EPEI in Rom 3.6 and GAR
alice-iver_larsen at wycliffe.org
Thu Dec 14 02:30:03 EST 2000
Alex Hopkins responded to Iver Larsen who wrote concerning EPEI in Romans 3:6
> Iver, I haven't time to do more than adumbrate a couple of matters.
> Carl wrote,
> To this you replied,
>> In the whole NT (UBS text) GAR occurs 1041 times and PWS 117, but only
>> once do the two occur together - somewhat by coincidence, and that is in Acts
> The statistics are interesting, but I'm not sure what you mean by saying
> that the one collocation of PWS GAR occurs "somewhat by coincidence"; the
> combination does occur elsewhere (outside the NT), and is remarked on by
> Denniston (page 86) who observes that it "confirms a negative statement".
Again, I am coming from the perspective of a discourse linguist, not a classical
scholar, and I must admit that I am not familiar with Denniston. In linguistics
we talk about different hierarchies or levels in the grammar. From the bottom up
these are: morpheme, word, phrase, clause, sentence, discourse. PWS functions
within the clause as an interrogative word. EPEI functions within the sentence
as a subordinating conjunction and connects two clauses together showing a
logical relationship. GAR functions not within the sentence, but within the
discourse. It binds sentences together and it indicates that the sentence in
which GAR occurs picks up a concept in the preceding sentence (it may skip
parenthetical material) and further explains or supports it. It is not helpful
to tag GAR as a subordinating conjunction. It would be better to tag it as what
it is, a discourse particle or a sentence connector. In fact, it occurs so
frequently that it deserves its own unique tag, for instance, "explanatory
discourse particle". In this respect GAR is in the same category as KAI, DE and
other Greek discourse particles. Kai is an "additive discourse particle" (when
it functions in the discourse. KAI can also function at lower levels) and DE is
"contrastive discourse particle." Since PWS and GAR function at different levels
they cannot form a unit in the grammar, and it is linguistically incorrect to
attempt to assign a meaning to a "unit" that does not function as a unit. There
is nothing in principle that prohibits one from occurring next to the other, but
they are independent words that have each their own meaning and function.
> Again, Carl:
> > Whether or not it is etymologically valid, I have come to understand the
> > way GAR works most adequately by seeing it as derivative from an original
> > enclitic, GE that follows upon a preceding word to underscore its
> > importance + an ARA which itself has a broad range of meanings, one
> > significant one being to mark a conclusion being drawn (= "then," "in that
> > case"), the combination yielding the linkage, "Yes indeed, because ..."
> > particularly when GAR follows upon an interrogative word (such as TIS in 1
> > Cor 2:11 or TIS in 1 Cor 2:15), its sense seems to me to be "in that case"
> > or "if that's the case."
> And your response,
> >ARA and GE do occur together in the NT four times. But the order is
> >ARA+GE. The hypothesis that GAR should be a contraction of GE+ARA does not
> >any linguistic basis (why would the final "a" be dropped, if that was the
> >case?), and therefore cannot be used to assist in deriving or describing the
> >meaning or function of GAR.
> I cannot comment on this from a linguistic standpoint, but note that
> Denniston says, "The derivation of GAR from GE and AR, though occasionally
> challenged (as, e.g. by H. Weber in Phil.Rundsch. iv 1078), has been pretty
> generally accepted by scholars." He says, "There appears little reason to
> doubt this etymology," but with that honesty that was alluded to in Carl's
> post of 4th November, he does indicate a couple of points that could be
> marshalled against this view.
> Best wishes,
> Alex Hopkins (Melbourne, Australia)
My first Masters degree was in mathematics, so I am very open to hypotheses.
However, I would like to see supporting evidence or reasons for the hypothesis.
And I do not consider statements like "most scholars agree" as evidence.
Etymologies belong to comparative and historical linguistics. If there is a time
period in the development of the Greek language when GE very often was followed
by ARA and at the same time GAR never occurred, then this would be solid
evidence for GE+ARA developing into GAR, because at the time of LXX and KOINE
Greek GAR is extremely common and GE is never followed by ARA in either LXX or
If there was a time where GE+ARA were a very common combination but GAR was also
common, then one could possibly suggest that GE+ARA contracted to a form GAR
which then fused with the other GAR. This would result in two different senses
for GAR, one of which could be derived from the senses ascribed to GE and ARA.
These are the kind of evidence a linguist is looking for. But if Denniston or
others can give good linguistic reasons for their claim, I am very open to
adjust my current conclusions.
alice-iver_larsen at wycliffe.org
More information about the B-Greek