Jonathan Robie wrote: ↑May 21st, 2020, 12:18 pm
So the hypothesis to test is purely about word order, and under this hypothesis A ἐστιν B => A is the predicate, A B ἐστιν => B is the predicate, and any nominative finite form of ἐστιν will do?
Almost.
You ask a very good question. Under the framework that Mike and I have been working with for a number of years and assumed in our discussion, the phenomenon is not purely about word order, but takes into account the prosody or intonation of the text using word order as clues in conjunction with knowledge of phonology both in Greek and in cross-linguistic studies.
Let me lay out a number of premises to this approach (as I understand them, can't speak for Mike):
1. Greek has a certain class of words known as
enclitics.
2. Enclitics are phonologically deficient in that their accentuation depends on the preceding word. We can call this preceding word the
host of the enclitic.
3. Traditional Greek orthography has a number of rules laying out how their accentuation is affected. If we've studied accents in our traditional grammars, we learned a lot of rules for them, which I will not rehearse, but most common one is flipping a preceding grave to an acute (or equivalently converting a barytone to an oxytone).
4. It is accepted that these rules largely, but not entirely, capture some sort of underlying phonological reality, and this reality makes sense when we compare their behavior to other languages with enclitics.
5. Wackernagel and other philologists have observed that enclitics tend to be in "second position." Other names for this phenomenon are the "peninitial tendency," the "post-positive" property, preferring the P2 position, and the like. These terms may be thought of as roughly synonymous; they are different attempts to grasp at and articulate in somewhat formal terms the same underlying phenomenon.
6. It is important to understand that the second position is not a graphical position. It is not literally the second printed word in a sentence, but in most cases it does correspond to it.
7. Nevertheless, there has been a lot of research over the past century to characterize this second position more precisely, which words on the page "count for position" and where the counting begins.
8. For example, proclitics like the article and prepositions do not count for position. Unfortunately, the orthographic conventions of Greek only disclose their proclitic nature when they have a breathing mark.
9. Fraenkel in particular discovered that there are places internal to the sentence where the counting for position effectively restarts, and he hypothesized that in (some of) these cases the sentence begins with a phonological element he called a
Kolon and in later work he argued that this
Kolon can be very short, even one word in length. These he called a
Kurzkolon.
10. In modern research,
Kola and
Kurzkola generally correspond to what are now know as "left dislocations" (LDs) and other extraclausal constituents (ECCs).
11. Back to enclitics, Greek actually has a number of different classes of enclitics with slightly different characteristics and different host preferences. (To futher complicate matters, some Greek enclitics have orthotonic counterparts that behave like full phonological words and are not subject to the peninitial tendency. Unfortunately, the orthography does not always reflect this difference, though it is sometimes reflected in the accentuation.)
12. The postpositive connectors (e.g., δέ, γάρ, etc.) are one such class; enclitic pronouns (με, σε, etc.) are another; and, importantly for our discussion, the present indicative of εἰμί is another.
13. These classes have different host preferences. For example, the postpostive connector can be hosted by proclitics, while the other classes generally need full phonological words (that, words with an accent)
13. When ἐστίν is an enclitic, we expect it to tend to the second position, and accordingly we can diagnose some sort of phonological prominence on its host. (Traditionally, such phonological prominence has been called emphasis.)
14. This phonological prominence still remains to be better characterized but it seems that there is a tendency for the first full word of a phonological segment (whether you call it a colon, intonation unit, or whatever) to be the most phonologically prominent element of the whole segment.
15. It is hypothesized that this prominence corresponds, under certain conditions still being worked out, to
focus in information structure.
16. Focus basically is a way to mark (an element of) the most informative part of the sentence, and this is sometimes referred to as the assertion.
17. In copular sentences (e.g., those with the linking verb ἐστίν), the assertion corresponds to the traditional predicate.
18. In other circumstances, however, phonological prominence can be used for tasks like establishing new topics.
19. In copular sentences (e.g., those with the linking verb ἐστίν), the topic corresponds to the traditional subject.
20. It can require great sensitivity in reading the context to figure out what the topics and assertions (and in our case, the subjects and predicates) are and how they develop throughout a passage.
21. Sometimes the way that an author arranges a series of topic and subtopic in the discourse is called the "topic structure."
Back to 1 Cor 11:3 παντὸς ἀνδρὸς ἡ κεφαλὴ [ὁ] Χριστός ἐστιν, we notice that the enclitic ἐστιν comes rather late in the clause. It tells us that the word Χριστός bears some emphasis/phonological prominence, but because it comes so late in the sentence, we have to account for the preceding παντὸς ἀνδρὸς ἡ κεφαλή.
One way to do so is to take it as a left-dislocation, which is common for establishing new topics. If so, then it is the subject, and [ὁ] Χριστός ἐστιν is the predicate. That's basically Mike's observation. What's interesting to me, that it's not just Mike and I who read it this way. Some Byzantine manuscripts of the New Testament are punctuated (albeit not quite consistently), and as see below, one of them that I looked at actually separates the pieces with a comma:
- [GG] Def 1C11.3 ms.png (587.07 KiB) Viewed 935 times
1 Cor 11:3 transcription wrote:
θέλω δὲ ὑμᾶσ εἰδέναι· ὅτι παντὸσ ἀνδροσ ἡ κε-
φαλὴ, ὁ χ(ριστό)σ έστιν· κεφαλὴ δὲ γυναικὸσ, ὁ ἀνήρ·
κεφαλη δὲ χ(ριστο)ῦ ὁ θ(εό)σ.
So in sum the hypothesis to test is that when we have
A, B ἐστιν and A is left dislocated, then A is the subject and B is the predicate.
Jonathan Robie wrote: ↑May 21st, 2020, 12:18 pm
How should I treat hyperbaton in examples like τὸ γὰρ ἐν αὐτῇ γεννηθὲν ἐκ Πνεύματός ἐστιν Ἁγίου?
We can take τὸ γὰρ ἐν αὐτῇ γεννηθὲν as the left-dislocated subject, so ἐστιν merely falls in its expected second position. The hard part actually happens in other cases when the expected hyperbaton is not there.