Word Order
clayton stirling bartholomew
c.s.bartholomew at worldnet.att.net
Fri Oct 6 23:51:04 EDT 2000
on 10/06/00 6:04 PM, Moon-Ryul Jung wrote:
> The above discussion led me to think that perhapse the unmarked word order
> of Koine Greek is a partial order. Given for examples elements A,B,C,D,E,
> a partial order may be defined by A < B, C< D , D < E. In this order,
> C can be placed anywhere, the order between A and C is not defined,
> etc. Any ordering can be said to satisfy this partial order as long as
> it is true that A < B, C <D, D < E.
Moon,
I think that this would be generally accepted by NT Greek students if you
limited it to constituents like prepositional phrases. The preposition
generally precedes its case but the prep. phrase can be located freely
within the higher level constituents (clause, paragraph). It would also
apply to article -> substantive patterns. I think it would be difficult to
find anyone who would argue with you about this.
On the other hand, if you try to make statements like this with pairs like
SV or VO or SO you are going to get have a fight on your hands.
Lets digress from your proposal a little. First of all the term "word order"
is probably infelicitous since the problem isn't really one of words but
functional units which we might as well call constituents because that is
the term which the guys from MIT have been using for 40 years now and it is
pretty well understood.
I did some more work today on this question and the more work I do on it the
less I agree with the posts I sent yesterday. This is a learning process.
I tried to apply the new information (NI) vs. old information (OI) typology
to a short span of Mark, applying it to Markan narrative only. The reason I
am restricting this to narrative is that reported speech and quotes from the
LXX are a whole different discourse "ball game" from narrative. Anyway, even
with these restrictions I found that the NI/OI typology was very difficult
to use. Givon's rule that there is only one "chunk" of NI per clause didn't
seem to pan out in practice. Some clauses seemed to have almost no old
information with the exception of anaphoric pronouns.
There did seem to be a pattern of OI preceding NI in the clauses I looked
at. However, the difficulty I had distinguishing between OI and NI didn't
give me much confidence in the results.
One discovery I made in the process was that other kinds of pragmatic
categories seem to intrude into the analysis while one is trying to
distinguish OI from NI. Notably the pragmatic categories of Topic and Focus,
Foreground and Background seemed to be always "in my face" when I was trying
to distinguish the new information (NI) from the old information (OI).
At one point I just gave up on OI/NI and decided to look for Foreground (FG)
and Background (BG) instead. This was a lot more fruitful. I used verb types
(finite, participle, infinitive) as one of the primary criteria. Participle
constituents functioning either as adjectives or adverbs are almost without
exception background (BG) material. The main verb (finite) in a clause is
typically foreground (FG).
I have not done thousands of lines yet, my sample was quite small. But
looking only at Markan Narrative it seemed like FG constituents were almost
always found after one or more BG constituents in the same clause. I also
found some sandwich clauses with BG -> FG -> BG patterns. There were some
clauses which had multiple fronted BG constituents: BG -> BG -> BG-> FG.
I think this line of investigation, using pragmatic constituent categories
to discuss NT Greek constituent order has the potential of being much more
fruitful than endless squabbles about SVO, VSO and so forth.
As I said before, the results of this kind of investigation may provide some
data from which some cautious general statements can be derived about SVO
issue.
Enough for now. I am going back in my cave for a while.
Clay
--
Clayton Stirling Bartholomew
Three Tree Point
P.O. Box 255 Seahurst WA 98062
More information about the B-Greek
mailing list