Discontinuous Syntax in the New Testament Part II

Previously, I surveyed the first half of of Devine and Stephen’s Discontinuous Syntax: Hyperbaton in Greek with the purpose of both summarizing the contents will also seeking to determine whether the argument and claims made regarding discontinuous phrases in Classical Greek carried over in the Koine Greek as well. The result of the summary showed that the phrasal discontinuity in Greek known traditionally as hyperbaton is represented in Koine at every point as it is in Classical Greek.

The implication of this is that on the basis of its uniformity, the phenomena of discontinuous phrases should not understood as proving that Greek has “free word order,” but that the Greek phrase structure has a consistent and representable pattern whether the phrase is a Noun Phrase, Verb Phrase, Adjective Phrase, or even a Prepositional Phrase.

This following post examines the second half of chapter one, which seeks to understand the word orderings of normal continuous phrases in order to lay a basis for understanding discontinuous phrases, which will be next week’s topic covering chapter two of the book.

Isn’t Greek fun?

1.1 Focus in Continuous Noun Phrases

In order to determine more accurately the meaning of Y­1 Hyperbaton,[1] The rest of chapter 1 examines the pragmatics and syntax of focus in continuous phrases where there is no hyperbaton. This requires determining the default word order in Noun Phrases, which is a massive task in of itself.[2] The majority of the discussion is statistical and since the evaluation of DS’s use of statistics is both beyond the focus of this paper and the author’s skill, we will simply summarize their relevant conclusions and continue to compare them with what is seen in the Koine Period. Several important observations should be noted though. For one DS recognizes, “There has been a tendency to underestimate the multivariate nature of the problem of adjective-noun order. Specifically, it is not very likely that the variability of order preferences can be reduced to the direct effect of a single parameter, whether it be semantic category, syntactic structure or pragmatic salience.”[3]

1.1.1 Adjective Usage and their Relation to the Noun

DS makes a distinction between two uses of the adjective: Descriptive versus Restrictive. For the most part, these two terms are self explanatory. Restrictive adjectives are those with limit the reference of a given noun with the implication that there is another similar noun with a different quality or trait. Descriptive adjectives merely express a property of the noun with no suggestion as to whether there was another possible reference. Restrictive adjective usage suggests a choice. In the sentence, “Jane had dinner with John’s best brother,” the adjective “younger” could be either restrictive or descriptive depending on how many brothers John had. If John has only one, it is descriptive, but if he has more than one, then the context restricts the meaning of the adjective to refer to one brother over against the other. If we were to say, “Jane had dinner with John’s best friend.” The adjective “best” necessitates a restrictive reading because it implies that John does indeed have other friends.

The claim DS makes for Classical Greek is that restrictive adjectives tend to neutrally occur in the postnominal position while descriptive adjectives tend to occur in the prenominal position. They confirm their analysis of the Classical Greek data on the basis of similar phenomena in other languages that allow both prenominal and postnominal adjectives, such as French, Italian, and Modern Greek where we find the same structures.

To test whether these conclusions were viable for the New Testament, we did two searches across the New Testament for modifiers from a single semantic domain from Louw and Nida, one for prenominal modifiers and one for postnominal modifiers. The goal was to chose a semantic domain that would lend itself to a restrictive usage rather than descriptive. To that end, Domain 11, “Groups and Classes of Persons and Members of Such Groups and Classes,” was chosen.[4]

The results were rather astounding.[5] A total of 115 instances of modifiers from Domain 11 were found and only 9 of these appear in a prenominal position. Of these, one of them were descriptive rather than restrictive. There were also four that while they had possible meanings in Domain 11, the particular occurrences did not have them functioning in Domain 11.[6] The one descriptive instance is seen in example (1) below.

(1) ἐπιλαβόμενοί τε αὐτοῦ ἐπὶ τὸν Ἄρειον Πάγον ἤγαγον
And they brought him to the Areopagus [lit. Ares’ Hill] Acts 17.19

This sort of modifier usage with Πάγον (hill) would normally be restrictive, but in this case, usage had become so stereo typed that the modifier is only descriptive, the proper name of a particular location.

The other eight instances are restrictive in nature. A few representative examples are below.

(2) τινὲς δὲ καὶ τῶν Ἐπικουρείων καὶ Στοϊκῶν φιλοσόφων συνέβαλλον αὐτῷ
Some of the Epicurean and Stoic Philosophers engaged in discussion with him. Acts 17:18.

(3) ψυχὴν δικαίαν ἀνόμοις ἔργοις ἐβασάνιζεν
He was tormented in his righteous soul by their lawless deeds. 2 Pet 2:8

(4) ἀλλʼ ἡ τελεία ἀγάπη ἔξω βάλλει τὸν φόβον
But perfect love casts out fear. 1 John 4:18

It is examples like this and their relation to hyperbaton that is the focus of the rest of the chapter for DS. If restrictive adjectives by default appear postnominally, why is it that they are placed before the noun here?

The answer from DS is that it’s a question of pragmatic focus. When there is no focus on the adjective or when there is only a broad focus on the entire phrase (“A BLACK CAT walked into the room;” as opposed to a brown dog, with focus on the entire NP), the restrictive adjective will appear after the noun. When there is a narrow focus on the adjective (A BLACK cat walked into the room;” as opposed to a white cat, with focus placed narrowly in the adjective), the restrictive adjective will precede the noun.[7]

But how does this fit with our previously examples; repeated below as example 5-7?

(5) τινὲς δὲ καὶ τῶν Ἐπικουρείων καὶ Στοϊκῶν φιλοσόφων συνέβαλλον αὐτῷ
Some of the Epicurean and Stoic Philosophers engaged in discussion with him. Acts 17:18.

(6) ψυχὴν δικαίαν ἀνόμοις ἔργοις ἐβασάνιζεν
He was tormented in his righteous soul by their lawless deeds. 2 Pet 2:8

(7) ἀλλʼ ἡ τελεία ἀγάπη ἔξω βάλλει τὸν φόβον
But perfect love casts out fear. 1 John 4:18

In example (5), it is clear that we have a restrictive, narrow focus on the words “Epicurean and Stoic”. Luke is intentionally highlighting these specific philosophical schools. This understanding is confirmed in commentaries. “The mention of these schools is not incidental. Paul would take up some of their thought in his Areopagus speech, particularly that of the Stoics, and thoroughly redirect it in line with the Creator God of the Old Testament.”[8] Likewise in example (6), ἀνόμοις explicitly has a narrow focus, contrasting it with the adjective δικαίαν. This same explicitness is seen in (7). The entire verse is says,

(8) φόβος οὐκ ἔστιν ἐν τῇ ἀγάπ ἀλλʼ ἡ τελεία ἀγάπη ἔξω βάλλει τὸν φόβον
There is no fear in love, but perfect love casts out fear.

The first half makes a claim about love in general, while the second puts the focus specifically on God’s love, which is perfect or complete. B. R. Westcott, intuitively recognized the distinctiveness of the construction in his commentary on John’s Epistles, “The arrangement ἡ τελ. ἀγ., which is common, for example, in 2 Pet., is unique in the Epistle…. It expresses a shade of meaning, as distinct from ἡ ἀγ. ἡ τελ., which is evidently appropriate here.”[9]

There is one other similar construction in the New Testament that can be explained by means of the thesis formulated by DS, shown in example (18)

(9) τὴν τοῦ θεοῦ δικαιοσύνην
“Righteousness of God” Rom 10:3

This particular construction is not discussed specifically by DS, but deserves discussion here, since its rather frequent in Koine Greek.[10] The genitive modifier Noun Phrases normally occur postnominally. The is a rather undisputable fact and easily documented. But what is interesting, in light of DS’s description of postnominal adjectives as typically restrictive rather than descriptive, Stanley Porter’s discussion of the genitive case in his grammar. “[T]he essential semantic feature of the genitive case is restriction.”[11] On the assumption that this claim is correct, the logical result is that in the numerous places where a genitive NP occurs between an article and its noun, the genitive NP has narrow focus. And example (19) shows this to be explicitly true with Rom 10:3 quoted in full.

(10) ἀγνοοῦντες γὰρ τὴν τοῦ θεοῦ δικαιοσύνην καὶ τὴν ἰδίαν [δικαιοσύνην] ζητοῦντες στῆσαι, τῇ δικαιοσύνῃ τοῦ θεοῦ οὐχ ὑπετάγησαν
Since they did not know God’s righteousness and they sought to establish their own righteousness, they did not submit to God’s righteousness. Rom 10:3.

Whether the variant in brackets is accepted as original or not, there is an explicit contrast between righteousness that is God’s and righteous that is not God’s.[12] Thus, Paul places the genitive NP, τοῦ θεοῦ, in the narrow focused position.

1.1.2 Implications for understanding Hyperbaton

Essentially, if the understanding of focus in continuous noun phrases is combined with the understanding of focus in Y1 Hyperbaton, DS argues, “it becomes obvious that we are not dealing with two separate phenomena but with a single syntactic process.”[13] The narrow focus position for restrictive adjectives is simply a lower position in a hierarchy of specifier-type positions. New Testament examples of this are shown below in examples 11-13.

(11) τί γὰρ ὠφεληθήσεται ἄνθρωπος ἐὰν τὸν κόσμον ὅλον κερδήσῃ τὴν δὲ ψυχὴν αὐτοῦ ζημιωθῇ (N A X)
What is a manʼs profit if he gains the whole world, but forfeits his soul? Matt 16:26

(12) ἐπιστάτα διʼ ὅλης νυκτὸς κοπιάσαντες οὐδὲν (A N X)
Master, we toilded all night and caught nothing. Luke 5:5

(13) ὅλη συγχύννεται Ἰερουσαλήμ (A X N)
“All Jerusalem was in confusion” Acts 21:31

Each of these examples contains the restrictive adjective. The first has the adjective in its unmarked or default restrictive position (N A). The second has the adjective in its more marked narrow focused restrictive position (A N). Finally, the third clause has the adjective in the most marked position, Y1 hyperbaton.

1.2 Conclusions for Chapter One

The fact that for every example given by DS for describing this phenomena, a similar example can be given from Koine texts makes it clear that more likely than not, we are dealing with the exact same pragmatic/syntactic phenomena. In the examples surveyed parallel examples show that the same sort of restrictive focus can be seen in the New Testament as it can in Classical texts several hundred years previously. When we move to examine chapter two, we’ll examine the phenomenon’s significance and meaning.


[1] Y­1 Hyperbaton being the word order: Modifier, XP, Noun; Y1, XP, Y2.

[2] The task is too massive, in fact, for most of their analysis to be discussed here. What follows only includes the most relevant details specifically for studying hyperbaton. What is left out is probably one of the most impressive and concise discussions of Greek word order in the Noun Phrase. Much of their data comes from Helma Dik’s Word Order in Ancient Greek: A Pragmatic Account of Word Order Variation in Herodotus (ASCP 5; J C Leiden, 1995) and her other articles.

[3] DS, 21.

[4] Johannes P. Louw and Eugene Albert Nida, Greek-English Lexicon of the New Testament: Based on Semantic Domains (electronic ed. of the 2nd edition; New York: United Bible societies, 1996, c1989), 1:119-135. The reasoning behind this choice was the claim, “[T]he neutral position for restrictive adjectives is postnominal. This is particularly so for intersective adjectives; these are prototypically simple properties denoting an extensional class determined for the most part independently of the noun they modify; for instance, adjectives of … nationality” (DS, 20). This claim seems intuitionally accurate. If one is description a particular class of something, more likely than not there will be another class that is then implicitly rejected.

[5] It must be acknowledge that statistics must be used with caution, particularly by those who have not been trained in statistical analysis, such as myself.

[6] Specifically, they were all instances of the word ἀλλότριος, with typically means “other” or “another,” but can carry the Domain 11 meaning, “stranger” in some contexts. The Lexham Greek-English Interlinear tags only three instances of this word as having the meaning “stranger” in John 10:5 (2); Heb 11:9. In each case, the word is used as a substantive rather than a modifier.

[7] This fits with what I have written previously about Adjective word orderings, HERE. But it also does not mean that all prenominal adjectives are restrictive with a narrow focus. As noted previously, descriptive adjectives will appear pronominally as well. This fact fits well with the comment by Dr. Rich Rhodes HERE.

[8]John B. Polhill, Acts (NAC 26; Nashville: Broadman & Holman, 1992), 367.

[9] Brooke Foss Westcott, The Epistles of St. John: The Greek Text With Notes and Essays (4th ed.; London: Macmillan, 1902), 159.

[10] Occurring 1322 times in the New Testament, LXX, Apostolic Fathers, Philo, and Josephus.

[11]Stanley E. Porter, Idioms of the Greek New Testament (Sheffield: JSOT, 1999), 92 (italics his).

[12] The case actually grows stronger if the bracketed reading is accepted because then we have two focused position modifiers: the τοῦ θεοῦ and then ἰδίαν.

[13] DS, 31.

13 thoughts on “Discontinuous Syntax in the New Testament Part II

  1. Since you asked for comments. I followed this reasonably well. It’s more technical than I’m used to reading, but that was okay. I think the significance of it all remains unclear, but perhaps that is a reflection of just being chapter one.

  2. I also found this comprehensible for a linguist or specialist which I believe is your target audience. I think the book is right to look for default vs. marked structures. However, my experience with Levinsohn is that we are intuitively almost always wrong about why the marked structure occurs. In the case of “the of God righteousness” there is first the concept of Natural Information Flow which states that established information comes first.

    That section in Rom 10:3-4 has righteousness plus a genitive three times and one is the default. I’m too out of practice to try to analyze why the genitive precedes righteousness twice in v.3 but follows it in v.4.

    If I weren’t so lazy I’d try to track down Levinsohn’s anlaysis but I think I’d rather take a Sunday afternoon nap.

  3. I’m not willing to give in on my proposal just yet. The typological evidence that DS provided is pretty strong in my book. But I’ll keep thinking about it.

    If you ever do come across Levinsohn’s analysis and have a digital camera handy, I’d be interested in it…

    I also think that perhaps the claims by DS are not necessarily in contradiction with the Natural Flow of Information Principle, because they argue that descriptive adjectives & modifiers – traits that have already been mentioned in a discourse or are simply well known in general have a pre-nominal position as default. As far as I can tell this fits when you’ve described perfectly.

    DS’s second claim is that there are times when restrictive adjectives (new clarifying information) are moved forward for emphasis or focus – particularly when there is a clear contrast between two referents. That’s what I see going on here.

  4. Mike and David,

    I do not think that Levinsohn addresses focus in the NP, just in the clause. I contacted him privately regarding the matter, and got some initial prnciples that he had been working on. I think that the direction that DS are heading is helpful. Most of what they seem to claim is consistent with NIF, so thanks for raising that David. I was thinking along the same lines.

    In the database underlying the LDGNT, I began marking NP focus, but was not comfortable enough with the analysis to publish the results. Perhaps reading DS would supply the needed framework.

  5. Steve, I’d be interested if your thoughts on this particular chapter if you read DS, particularly since you would be able to evaluate their statistical claims better than I would.

    The next chapter is going to look more closely at Adj XP Noun hyperbaton with an eye toward determining its significance more precisely than simply “focus,” which by itself is a bit of an umbrella term.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s