vastusaṅgrahaṇī (She shifen 攝事分) of the Yogācārabhūmi are
shown in sequence thus:
- Spoken by the Tathāgata (如來所說, tathāgata-bhāṣita)
- Spoken by Śrāvakas (諸弟子所說, śrāvaka-bhāṣita)
- Aggregates (蘊, skandha)
- Elements (界, dhātu)
- Sense Spheres (處, āyatana)
- Causal Condition (緣起, pratītya-samutpāda)
- Nutriments (食, āhāra)
- Truths (諦, satya)
- Stations of Mindfulness (念住, smṛtyupasthāna), Right
Effort ( 正斷, saṃyak-prahāṇa), Bases of Supernormal
Power (神足, ṛddhipada), Faculties (根, indriya), Powers
( 力, bala), Enlightenment Factors ( 覺支, bodhyaṅga),
Path Factors (道支, mārga), Mindfulness of Breathing (入
出息念, ānāpānasmṛti), Training ( 學, śikṣā), Definite
Purity/Faith (證淨等, avetyaprasāda).
- Eight Assemblies (八眾, aṣṭau pariṣadaḥ)
A list similar to the above is found in the Bahubhūmika of the Yogā-
cārabhūmi. It sets out the nine topics (九事, navavastuka) that the
teachings of the Buddha or of Buddhas (諸佛語言, buddhavacana)
should contain.
The Mūlasarvāstivāda Vinaya contains a similar list
regarding the content of the Saṃyukta-āgama."
This is, as far as I can tell, the entire argument;
1.) In the Taishō edition of the Chinese Tripiṭaka the vyākaraṇa portion of the Saṃyukta-āgama is marked off by the editors with the heading Dizi suoshuo song 弟子所說誦 (‘Section Spoken by Śrāvakas’).
2.) The names of these two collections, i.e., ‘Section Spoken by Śrāvakas’ and ‘Section Spoken by the Buddha’, are relevant to the Sarvāstivāda tradition of the Saṃyukta-āgama pre- served in the Yogācārabhūmi which preserves a 9 part list including the first 2 sections, (九事, navavastuka) that the teachings of the Buddha or of Buddhas (諸佛語言, buddhavacana) should contain.
and
3.) The Mūlasarvāstivāda Vinaya contains a similar list regarding the content of the Saṃyukta-āgama."
There simply is no force to this argument! it amounts to "a 4th century sarvastivada encyclopedia has the same headings as as a (manuscript witnessed) 4th century sarvastivadan edition of the SA, which is also evident in a (maha)sarvastivadan Vinaya.
This is the exact same problem that Buddhaghosas 4-5th centrury claim has, we have a group, from lets say 800 years after the event, looking back to a pre-existing body of literature that they have held since time immemorial, and claiming (if they are in fact claiming) that the SA is the sutta-geyya-etc. Why should we believe this? apart form the claim in the very late manuscript that alludes to it in the sarvastivada SA,and e refrence in yogacarabhumi of he same period, there is simpoly no evidence, given or inferred, that aught to make anyone think that DA or MA have any less claim to venerability than SA.
Choongs list just goes on repeating the same appeal;
“(e) These topics of the saṃyuktas, the connected units of the entire
Saṃyukta-āgama, are also grouped into three categories, according
to the Vastusaṅgrahaṇī of the Yogācārabhūmi.”
Yogacarabhumi.
“(f) The texts of these two collections, the ‘Sections Spoken by
Śrāvakas and by the Buddha’, are found in the extant Saṃyukta-
āgama. The Tibetan version of the Vastusaṅgrahaṇī of the Yogā-
cārabhūmi confirms the two collections identified by Mukai Akira
向井 亮 (1985: 20–22) (cf. also Yinshun 1983: I 28–9, and Nagasaki
2004: 53, 56–58 and 60).
12
The Saṃyutta-nikāya counterparts of the
vyākaraṇa-aṅga portion in the Saṃyukta-āgama are also found
(Yinshun 1971: 684–694 and 697–701, and 1983: I 32, 43 and 56–
57, and Choong 2000: 21–23, note 22 and 248–250) (see Appen-
dices 1 and 2 below).”
Yogacarabhumi.
And it gets even worse;
"For the vyākaraṇa-aṅga texts (i.e., ‘Sections spoken by Śrāvakas
and by the Buddha’), there are substantial differences between the
Saṃyukta-āgama and the Saṃyutta-nikāya (Choong 2000: 21–22).
The Saṃyutta-nikāya version is not marked off with a heading cor-
responding to Dizi suoshuo song 弟子所說誦 (‘Section Spoken by
Śrāvakas’).
Master Yinshun (1971: 700–701 and 1983: I 32, 43,
56–57) suggested that historically the vyākaraṇa-aṅga discourses
were at first attached to, or subordinated to, the relevant sūtra-aṅga
sections, and that later editors decided to group them into saṃyuktas/
saṃyuttas collected in a single section (Choong 2000: 23, note 22;
also Nagasaki 2004: 52).
That is, the two sections were new crea-
tions within the Saṃyukta-āgama transmission. "
So here it appears that even Yin Shun is admitting that in their scheme the distinction between sutta and vyakarana is actually introduced into the sarvastivada SA, not original, so we have even less reason to think that the original S from which SN and SA and SAB derive preserved a sutta-geyya-vyakarana structure originally.
“However, the next statement, “particularly the entire Sagātha-
vagga in the Saṃyutta”, gives a concrete example of a textual col-
lection representing geyya-aṅga, i.e., the Sagātha-vagga section of
the Saṃyutta-nikāya (SN 1). This is supported by the findings of
Master Yinshun (1971: 517 and 1983: I 23)”
Here again is a literal appeal to religious authority as the only argument given as to why we should take the poetry of the S collection as the “geyya-anga”.
“According to Master Yinshun (1983: I 3 and 9), as mentioned
above, the structure of the Saṃyukta-āgama consists of three aṅgas,
according to the Sarvāstivāda and Yogācāra traditions. It should be
noted that the extant Saṃyukta-āgama belongs to the Sarvāstivāda
tradition, so the treatment of the Saṃyukta-āgama should also follow
closely that tradition.”
This is circular, just because a particular sectarian version of a collection belongs to a particular sect tells us literally nothing about the veracity of claims it makes to a pre-sectarian original, the exact same reasoning would justify taking Therevada commentarial positions on the basiss that the Pali collection belongs to them. it’s incoherent. also I not that the whole circular argument is once again pre-faced with :according to Yin-Shun". “according to Yin Shun” is not an argument.
"Nevertheless, the sūtra-mātṛkā (sūtra matrix, 契經, 摩呾理迦 or
本母), essentially a commentary on a portion of the Saṃyukta-āgama,
in the Vastusaṅgrahaṇī of the Yogācārabhūmi,
20
follows the sequence
of the Saṃyukta-āgama, as was first noted by Lü Cheng 呂瀓 (1896–
1989) (Yinshun 1971: 630–631 and 1983: I 2–3). This discovery
also confirms that the Sarvāstivāda tradition regarding the Saṃyukta-
āgama is attested to in the Yogācārabhūmi. The sūtra-mātṛkā con-
tains only these seven topics:
- ‘Discourses Connected with the Aggregates’
- ‘Discourses Connected with the Sense Spheres’
- ‘Discourses Connected with Causal Condition’
- ‘Discourses Connected with the Nutriments’
- ‘Discourses Connected with the Truths’
- ‘Discourses Connected with the Elements’
- ‘Discourses Connected with the Path: the Stations of Mindful-
ness, etc., of the Enlightenment Factors’
These seven topics (without the sections spoken by Śrāvakas and the
Tathāgata) are considered by Master Yinshun to be the most funda-
mental and earliest portion of the ‘Connected Discourses’ (相應教,
*saṃyukta-kathā) of the Saṃyukta-āgama.
They are found in the
five major sections (varga) on aggregates, sense spheres, causal con-
dition (including nutriments, truths and the elements) and path of the
extant Saṃyukta-āgama/Saṃyutta-nikāya.
These sections of the
‘Connected Discourses’ are identified by Master Yinshun (1983: I
6–12) as the sūtra-aṅga portion of the Saṃyukta-āgama/Saṃyutta-
nikāya "
So once again we have a list, once again it is observed that the “spoken by the buddha” and spoken by the sravajkas" is missing from the list, but nevertheless, without any supporting argument other than “Yin Shun says so” that this is the “sutra-anga” portion of the materials.
“Regarding the sections on the major subject items, i.e., aggregates,
sense spheres, causal condition and path of the extant Saṃyukta-
āgama/Saṃyutta-nikāya, they are evidently the core teachings of
early Buddhism and early Abhidharma Buddhism. For example, the
subject items of the Saṃyukta-āgama/Saṃyutta-nikāya bear certain
resemblances to the structure of these two early Abhidharma books:
the Pali Vibhaṅga and the Sarvāstivāda *Abhidharma-dharmaskandha-
pāda (Apidamo fayun zu [lun] 阿毘達磨法蘊足[論]) (Choong 2000:
252) (see Appendix 3 below).”
Here again we have an argument that is just as easily interpreted as evidence of the LATENESS of S as opposed ot it’s earliness, that it resembles abhidhamma (as movement universally agrred to be late, sorry buddhaghosa). This is not evidence of earliness and it is hard to see how it can be made to be so.
"Also, Sāratthappakāsinī, ‘Revealer of
the Essential Meaning’, is the title of Ācāriya Buddhaghosa’s com-
mentary on the Saṃyutta-nikāya. This suggests that the Pali
tradition also recognized the practical and essential values of the
Saṃyutta-nikāya suttas for Buddhist monks. "
And after relentlessly attacking straw man Buddhaghosa in order to privilege a more or less contemporaneous sarvastivadan commentarial tradition the author then seeks support, again in a very flimsy way, from the author who he has said we should not regard as coherent. weird.
I would like to finish up my analysis of this article by quoting in full the glorious footnote 21:
“As noted in Choong 2010: 57, note 7, the Sanskrit term *saṃyukta-kathā
is inferred from the corresponding Tibetan term, ldan pa’i gtam (ldan
pa’i, ‘connected’; gtam ‘talk, discourse, report’): 即彼一切事相應教間
廁鳩集。是故說名雜阿笈摩 (T 1579 at T XXX 772c23) = gzhi thams cad
dang ldan pa’i gtam de yang dag par ldan pa las ’byung bas na de’i
phyir yang dag par ldan pa zhes bya’o (P 5540, sems tsam, ’i 144a1).
Also, according to the Vastusaṅgrahaṇī of the Yogācārabhūmi, the
Saṃyukta-āgama is the foundation of all four Āgamas (cf. Yinshun
1971: 507–508 and 1983: 7–9 and 39). T 1579 at T XXX 772c23–28 (=
P 5540, sems tsam, ’i 144a1–2): 即彼一切事相應教間廁鳩集。是故說名
雜阿笈摩 = gzhi thams cad dang ldan pa’i gtam de yang dag par ldan
pa las ’byung bas na de’i phyir yang dag par ldan pa zhes bya’o. 即彼
相應教。復以餘相處中而說。是故說名中阿笈摩 = de dang ldan pa’i
gtam nyid rnam pa gzhan du bar gyi mdo sde rnams kyis bstan pas na
de’i phyir bar ma zhes bya’o. 即彼相應教。更以餘相廣長而說。是故說
名長阿笈摩 = de nyid rnam pa gzhan du rgyud ring po’i mdo sde rnams
kyis bstan pas na de’i phyir ring po zhes bya’o. 即彼相應教。更以一二
三等漸增分數道理而說。是故說名增一阿笈摩 = gcig dang gnyis dang
gsum la sogs pa nas gcig nas gcig tu sde tshan gyi tshul gyis ’byung bas
na gcig las ’phros pa zhes bya’o, “Because the connected discourses/teach-
ings (相應教, *saṃyukta-kathā) are grouped together according to all
the topics/subject matters (事, vastu) into connected units (saṃyuktas),
it is called Saṃyukta-āgama. Because the connected discourses are ex-
pounded in another manner by means of medium-sized discourses, it is
called the Madhyama-āgama. Because the connected discourses are ex-
pounded in another manner by means of lengthy discourses, it is called
the Dīrgha-āgama. Because the connected discourses are arranged seq-
uentially in sections going from one [topic/subject matter], to two, three
and so forth, it is called the Ekottarika-āgama.” Thus, according to Yin-
shun 1971 and 1983, the Saṃyukta-āgama is so called because the con-
nected discourses are grouped together according to their topics into
connected units. Then, according to other intensions by means of dif-
ferent structures, the connected discourses associated with their topics
subsequently expanded and yielded the other Āgamas in the sequence
Madhyama-āgama, Dīrgha-āgama, Ekottarika-āgama. Therefore, the
Saṃyukta-āgama is the foundation of all four Āgamas in the formation
of early Buddhist texts, according to the Sarvāstivāda tradition of the
Vastusaṅgrahaṇī of the Yogācārabhūmi.”
This perfectly encapsulates everything I find frustrating with this scholarship. First, it is wildly confusing and verbose, literally opening a footnote with a reference to another footnote form a different piece by the same author, then making an utterly obscure comment about Tibetan terminology before simply quoting, at length, you guessed it, the sarvastivadan encyclopedic commentary of the 4th century, the yogacarabhumi, before appealing to Yin Shun and then simply repeating what the sarvastivadan text has claimed, that the SA is the primary text.
“Thus, ac-
cording to Master Yinshun (1971: 690), these discourses may reflect
the essential characteristics of the early Saṅgha council (saṅgīti) for
collecting the dharmas or ‘teachings’ of the Buddha”
another according to Yin Shun. This one in fairness alluding to a more complex and nuanced argument apparent in formations, that the sarvastivadan accounts of the first council appear to better capture the realistic picture of how exactly the material would be recited, again, I like the argument, but it’s simply not that forceful, being from a period centuries after the events.
"This suggests that the
structure of the two versions is largely pre-sectarian. If the Saṃyukta-
āgama version had the three-aṅga structure, then it is only to be ex-
pected that the Saṃyutta-nikāya version would have been the same. "
This literally contradicts what the author themselves has claimed in the qoute above that the “Master Yinshun suggested that historically the vyākaraṇa-aṅga discourses were at first attached to, or subordinated to, the relevant sūtra-aṅga sections”.
"In my opinion, Bhikkhu Anālayo (2011: 697, note 69) offers no ex-
planation in support of his claim that “this would become a circular
argument”; I have difficulty seeing how it would. "
Really? relying on a 4th century text to reorder an ancient manuscript and then claiming that this new order proves its ancientness because it (now) agrees with the same 4th century text is not only “not circular” but it is impossible for the author to see “how it could be”?!
“This will reveal that the Vastusaṅgrahaṇī list is, in effect, the table
of contents of a text closely resembling the extant Saṃyukta-āgama”
and
“Comparison reveals that most of the nineteen items in this list are
also represented among the Saṃyukta-āgama titles contained in
Table 1 of Appendix 2 below, though with some differences in the
sequence.”
again, this is demonstrably circular if we use it to infer something about the “earliest” S.
In conclusion:
The argument given by Choong in this paper amounts to “Yin Shun says that the Yogacaribhumi says that S is the earliest collection.”
There is a further argument to the effect that “Yin Shun says that the Yogacarabhumi says that S is arranged in a 3-anga structure”.
The first argument is simply the testimony of a tradition, as Analayo points out, it is perfectly possible that it exists simply to justify the (mula)sarvastivada Vinaaya account that gives S as the first agama recited by Ananda. I would add another thought here, which is that we should be unsurprised that the sect renowned as the “abidharma-par-exellance” school, should privilege that collection that most resembles the subsequent systematic treatises is hardly surprising, and once again is suggestive, if we admit that abbhidhamma systematics is evidence of lateness, that the S collection’s resemblance to such systematics is evidence of lateness more than it is evidence of earliness.
The second argument is even weaker than the first. it relies on taking one sutta, MN122/MA191 which gives a shortened list of genres that is given as a list of 9 items in 26 other places and connecting that list to a different list of items in the aforementioned yogacarabhumi that is never really made adequately clear, before then re-ordering all the known collections of S to match the structure thus uncovered, and then claiming that this re-ordering is evidence of the earliness of S. This is circular.
I am currently reading Yin Shun’s “formations” and enjoying it. However, my impression is that it is not ground-breaking or revelatory scholarship, rather it is a nuanced plea for a greater weight being given to sarvastivada materials when assessing speculative reconstructions of the early texts in a context where western scholarship especially has disregarded it in favour of an over-emphasis of Pali. This all seems legitimate and to be welcomed, and there is much food for thought in Yin Shun. But I have yet to come across, either in Choong or in the original, any actually substantive argument that would show that the original S had a “3-anga” structure, or that it was the first of the Agamas to be collected.
All the qoutes and comments above are a very lightly edited copy from my personal rough notes on: (PDF) Ācāriya Buddhaghosa and Master Yinshun 印順 on the Three-aṅga Structure of Early Buddhist Texts | Mun Keat Choong - Academia.edu
Metta.
