I wanted to make some remarks about the structure of the new DA
findings, but Stephen and Rod Bucknell beat me to it. Rather than
write it all out again, here's a copy of some notes that Rod sent me
immediately after receiving Hartmann's essay. Please note that these
are preliminary observations only:
The table [below] is based on Table 17.1 in the draft I sent you. I
have added the sutra numbers supplied by Hartmann for the Sanskrit
DA he is studying. It shows that 10 DN suttas are not represented in
the Skt, e.g. MN15 Mahanidana. The interesting thing is that these
are precisely the 10 DN suttas that have counterparts in the Chinese
MA. This fact fits with the proposition that the Skt DA belongs
together with the Chinese MA: they are the DA and MA of one and the
same (Sarvastivada) tradition.
The second observation relates directly to Hartmann's table
1.3, "Table of the sutras and their parallels" (which bears a
remarkable overall resemblance to my table). His table shows that 10
of the sutras in the Skt DA have their Pali counterparts not in DN
but in MN. Now it turns out that not one of these 10 MN suttas has a
counterpart in the Chinese MA, and indeed eight have no Chinese
counterpart at all. This lack of counterparts in MA is in keeping
with the proposition that the Sarvastivadin tradition located these
sutras in its DA rather than in its MA. That is, this second
observation reinforces the first: the Chinese MA and the Skt DA are
mates, complementing each other as regards sutta content.
Nothing remarkably new so far; just confirmation that we do, after
all, know the structure of the Sarvastivadin DA. But this gives us
further data on which to infer how the Collection of Long Discourses
evolved, and I suggest some significant inferences can be drawn..
The two observations amount to the following. (Henceforth DA means
the Skt DA)
1) 10 suttas of DN have counterparts in MA
2) 10 suttas of DA have their Pali counterparts in MN rather
than in DN
These are two different sets of ten. Thus, if we compare the
Sarvastivadin DA+MA with the Pali DN+MN, we find two sets of 10
suttas apparently switching places:
DN 15, 17, 21, 22, 23, 25, 26, 27, 30, 31 have their Sarv
counterparts in MA, not DA
DA 7, 10, 11, 12, 17, 19, 20, 21, 22, 43 have their Pali
counterparts in MN, not DN
As the numbers show, the ten suttas in each set of ten are inclined
to be clustered together. All this strongly suggests that we have
here two vaggas, and that they switched places between the D and M
collections in one or other of the two traditions.
One small problem with this is that DN22 really shouldn't be
counted or should it? The Skt DA also seems to refute my
assumption about the derivative nature of DN 6, 7, 10.
This is a first response. I will meditate further on it all and get
back to you! Anyhow, many thanks again for sending this interesting
material. Do you have any of his earlier publications mentioned in
the article?
Best wishes. Rod.
Comparison of DN and DA, including counterparts in MA.