[B-Greek] Assimilation of the Relative (was Attraction oftheRelative ...)
Iver Larsen
iver_larsen at sil.org
Wed Dec 16 04:00:16 EST 2009
Dear Elizabeth,
It is always difficult to know how much background one needs to include in a
short e-mail in order to communicate clearly. Since you did not understand what
I said, I need to add some background below:
----- Original Message -----
From: "Elizabeth Kline" <kline_dekooning at earthlink.net>
To: "greek B-Greek" <b-greek at lists.ibiblio.org>
Sent: 15. december 2009 22:44
Subject: Re: [B-Greek] Assimilation of the Relative (was Attraction
oftheRelative ...)
> Thank you Iver,
>
>
> On Dec 14, 2009, at 11:42 PM, Iver Larsen wrote:
>
>> Cooper's theory does not stand up to scrutiny, which is easy to see
>> if you look
>> at relatives in the accusative. The first one in the GNT is in
>>
>> Mat 2:9: ὁ ἀστήρ, ὃν εἶδον ἐν τῇ
>> ἀνατολῇ, προῆγεν αὐτούς
>> hO ASTHR, hON EIDON EN THi ANATOLHi, PROHGEN AUTOIS
>> the star, which they had seen in the east, went ahead of them
>>
>> In this case the relative clause qualifies a noun in the nominative.
>> In that way
>> it is like an adjective and it is part of the noun phrase that
>> functions as
>> subject in the main clause. A sentence as a whole does not have a
>> function in
>> another sentence. A sentence does not have a case function. It is
>> phrases that
>> have syntactic functions in a clause and these functions are marked
>> by case. hON
>> is accusative above because it is the object of EIDON within the
>> relative
>> clause.
>>
>> An interesting example of the nominative relative is found in
>>
>> Mat 18:28: ἐξελθὼν δὲ ὁ δοῦλος ἐκεῖνος
>> εὗρεν ἕνα τῶν συνδούλων αὐτοῦ, ὃς
>> ὤφειλεν
>> αὐτῷ ἑκατὸν δηνάρια
>> EXELQWN DE hO DOULOS EKEINOS hEUREN hENA TWN SUNDOULWN AUTOU, hOS
>> WFEILEN AUTWi
>> hEKATON DHNARIA
>> That slave went out and found one of his fellow slaves, who owed him
>> hundred
>> dinarii
>>
>> hOS refers back to hENA - the head of the noun phrase "one of his
>> slaves", and
>> it is in the nominative because it functions as subject in its own
>> relative
>> clause.
>>
>> So, attraction does not occur for nominative or accusative
>> antecedents, only for
>> genitive and dative - unless someone can show me an example against
>> this
>> theory.
>
>
> I don't see how the accusative examples really undermine G.Cooper's
> approach.
Which shows that I need to explain it better.
>
> It seems reasonable to start out with certain assumptions about
> G.Cooper. Generally speaking someone who manages to get an academic
> press (U Michigan) to publish a Greek reference grammar running four
> volumes and 3500 pages will probably be aware that assimilation of the
> relative is found with datives and genitives, a fact that should be
> known by any third semester seminarian. Cooper mentions this
> restriction in the quoted paragraph but I snipped "..." it out.
>
> Cooper's treatment in vol 1 51.10.0-10.13 and vol 3 51.9.0-9:12 is a
> detailed nuanced discussion of a complex topic. I quoted less than one
> paragraph from several thousand words and hundreds of examples. We
> cannot hold e-mail discussions about details if we are always required
> to build the grammar from the ground up. So why don't start by
> assuming we are dealing with intelligent and informed authors,
> competent in their field rather than waste time typing paragraphs like
> this one?
>> A sentence as a whole does not have a function in
>> another sentence. A sentence does not have a case function. It is
>> phrases that
>> have syntactic functions in a clause and these functions are marked
>> by case.
>
>
> It appears that we have a problem here with meta-language. To say
> that a relative sentence/clause can function as a complex **virtual**
> constituent in another sentence/clause is a perfectly acceptable
> statement in the several linguistic frameworks I am familiar with.
> This is reflected in the way subordinate sentence/clauses are linked
> to their antecedents when building parsing diagrams of complex
> hypotaxis found in NT books like Hebrews. The subordinate sentence/
> clause as a whole can be understood as performing some function
> relative to sentence/clause which contains the antecedent. I don't see
> anything controversial about Cooper's proposal.
The problem we are dealing with may in part be caused by failure to
differentiate between a clause and a sentence. So, let me clarify my framework.
I thought you were familiar with the following, but maybe you are not, and
others may not be. So, here goes:
I assume a grammatical hierarchy where structure and function are important
concepts. The grammatical hierarchy starts with morphemes at the bottom. A
morpheme has a function in the hierarchy above which is the word level. A word
has a structure in terms of morpheme constituents. A word functions in the next
higher level which is the phrase. A phrase has a structure in terms of words. A
phrase functions in the next hierarchy which is the clause and a clause has a
structure in terms of constituent phrases. A clause functions in the next
hierarchy which is the sentence. A sentence consists of one or more clauses, one
of which is usually the main clause, others subordinate, but coordinate clauses
also exist. A sentence functions in the next hierachy which is the paragraph,
and above that you come into episodes and discourse, but the terms are not quite
fixed at the higher level.
Now, a relative clause is in this system called a rank-shifted clause because it
is not a constituent of a sentence or a clause nor does it have a function in a
sentence or a clause. It is a constituent of a phrase and it normally modifies a
head noun similar to how an adjective, demonstrative, numeral etc modify the
head noun.
It is from this background that I could not make much sense out of the quote you
gave from Cooper. Maybe if I had access to his many pages and had read them, I
would understand him better.
I have no idea what linguistic framework would accept "that a relative
sentence/clause can function as a complex **virtual** constituent in another
sentence/clause". Maybe you can tell me which framework you refer to?
> Assimilation of the relative is a special case, the general rule
> (Cooper v1 51.9.1) "The case of relative is determined within the
> relative sentence ... accordingly the relative may be any of the the
> four cases depending on the function ..." within the relative sentence/
> clause. We need an explanation of assimilation which goes beyond
> attaching a label "attraction" which tells us nothing about the why
> assimilation takes place. Cooper ventures in to the domain of "telling
> us why", there are with out a doubt going to be objections to his
> proposal, but the other grammars in my library don't even venture into
> the domain. The seem to assume that attaching a meaningless label
> "attraction" solves the problem.
My first task as a descriptive linguist is to analyse the text and discover
rules, and none of the quotes I have seen so far give a clear and comprehensive
analysis. After that I may or may not try to suggest a reason why the langauge
works the way it works.
> Let's not get bogged down arguing about G.Cooper. If someone wants to
> propose another explanation of assimilation which takes into account
> all the data, I am willing to listen.
It is a big task to take all the data into account, and I am not saying that I
have done that. However, my approach is to look at the data first and try to
analyse them before looking into what others may have written. I have limited
access to grammars, and BDF did not give an adequate analysis of this.
Where I have come to so far is that all the data I have looked at suggest that
the assimilation between the relative and its antecedent in terms of case does
not take place in nominative and accusative, but it normally does take place in
genitive and dative, although there are cases where it does not take place, and
this then needs to be explained.
I have other things to do, so I am not ready with a comprehensive analysis yet.
Another factor is the relative placement of the relative pronoun and its
"antecedent", which in my framework is called the head noun.
Another factor is that the head noun is at times not explicit, so that the
relative can stand alone without such a head.
Another factor is that the relative often functions both as demonstrative and
relative merged into one.
It is really too complex to describe all this properly in a short e-mail.
All for now,
Iver
More information about the B-Greek
mailing list