First, what type of compound is "paṭicca-samuppāda".
Ole Holten Pind wrote:
This compound is somewhat peculiar. pa.ticca is an absolutive. Now
absolutives do not normally occur as first member in compounds. In the
present case we need a syntactical complement to understand it e.g. a noun
in the accusative like hetum or kara.nam or any other term in the
accusative. As you can imagine the fact that absolutives normally denote
actions preceding the action denoted by the finite verb, provided that the
two actions have the same agent, this particular compound has generated a
heated controversy among Buddhists interested in grammar, because the action
denoted by the absolutive normally precedes that of the finite verb. Now all
we can say is that pa.ticcasamuppaada means "origination dependent (on
something). " The usual translation "dependent origination" is meaningless
and ungrammatical, besides being not very intelligent considering the
canonical context, in addition to the grammatical constraints on the
semantics of absolutives. Cf. the vinaya term pa.ticcakamma which means " an
action that is due (to someone else, i.e. caused by someone else). " For
instance, the crime that someone who has made you commit would be a
pa.ticcakamma. In short, it is a "syntactical compound" in the sense that it
is syntactically dependent upon an explicit or implicit term that is
independent and syntactically external to the terms of the compound.
Therefore the peculiar term "syntactical compound."
Rett Thiele wrote:
My best guess is that it would be considered a tappurisa of a sort called 'nicca' (unanalysable). However it's obviously not a 'classic tappurisa' (suddhatappurisa) where the first member stand in an oblique case relation to the latter member when analysed.
As far as I can see, Aggava.msa (author of Saddaniiti) doesn't really take a clear stand on this point. In Paa.nini, tatpuru.sa has the more general sense 'determinative compound' in addition to the more restricted sense of a compound where the first element stands in an oblique case relation to the second when resolved (the latter being the one we are most familiar with). In this system, pa.ticcasamuppaada can belong to a class of irregular determinative (tatpuru.sa) cpds called mayuuravy.amsaka-s. Some examples of the latter (ashtadhyayi II.1.72) are pitvaasthiraka and nipatyarohi.nii. (having drunk, steady? I.e. able to hold his drink? Or strong after drinking?) and (having fallen, stood up? Standing up after falling down?).
So far, however, I haven't seen this more general sense of tappurisa explicitly stated in Aggava.msa. He goes straight to defining tappurisa as a compund with an oblique (from accusative onwards, lit. amaadayo, having the (ending -am and so forth) relation between the words.
My assumption (from Kahrs) is that Aggava.msa has borrowed material from mainstream Sanskrit grammar without always maintaining its structural integrity or systematic function in its context. Aggava.msa was more of a compiler than a systematizer or synthesizer.
The main place where Aggava.msa mentions the cpd pa.ticcasamuppada (and the similar upaadaayaruupa.m/upaadaaruupa.m) is in sutta 683 of the suttamaalaa, where he states that compounds beginning with an absolutive are "niccasamaasa". This is a type of compound (samaasa) which is always (nicca.m) a compound. The point is that you can't resolve the compound into its component parts. (at least not using only the words in the compound). (in Sanskrit: nityasamaasa)
As Ole pointed out, the analysis of pa.ticcasamuppaada would require another word , for example "hetu.m paticca samuppada". Aggava.msa illustrates the same idea in this way: a~n~nama~n~na.m pa.ticca sahite dhamme uppaadetii ti pa.ticcasamuppaado. "Mutually depending on connected things it arises" = dependent origination.
Another sort of compound which is considered 'nicca' is what is called an upapadasamaasa "prefix (upapada) compound". In Pa.nini this is specifically a compound where a word is prefixed to a verbal root, and the root becomes an action noun through addition of a suffix, but where that action noun could not exist apart from the prefix. Ex: kumbhakaaro: jar maker, potter.
This upapadasamaasa, which is nicca, is included in the discussion of tappurisas, not surprisingly since the is an accusative relation between 'making' and the thing made (the kumbha).
Since this latter 'niccasamaasa' is taken as an example of a tappurisa, I would be inclined to take cpds whose first member is an absolutive also as such, though again, in the wider sense of tappurisa....
...Sutta 707 appears to go into this.
Alan McClure wrote:
Of course, much of the above exchange is dealing with grammar and technical terms, etc. However, there is an important reason that I wanted to go into all of this. If it is indeed true that "pa.ticca-samuppada" is different from a "normal" determinative compound, i.e. it is a "syntactical compound," then might it be translated differently too?
Dr. Pind notes that this idea is a source of Buddhist debate, and I can see why, for it does indeed seem that the common translation of "Dependent Origination" is not quite accurate. Again, Dr. Pind says that:"all we can say is that pa.ticcasamuppaada means "origination dependent (on something). The usual translation "dependent origination" is meaningless and ungrammatical, besides being not very intelligent considering the canonical context, in addition to the grammatical constraints on the
semantics of absolutives."
"Origination dependent upon [something]," however, is not a very smooth translation. However, it seemed to me that it is more specific and accurate, so I started to think about what could replace the "something" so that the translation would sound better and would be more accurate grammatically, etc. than "Dependent Origination." Since the compound is actually lacking a concept, the "something" the idea was to figure out what that something should be in English.
As I thought about different possibilities, I recalled a recent post by Dmytro where he mentions that "paccaya" in the context of pa.ticca-samuppada can be translated as 'requisite condition,' no doubt focusing on the fact that other conditions can be present as well, but that the preceeding condition, or the one that the orgination of the following condition is dependent upon, is required.
So, it occured to me, that along these lines, a good translation of pa.ticca-samuppada might be:
"Origination that is dependent upon a requisite condition"