I suspect that not many list members are interested in this thread,
based on the relative paucity of input. I hope they will tolerate at
least one more post on the topic, perhaps the last lengthy one.
I checked out Don Wilkins' reference from Smyth, and I find that the
example cited does NOT falsify my understanding, because the
pronoun's antecedent is suppressed (what Goodwin calls
"assimilation"). The exact wording of Thucydides' phrase is
BLAPTESQAI AF' hWN hHMIN PARESKEUASTAI ("to be harmed by what has
been prepared for us"). If the antecedent were expressed, the phrase
would have to read, I contend, BLAPTESQAI APO TOUTWN *hA* hHMIN
PARESKEUASTAI, not BLAPTESQAI APO TOUTWN *hWN* hHMIN PARESKEUASTAI.
The antecedent is genitive (object of APO), and the relative is
nominative (subject of PARESKEUASTAI).
Now, if the writer is going to suppress the antecedent (as we do in
English when we say "what" instead of "that which"), he has a choice
to make regarding the case of the relative: leave it in the
nominative or change it to genitive. Leaving it in the nominative
(BLAPTESQAI AF' hA hHMIN PARESKEUASTAI) would be atrocious: APO
absolutely demands a genitive object.
Let me try to summarize my understanding so that it can be criticized
accurately. Here are the three variations of Thucydides' phrase I
have mentioned, all meaning the same thing.
A. BLAPTESQAI *AF' hWN* hHMIN PARESKEUASTAI
B. BLAPTESQAI *APO TOUTWN hA* hHMIN PARESKEUASTAI
C. BLAPTESQAI *APO TOUTWN hWN* hHMIN PARESKEUASTAI
Phrase A. is what he actually wrote. It is, of course, smooth and
idiomatic Greek. Phrase B. is also grammatically acceptable, though
perhaps a bit stilted. Phrase C. is grammatically impossible.
A. is Goodwin's "assimilation"; a species of attraction according to
Smyth. B. involves no case modification of any kind. C. is agreed to
be attraction (though bad grammar in my opinion).
I think Carl Conrad and I agree that A. and B. are good grammar;
perhaps we differ about whether A. is an example of attraction. I
hope we agree that C. is bad grammar. This is the point about which I
am uncertain based on what he has written in previous posts.
It might help for me to reiterate that the point I am raising is
whether the NOMINATIVE can be attracted. Here's another set of
phrases; in this case all grammarians will approve all three (I
hope!), since the cases involved are oblique. All the phrases mean
"to be harmed by what (or 'the things which') they prepared for us."
The relative is now the object of an active voice verb (aorist tense,
but don't try to read too much into it!) rather than the subject of a
passive. I hope I got the form of the aorist right; I didn't want to
change verbs, and I don't have any tools with which to verify the
A. BLAPTESQAI *AF' hWN* EKEINOI hMIN PARESKEUASAN
B. BLAPTESQAI *APO TOUTWN hA* EKEINOI hMIN PARESKEUASAN
C. BLAPTESQAI *APO TOUTWN hWN* EKEINOI hMIN PARESKEUASAN
Again, phrase A. is assimilation/attraction; B. involves no case
modification, and C. is certainly a legitimate instance of
Now, one last point. Along a slightly different line, Smyth observed
something that was quite helpful to me. "Attraction . . . takes place
only (but not always) when the relative clause is essential to
complete the meaning of the antecedent. When the relative clause is
added merely as a remark, attraction does not take place." This, of
course, is the distinction that many grammarians call "restrictive"
vs. "non-restrictive." I had never observed any correlation between
this semantic feature and the surface-structure feature of
attraction. I'll be keeping my eyes open to see how well this works
out in the New Testament.
Smyth's observation is certainly helpful in connection with Acts 1:1,
where the relative clause is clearly restrictive; to use Symth's
words, it is "essential to complete the meaning of the antecedent."
Luke didn't recount "all things," he recounted "all that Jesus began
to do and teach." The restrictive nature of the relative clause
becomes apparent when we consider what difference it would make if we
inserted a comma: "concerning all things, which Jesus began to do and
teach..." Smyth's observation draws attention to the close
relationship between hWN and PANTWN that has driven some of our
number to try to construe the two words together rather than allowing
each to stand in its own clause.
Thanks for your patience with a long post. I'm satisfied now, and I
hope someone else will get some benefit from the discussion, arcane
as it may have been.
In Love to God and Neighbor,
Bob Jones University