syntax of disjunctive phrase (Matt. 5:36)

[Charlie Law]

I have never known what to make of this sort of syntax in the ὅτι clause of Matt. 5:36:

... ὅτι οὐ δύνασαι μίαν τρίχα λευκὴν ποιῆσαι ἢ μέλαιναν

I know that the syntax of each language is governed its own logic -- or to put a finer point on it, the syntax of each speaker's speech is governed by its own logic in a specific context -- but I just don't understand the logic behind placing λευκὴν 'white' before the infinitive ποιῆσαι 'to make' but putting the remainder of the phrase after the infinitive.

It seems counterintuitive that the two object complements in the phrase λευκὴν ἢ μέλαιναν 'white or black' would be separated like this.

Can anyone account for this? I don't see any sort of discourse motivation or grammatical or lexical constraint at work. Does it have to do with a limitation on the number of items that can fill that pre-verbal complement slot? Are there possibly other factors like euphony or some sort of prosody (of which I know next to nothing)? I'm shooting in the dark here.

[Stephen Carlson]

Nice example. You can find some discussion of this phenomenon in the literature under the rubric "hyperbaton" and especially in Devine & Stephens' book Discontinuous Syntax: Hyperbaton in Greek.

I agree that it appears counterintuitive to English speakers (it certainly did to me), but there is a lot of evidence for it in Greek, not just the NT. While different theorists may have different explanations, it does appear that Greek speakers had a preference for putting only one phonological word of a larger constituent in the pre-verbal focus slot (or initial position in a prosodic domain, depending on your theory).

[Jonathan Robie]

Weird language ;->

I'm sure you know that the objects of the verb can precede it, and you seem to know about fronting. I suspect the part that is throwing you off is the split constituent λευκὴν ἢ μέλαιναν, with the verb ποιῆσαι landing smack dab in the middle of one of its objects. Here are three examples where it happens with ποιῆσαι:

μήτε ἐν τῷ οὐρανῷ, ὅτι θρόνος ἐστὶν τοῦ Θεοῦ· μήτε ἐν τῇ γῇ, ὅτι ὑποπόδιόν ἐστιν τῶν ποδῶν αὐτοῦ· μήτε εἰς Ἱεροσόλυμα, ὅτι πόλις ἐστὶν τοῦ μεγάλου Βασιλέως· μήτε ἐν τῇ κεφαλῇ σου ὀμόσῃς, ὅτι οὐ δύνασαι μίαν τρίχα λευκὴν ποιῆσαι ἢ μέλαιναν.

λευκὴν ἢ μέλαιναν.

Τῷ δὲ δυναμένῳ ὑπὲρ πάντα ποιῆσαι ὑπερεκπερισσοῦ ὧν αἰτούμεθα ἢ νοοῦμεν κατὰ τὴν δύναμιν τὴν ἐνεργουμένην ἐν ἡμῖν, αὐτῷ ἡ δόξα ἐν τῇ ἐκκλησίᾳ καὶ ἐν Χριστῷ Ἰησοῦ εἰς πάσας τὰς γενεὰς τοῦ αἰῶνος τῶν αἰώνων· ἀμήν.

ὑπὲρ πάντα ὧν αἰτούμεθα ἢ νοοῦμεν

οὔτε ἁλυκὸν γλυκὺ ποιῆσαι ὕδωρ.

γλυκὺ ὕδωρ

I'm not sure what the right explanation is. I can think of at least two, and this does happen with other verbs and other moods. One is to say that this is like Matthew 1:20:

τὸ γὰρ ἐν αὐτῇ γεννηθὲν ἐκ Πνεύματός ἐστιν Ἁγίου

I've heard a few explanations for the interruption in Matthew 1:20, one involves enclitics, another involves predicate constructions ...

But there's at least one other explanation that might explain the three examples above that use ποιῆσαι: it might actually be an indicator of an implied verb that is not explicitly repeated. If so, we might interpret these constructions more or less the same way we interpret Mark 3:4:

καὶ λέγει αὐτοῖς Ἔξεστιν τοῖς σάββασιν ἀγαθὸν ποιῆσαι ἢ κακοποιῆσαι, ψυχὴν σῶσαι ἢ ἀποκτεῖναι;

I'd really like to hear other opinions on this ...

[Stephen Carlson]

But there's at least one other explanation that might explain the three examples above that use ποιῆσαι: it might actually be an indicator of an implied verb that is not explicitly repeated.

You mean like this?

ὅτι οὐ δύνασαι μίαν τρίχα λευκὴν ποιῆσαι ἢ μέλαιναν [ποιῆσαι]

or even:

ὅτι οὐ δύνασαι μίαν τρίχα λευκὴν ποιῆσαι ἢ [μίαν τρίχα] μέλαιναν [ποιῆσαι]

I'm not a fan ellipsis theories myself. Another, somewhat related idea, is an afterthought in a separate unit "... that you cannot make one hair white--or black." At any rate, hyperbaton is a more general phenomenon, and this appeal to ellipsis seems to reach only some of them.

[Jonathan Robie]

While different theorists may have different explanations, it does appear that Greek speakers had a preference for putting only one phonological word of a larger constituent in the pre-verbal focus slot (or initial position in a prosodic domain, depending on your theory).

Hmmm.

An awful lot of multi-word constituents appear before the verb, even when an object is fronted.

Matt.16.3 τὸ μὲν πρόσωπον τοῦ οὐρανοῦ γινώσκετε διακρίνειν, τὰ δὲ σημεῖα τῶν καιρῶν οὐ δύνασθε;
Matt.16.3!10 τὸ πρόσωπον τοῦ οὐρανοῦ

Luke.12.56 ὑποκριταί, τὸ πρόσωπον τῆς γῆς καὶ τοῦ οὐρανοῦ οἴδατε δοκιμάζειν, τὸν καιρὸν δὲ τοῦτον πῶς οὐ δοκιμάζετε;
Luke.12.56!2 τὸ πρόσωπον τῆς γῆς καὶ τοῦ οὐρανοῦ

John.8.44 ὑμεῖς ἐκ τοῦ πατρὸς τοῦ διαβόλου ἐστὲ καὶ τὰς ἐπιθυμίας τοῦ πατρὸς ὑμῶν θέλετε ποιεῖν.
John.8.44!9 τὰς ἐπιθυμίας τοῦ πατρὸς ὑμῶν

2Cor.5.2 καὶ γὰρ ἐν τούτῳ στενάζομεν, τὸ οἰκητήριον ἡμῶν τὸ ἐξ οὐρανοῦ ἐπενδύσασθαι ἐπιποθοῦντες, εἴ γε καὶ ἐνδυσάμενοι οὐ γυμνοὶ εὑρεθησόμεθα.
2Cor.5.2!6 τὸ οἰκητήριον ἡμῶν τὸ ἐξ οὐρανοῦ

2Pet.1.15 σπουδάσω δὲ καὶ ἑκάστοτε ἔχειν ὑμᾶς μετὰ τὴν ἐμὴν ἔξοδον τὴν τούτων μνήμην ποιεῖσθαι.
2Pet.1.15!11 τὴν τούτων μνήμην

I can easily do queries that find split constituents vs. constituents that wholly precede the verb, but I don't yet have an explanation that fits all the data I've seen. I'd love to hear one.

[Jonathan Robie]

At any rate, hyperbaton is a more general phenomenon, and this appeal to ellipsis seems to reach only some of them.

I don't have an explanation that fits all the instances I see. I may just need to see the right explanation. Or there may be more than one reason for hyperbaton in Greek.

I just noticed that ellipsis seems to fit pretty well with the examples that use this particular infinitive verb in the Greek New Testament. It may not be the right explanation, it's just one possible guess.

[Stirling Bartholomew]

Something may be learned looking at the variations in the witnesses to the text.

Matt. 5:36
R-P μητε εν τη κεφαλη σου ομοσης οτι ου δυνασαι μιαν τριχα λευκην η μελαιναν ποιησαι
NA27 μήτε ἐν τῇ κεφαλῇ σου ὀμόσῃς, ὅτι οὐ δύνασαι μίαν τρίχα λευκὴν ποιῆσαι ἢ μέλαιναν.

The Byzantine text including the TR show verb final position. A decent textual apparatus (Reuben Swanson) shows the same verb after δύνασαι in front of μίαν τρίχα in Codex Bezae Dc(*) and f1. The Gothic Version agrees with the Byzantine text[1].

[1] A STUDY OF PURPOSE, RESULT, AND CAUSAL HYPOTAXIS IN EARLY INDO-EUROPEAN GOSPEL VERSIONS,
JOSEPH ALLEN PENNINGTON, p423 (443 pdf).

[Jonathan Robie]

Something may be learned looking at the variations in the witnesses to the text.

Matt. 5:36
R-P μητε εν τη κεφαλη σου ομοσης οτι ου δυνασαι μιαν τριχα λευκην η μελαιναν ποιησαι
NA27 μήτε ἐν τῇ κεφαλῇ σου ὀμόσῃς, ὅτι οὐ δύνασαι μίαν τρίχα λευκὴν ποιῆσαι ἢ μέλαιναν.

The Byzantine text including the TR show verb final position. A decent textual apparatus (Reuben Swanson) shows the same verb after δύνασαι in front of μίαν τρίχα in Codex Bezae Dc(*) and f1. The Gothic Version agrees with the Byzantine text[1].

[1] A STUDY OF PURPOSE, RESULT, AND CAUSAL HYPOTAXIS IN EARLY INDO-EUROPEAN GOSPEL VERSIONS,
JOSEPH ALLEN PENNINGTON, p423 (443 pdf).

Interesting. Here's Bezae, according to Alan Bunning's site:

μητε εν τη κεφαλη σου ομοσησ οτι ου δυνασαι ποιειν τριχα μειαν λευκην η μελαιναν

[Stephen Carlson]

While different theorists may have different explanations, it does appear that Greek speakers had a preference for putting only one phonological word of a larger constituent in the pre-verbal focus slot (or initial position in a prosodic domain, depending on your theory).

Hmmm.

An awful lot of multi-word constituents appear before the verb, even when an object is fronted.

As far as I can tell, all your examples involve topics, not focus constituents. (Topical constituents are fairly often multi-word.) And none of them involve hyperbaton.

I can easily do queries that find split constituents vs. constituents that wholly precede the verb, but I don't yet have an explanation that fits all the data I've seen. I'd love to hear one.

It would really be helpful to query information structure as well. All the viable theories of Greek word order I've seen involve information structure.

[Jonathan Robie]

As far as I can tell, all your examples involve topics, not focus constituents. (Topical constituents are fairly often multi-word.) And none of them involve hyperbaton.

I simply searched for multi-word objects that directly precede the verb - this search explicitly excluded hyperbaton. Earlier in the thread, I showed examples that illustrate hyperbaton using a search that looked for it.

I can easily do queries that find split constituents vs. constituents that wholly precede the verb, but I don't yet have an explanation that fits all the data I've seen. I'd love to hear one.

It would really be helpful to query information structure as well. All the viable theories of Greek word order I've seen involve information structure.

It's quite possible that topic vs. focus is key here. I should be able to do queries involving information structure soonish, I'm in the process of setting up an environment to do various kinds of queries. When I can, I'll do those queries and post the results.