Editor’s note: I was asked to comment on Greek word order in a Facebook thread and the following discussion grew out of that. I would like to do a Brief Guide series for word order along the lines of my aspect series: A brief guide to aspect in Greek. Filed under: Goals I’m not sure I can get to any time soon.

Problems with common word order approaches

Many discussions of Postclassical Greek word order in NT studies continue to be defined by the ordering of the verb and its grammatical relations (subject, object, etc.). There are insights to be had in that context, but overall, I grow more convinced that using grammatical relations as an organizing principle for word order should be viewed as a distraction from more interesting insights into language.

Greek word order, as a non-configurational language, is driven by information structure in more dramatic ways than configurational languages like English, where constituent orders like SVO or VSO are more regular and consistent. Because of the non-configurational nature of Greek, looking for the proper order of Verb, Subject, and Object is already starting off on the wrong foot. Greek word order primarily involves the ordering of (1) old information and new and (2) assumed information and asserted information. These categories tend to be referred to with the labels “topic” and “focus”. All of this plays a large role in word order regardless of the grammatical category in question (subject, object, indirect object). Additionally, there is an importance/salience that speakers attribute to this information and how it is contrasted with other information.

What I think needs to be understood first, in the most simple terms, is that in Greek, things are placed in front of the verb you want to put prosodic sentence stress on them. Let’s first consider some English examples of what I mean before we look at Greek directly. In English someone might say:

1. John wants the APPLE not the orange.

The speaker asserts something about the apple and not about the orange, so they place extra prosodic stress on ‘apple’. This is the sentence accent.

In a language like Greek, the way the information is packed would require us to place ‘apple’ first. This would place the stress at the front and move the constituent “apple” forward. English also allows this kind of construction as more salient word order.

2. The APPLE John wants not the orange.

At a basic level, Greek sentence stress always prefers to be on the verb or in front of the verb. We can see this by how regularly words that cannot take accents at all (clitics) attach to the verb or an element in front of the verb. Pronominal clitics default to attaching to the verb. In example (3) that attachment is marked by an ‘=’.

3. a.     ἐὰν θέλῃς δύνασαί=με καθαρίσαι
           If you are willing, you are able to heal me (Luke 5:12).
   b.     ἀπέστειλέν=με ὁ ζῶν πατὴρ
           The living father sent me (John 6:57)

These accent-less pronouns are attracted to the sentence accent of the clause. We know this is the case because while their default position is immediately after the verb, if a speaker wants to emphasize something else with the sentence accent, the clitic pronoun moves there instead—also in front of the verb. In example (4), we see that Greek question words, like τίς, ‘who’ receive sentence accent.

4. Τίς=με κατέστησεν κριτὴν ἢ μεριστὴν ἐφʼ ὑμᾶς;
   Who made me just or arbitrator over you? (Luke 12:14).

Similarly in Luke 10:35 (example 5), the clitic pronoun moves in front of the verb to attach to the word most salient to Martha’s assertion μόνην, ‘alone’.

5. ἡ ἀδελφή μου μόνην=με κατέλειπεν διακονεῖν
   My sister has left me alone to make preparations (Luke 10:35)

In example (6) there are two clitic pronouns pulled forward by the sentence accents on οὐδείς, ‘no one’ and on πάντες, ‘everyone’.

6. Ἐν τῇ πρώτῃ μου ἀπολογίᾳ οὐδείς=μοι παρεγένετο, ἀλλὰ πάντες=με ἐγκατέλιπον
   At my first defense, no one came to me, rather, everyone deserted me (2 Tim 4:16).

Paul uses sentence position and the sentence accent together to stress his assertion about his companions during his trial.

Understanding and thinking about these kinds of examples with clitic pronouns is important for two reasons. First, they illustrate clearly the default position of the sentence accent (the verb) along with its position in front of the verb when another constituent is more salient. Second, they illustrate a fundamental problem with traditional approaches to word order. Note that in all these examples the clitic pronouns are objects/indirect objects in the accusative or dative cases. That means they are also grammatical relations. Trying to ascertain a word order, VSO, SVO, VOS, and so forth one when often one of those elements is phonologically bound to another word like these clitics makes the entire endeavor a fool’s errand. Those who prioritize statistical approaches to constituent order will find their statistics skewed by clitics whose position in the clause is determined by sentence prosody, independent of syntax.

A similar challenge arises with discontinuous noun phrases like δευτέραν ὑμῖν γράφω ἐπιστολήν (2 Peter 3:1). Again, this is sentence stress in front of the verb. The SECOND letter I have written to you—note also the pronoun ὑμῖν being pulled forward because of that sentence accent/stress. This pronoun is a clitic, too. Its word accent was not pronounced in speech when it attaches to the sentence accent. We can describe the motivation of this kind of discontinuity by analogy with English.

7. John wants the GREEN apple, not the red one.

The sentence introduces a contrast of assertions. The speaker wants to assert, not merely that John wants an apple, but that he wants a specific kind of apple. English sentence stress allows high to place the focus on attention on the adjective that specifies that type: the GREEN apple. In example (8), if English were a language with a Greek-like word order, this would be realized more like:

8. The GREEN John wants apple, not the red one.

Here ‘the green’, adjective portion of the full noun phrase ‘the green apple’ is positioned in front the verb and the contrasted as the asserted part of the sentence. It receive the primary sentence stress. Finally note again, that because a phrase like this is a grammatical relation (object), discontinuous constituents also disrupt constituent order analyses: δευτέραν … ἐπιστολήν, ‘second letter’ is both before and after the verb at the same time.

Sentence accents in Ancient Greek

Here are a variety of places where sentence accents appear in front the verb with some brief discussion. In Acts 17:18 (example 9), Luke gives sentence stress to the subject to highlight the types of people engaging with Paul.

9. τινὲς δὲ καὶ τῶν Ἐπικουρείων καὶ Στοϊκῶν φιλοσόφων συνέβαλλον αὐτῷ.
   Even some of the EPICURIAN and STOIC philosophers engaged in discussion with him (Acts 17:18).

I suggest we place the stress on Ἐπικουρείων and Στοϊκῶν because we already know from previous context that he would engage with anyone who happened to be there (τοὺς παρατυγχάνοντας) and now this clause specifies who was included in that group: τῶν Ἐπικουρείων καὶ Στοϊκῶν. The use of καὶ, ‘even’, is also a good signal that this is new, assert information.

Similarly in Luke 5:5 (example 10) the disciples place their sentence stress on διʼ ὅλης νυκτὸς and thus put it in front of the verb:

10. Ἐπιστάτα, διʼ ὅλης νυκτὸς κοπιάσαντες οὐδὲν ἐλάβομεν
    Master, toiling ALL NIGHT, we caught nothing (Luke 5:5).

The disciples want to communicate and emphasize how much they toiled in the face of failure to catch anything. In example (11), the speaker contrasts what a man might gain versus what he forfeits, both fronted before the verb and stressed.

11. τί γὰρ ὠφεληθήσεται ἄνθρωπος ἐὰν τὸν κόσμον ὅλον κερδήσῃ τὴν δὲ ψυχὴν αὐτοῦ ζημιωθῇ
    What is a manʼs profit if he gains the WHOLE WORLD, but forfeits his LIFE? (Matt 16:26).

Let’s look at some clear examples of this appearing with discontinuities, where part of a noun phrase is fronted before the verb. In example (12) the speaker strongly asserts, with stress, how much of Jerusalem was in confusion.

12. ὅλη συγχύννεται Ἰερουσαλήμ
    ALL Jerusalem was in confusion (Acts 21:31).
13. ἴδετε πηλίκοις ὑμῖν γράμμασιν ἔγραψα τῇ ἐμῇ χειρί
    See what LARGE letters I write to you in my own hand (Gal 6:11).
14. κύριε ἡ μνᾶ σου δέκα προσηργάσατο μνᾶς
    Lord your mina has made TEN minas (Luke 19:16).

Similarly, in example (13), Paul asserts with sentence stress the size of the letters, not merely their existence, because the size, specifically, is salient for their being in his own hand. And thus, he fronts the adjective πηλίκοις before the indirect object. The verb still comes first for other reasons related to the salience of imperatives. In example (14), the servants assert with sentence stress the quantity of mina—TEN!—and the numeral appears in front of the verb. The subject, ἡ μνᾶ σου, functions as the topic/given information that sets the stage for the strong assertion.

Finally, let us think, again about 2 Peter 3:1 briefly mentioned above.

15. ταύτην ἤδη ἀγαπητοί δευτέραν ὑμῖν γράφω ἐπιστολήν ἐν αἷς διεγείρω ὑμῶν ἐν ὑπομνήσει τὴν εἰλικρινῆ διάνοιαν
    This is now, beloved, the SECOND letter I am writing to you; in this one, as in the first, I am trying to arouse your sincere intention by reminding you (2 Peter 3:1).

The discontinuous phrase is δευτέραν … ἐπιστολήν. Peter explicitly states the reason for the sentence stress of δευτέραν in front of the verb. The author is trying to arouse his audiences’ sincere intention.

There is much more to say about all of this, but that would require more time than I have right now. And there are numerous great books and articles that engage with these concepts. Beyond this, there are basic concepts like topic and focus that linguists have developed for discussing information structure/flow that also deserve discussion. For now, here is a brief bibliography of important works.

Bibliography (with links and a few notes)

This bibliography is not comprehensive, but it does highlights what I consider a few excellent resources on Greek word order. All of this work is really good and they are one of the reasons that I have not pursued publishing on this topic myself—I don’t have anything particularly new to say because so many others have already done such great research. But eventually, I probably still will simply because we need a better synthesis of all these somewhat disparate works. All of these cover a wider collection of topics than the content of the essay above. Dik (1995, 2007), Levinsohn (2000) and Runge (2010) represents analyses of word order with clear discussions of the concepts of Topic and Focus as organizing principles. Bailey (2009) does similarly with regard to one specific type of sentence (a published version is forthcoming) and particularly highlights how presentational clauses have specific realizations in Greek. Goldstein (2010, 2015), Dik (2003), and Janse (1993, 1994, 1996) represent important work on clitics, prosody, and word order. Devine & Stephens (2000) gives a fairly comprehensive analysis of discontinous noun phrases, but they often seem hamstrung by their theoretical framework and their focus on finding a purely syntactic explanation leads them to miss better generalizations/motivations. Finally, Scheppers (2011) is insightful, but also a challenging read and idiosyncratic (see Goldstein’s review in Bryn Mawr Classical Review). Unfortunately, it also seems to be out of print. I would finally emphasize that these concepts, particularly ones about enclitics are not remotely new. Wackernagel (2020) is the first English translation of Jacob Wackernagel’s important publication on second position phenomena and everything builds on his insights. Similarly Edwin Abbott evinces an surprisingly nuanced awareness of enclitics pronouns behavior in his (1906, 601-8) Johannine Grammar.

Abbott, Edwin A. 1906. Johannine Grammar. London: Adam and Charles Black.

Bailey, Nicholas (2009) “Thetic Constructions in Koine Greek: with special attention to clauses with εἰμί ‘be’, γίνομαι ‘occur’, ἔρχομαι ‘come’, ἰδού/ἴδε ‘behold’, and complement clauses of ὁράω ‘see’.” PhD Dissertation. Vrije Universiteit Amsterdam.

Devine, A.M. & Laurence D. Stephens. 2000. Discontinous syntax: Hyperbaton in Greek. Oxford: Oxford University Press.

Dik, Helma. 1997. Word order in Ancient Greek: A pragmatic account of word order variation in Herodotus. Amsterdam: Brill Academic.

Dik, Helma 2003. “On Unemphatic ‘Emphatic’ Pronouns in Greek: Nominative Pronouns in Plato and Sophocles.” Mnemosyne 56: 535-550.

Dik, Helma. 2007. Word Order in Greek Tragic Dialogue. Oxford: Oxford University Press.

Goldstein, David. 2010. “Wackernagel’s Law in fifth-century Greek.” PhD Dissertation. University of California, Berkeley.

Goldstein, David. 2015. Classical Greek Syntax: Wackernagel’s Law in Herodotus. Amsterdam: Brill Academic.

Janse, Mark. 1993. “The Prosodic Basis of Wackernagel’s Law.” Pages 19-22 in Les Langues Menacées. Actes du Xvu Congrès International des Linguistics, Québec, Université Lavel, 9-14 Août 1992. Ed. André Crochetière, Jean-Claude Boulanger and Conrad Ouellon. Sainte-Foy: Presses de i’Université Lavel.

Janse, Mark. 1994. “Clitics and Word Order since Wackernagel: One hundred years of research into clitics and related phenomena.” Orbis, 37(0), 389–410. doi:10.2143/orb.37.0.2012786

Janse, Mark. 1996. “Phonological aspects of clisis in Ancient and Modern Greek.” Glotta, 73: 155-167.

Levinsohn, Stephen. 2000. Discourse Features of New Testament Greek: A Coursebook on the Information Structure of New Testament Greek. 2nd. Edition. Dallas, TX: SIL International.

Runge, Steven E. 2010. Discourse grammar of the Greek New Testament: A practical introduction for teaching and exegesis. Peabody, MA: Hendrickson.

Scheppers, Frank 2011. The Colon Hypothesis: Word Order, Discourse Segmentation and Discourse Coherence in Ancient Greek. Brussels: VUB Press.

Wackernagel, Jacob. 2020. On a law of Indo-European word order: Über ein Gesetz der indogermanischen Wortstellung. Berlin: Language Science Press.