[B-Greek] Subject, or appositive to the subject indicated by the person inflection?

Chet Creider creider at uwo.ca
Fri Mar 3 10:16:29 EST 2006

Randy Leedy wrote:
> It seems to me that I have seen in numerous grammar books the idea that
> the nominative that is commonly called the subject of a finite verb is
> actually an appositive renaming and specifying the subject indicated by
> the verb's personal ending. I am not aware of any widespread tendency,
> though, to insist on this complication in naming the use of the
> nominative with 3rd-person verbs. Any grammarian who insisted on calling
> it an appositive to the subject contained in the verb's person
> inflection, it seems to me, would have no need at all of a
> nominative-case usage category named "subject."
> But what about 1st- and 2nd-person verbs? For example, Eph 4:13 has
> MECRI KATANTHSWMEN hOI PANTES... ("until we all arrive"). I don't see
> any grammatical reason that I should not call this nominative the
> subject, even though English does not provide a construction in which
> "all" can function directly as subject of a first-person verb. It seems
> to me that the person information that fills the English subject slot is
> in Greek contained in the verb ending, freeing the Greek subject slot to
> accept a subject specifier that the English slot cannot accept without
> losing the person information. A colleague of mine takes the opposite
> view, that a nominative with a 1st- or 2nd-person verb cannot be
> construed as its subject. I have not yet found the question directly
> addressed in any of the half dozen or so Koine grammars on my shelf.
> I would appreciate any educated opinions and especially any pointers to
> any sort of grammar or linguistics literature where the question is
> treated explicitly.

There are two very interesting questions here for which I only have time 
for a few hastily written notes which don't do justice to the topics.

(1) Is the "nominative that is commonly called the subject of a finite 
verb ... actually an appositive"?
This idea doubtless has considerable antiquity.  I only mention three 
versions of it:
(a) In an interesting article, "On Semantically Based Grammar", 
published in 1973 in Linguistic Inquiry, the UCLA linguist Ed Keenan 
suggested that the inflections marking subject (and object if present) 
on a verb in some languages are actually pronominal arguments of the 
verb.  At the time, I think it would be fair to say, linguistic theory 
was not at a stage where this idea could be proven or disproven.
(b) The Homeric scholar, Egbert Bakker, suggested in an article written, 
I believe, in the 80s (I can supply the reference but don't have it at 
hand), that postverbal subjects in Homer were actually appositional 
"afterthought" words.  Implicit in this claim is the idea that the true 
subjects then are the verbal desinences.
(c) In recent years, primarily in connection with various Amerindian 
languages, the idea has been advanced very explicitly that subject (and 
sometime object) inflections can be incorporated pronominal arguments 
rather than just agreement morphemes.  This works is particularly 
associated with Eloise Jelinek (Linguistics (Emeritus), U Arizona). 
Joan Bresnan (Stanford) and Sam Mchombo (UC Berkeley) have discussed in 
considerable detail in an article in Language the case of Chichewa, a 
Bantu language spoken in Malawi (and some surrounding countries) in 
which they argue that some elements of verbal morphology are 
incorporated pronouns and others agreement morphemes.  The argumentation 
is very carefully done.

I am not aware that anyone has made a similar argument for any variety 
of Greek, including NT Greek, and FWIW, I was not convinced myself by 
Bakker's arguments for Homeric Greek.  However, I haven't thought this 
matter through for Greek myself.  Until someone does, the best advice I 
can give is to stay with the traditional view that the inflections on 
the Greek verb are agreement morphemes and that when an overt subject 
noun is present this is the true subject.  Note with reference to the 
example in Eph 4:13, MECRI KANTANTHSWMEN hOI PANTES, most linguists 
today would feel that hOI is the subject or the head of the subject noun 
phrase.  This is known as the DP (Determiner Phrase) hypothesis and is a 
separate matter.

(2) Is there a need for "a nominative-case usage category named "subject"?
Most grammars of Greek (e.g. Smyth) state that the nominative is the 
case of the subject, but this is only a careless statement of an 
introductory sort as these same grammars go on to discuss subjects in 
other cases (for example that the default subject of an infinitive is in 
the accusative case, that of a genitival absolute in the genitive case). 
  In addition of course the nominative case is used for elements which 
are not subjects (such as predicate adjectives and nouns). In a 
just-published article, I've argued that the facts of agreement in Greek 
force upon us the need to recognize covert elements, specifically 
subjects (and probably objects as well).  Consider, for example, the 
following sentence from Homer in which a participle in the accusative 
masculine plural depends on an infinitive (with no overt subject) which 
in turn depends on a finite verb with an overt masculine plural subject. 
  There is no way to explain the accusative case on the participle 
without assuming some kind of (covert) subject for the infinitive:


Here is a word for word gloss: then me(acc.) then first-of-all 
companions(nom.) they-begged words(dat.) cheeses(gen.) they-taking(acc.) 
to-go back.  Roughly translated: thereupon (my) companions then first of 
all begged me with words (for them) to go back taking (some) of the 
cheeses.  {I'm sure Carl can improve on this.) Traditional grammarians 
didn't have any problem with the notion of a covert subject ("understood 
subject"), but obviously it would be simpler in a carefully worked out 
theory of language if one could do without them.

The point of relevance to Randy Leedy's query is that yes there is most 
definitely a need for a notion of subject in Greek.  (I suspect that the 
reason the ancient grammarians didn't recognize the notion is that 
subject in Greek appears is so many different guises.)  The case is 
determined by the syntactic context and the number, person, gender 
(where relevant, as with participles) is determined by the semantics.

Reference: Creider, C. and R. Hudson. Case agreement in Ancient Greek: 
implications for a theory of covert elements. In: Word Grammar: New 
Perspectives on a Theory of Language Structure, K. Sugayama and R. 
Hudson, eds. London: Continuum, 33-53 (2006).

Sorry for the detail, but your questions raise very complicated issues.

Chet Creider
Professor of Anthropology &
Co-Director, Linguistics Programme
University of Western Ontario
London, Ontario, Canada

More information about the B-Greek mailing list