Changes in complement structure from Classical to Byzantine Greek

The post below was originally published in 2017. It was among a large number of posts that were lost for several years, and I recently recovered it. I have added a brief comment below after the republished post.


Journal of Greek LinguisticsIn the first issue of the Journal of Greek Linguistics of 2017, Klaas Bentein examined changes in the way verbal complements were formed between the Classical and Byzantine periods. Here’s what the abstract of his paper says:

While Classical Greek has a particularly rich complementation system, in later times there is a tendency towards the use of finite complementation. In this context, Cristofaro (1996) has claimed that the Classical opposition whereby the accusative and infinitive is used for non-factive complements, and ὅτι with the indicative and the accusative and participle for factive ones, is disappearing, ὅτι being used as a ‘generic’ complementiser. In this article, I investigate to what extent Cristofaro’s (1996) claim of the pragmatic neutralisation of complementation patterns can be upheld, and whether it could be claimed that a new pragmatic opposition, in terms of ‘register’, is being established. For this purpose, I turn towards documentary papyri, a corpus which is particularly fruitful for socio-historical investigations.

You can read this article here here. I have added it to the Comprehensive Bibliography of Hellenistic Greek Linguistics.


The title of Bentein’s article is “Finite vs. non-finite complementation in Post-classical and Early Byzantine Greek: Towards a pragmatic restructuring of the complementation system?”

What he calls “complement structure” is referred to as “argument structure” in other theoretical frameworks and “case frames” in still others. These two alternative models both place a strong emphasis on the relations of these complements to the verb with which they appear.

Dissertation Idea? Case and Infinitival Clauses

I’ve been reading Christina Sevdali’s 2007 dissertation, Infinitival Clauses in Ancient Greek: Overt and Null Subjects, the Role of Case and Focus. She deals primarily with Classical Greek, but delves into some Modern Greek data as well, but she does not deal with the hellenistic period. Her work, though, does raise some questions that should be answered for the hellenistic literature.

Sevdali concludes that both agreement and focus play a role in Case marking in Ancient Greek. Here is part of a long paragraph from pages 209—210 in her last chapter that I think could suggest a dissertation idea for someone working specifically on Hellenistic Greek:

There are various languages [in which] Case can be related to discourse phenomena: Blake, 2001 for example reports Australian languages Nyigina and Gooniyandi where this is true. These languages do not show Case concord within a noun phrase, where Case and number and person are marked on every constituent, i.e. the determiner, the noun, the adjective etc., but Case mark only one constituent, the final one, or the head etc. In some cases, they mark the one that is focalised, essentially using Case as a discourse marker. Miyagawa, 2005 argued that languages can either be agreement prominent (like most Indo-European ones) or focus-prominent (like Japanese), implying that Agreement and Focus are the two sides of the same coin. Assuming that Case exists in both types of languages, it is not unreasonable to assume that it can be linked to Agreement and Focus respectively. On top of that nothing prevents us from arguing that there also exist mixed language types. We want to suggest that AG is a mixed language, being agreement prominent in finite clauses, where Case is linked to agreement, and being focus-prominent in non-finite clauses, where Case is linked to focus as we showed.

Okay… Here are dissertation ideas for the hellenistic period: Can recognizing two different ways in which Case may be assigned (Agreement vs Focus) lead to a clearer understanding of morphological Case assignment in Hellenistic Greek? Under what specific circumstances might morphological Case be controlled by Focus in Hellenistic Greek? Does Focus play any role in the Case assignment of optional arguments of a verb? Does it control the Case assignment of any DPs governed by a preposition, especially prepositions whose object DP is not always assigned the same Case. Prepositions played a larger role in the hellenistic koine than they did in the classical period. How does this affect the agreement/focus split if at all?

Any takers? I’d love to see a dissertation addressing any of these issues.

Argument Structure of ἀγαπάω

Simon Wong’s A Classification of Semanti Case-Relations in the Pauline Epistles lists the Case Frame (Argument Structure) of ἀγαπάω as [Event: EXPERIENCER, COMPLEMENT/PATIENT]. I think this argument structure is quite appropriate for the English word “love,” but I’m not sure it really fits ἀγαπάω.

My disagreement is with the designation of first argument as EXPERIENCER. In English we think of love as an emotion, in which case it is quite appropriate to think of the first agument (the subject of an active verb) as EXPERIENCER rather than AGENT. Love is something we experience more than do.

In Hellenistic Greek, though, ἀγαπάω represents a way of acting more than an emotion. Jesus commands his disciples ἀγαπᾶτε τοὺς ἐχθροὺς ὑμῶν (Matthew 5:44; Luke 6:27 and 35). He is not commanding them to feel warm and fuzzy toward their enemies, but to treat their enemies with good will.

Does it even make sense to command an emotion? If I tell you, “Be angry!” will you be able to simply decide to do so? In Hellenistic Greek, ἀγαπάω represents something that can be commanded. It represents something that a person can decide to do.

I propose the following revision to Wong’s case frame (argument structure) for ἀγαπάω: [Event: AGENT, COMPLEMENT/PATIENT]. The verb implies an actor/AGENT (the person who acts with good will) and a PATIENT (the person who is treated with good will).

Feel free to disagree. Please offer examples that you think demonstrate whether the first argument (the subject of ἀγαπάω when it is active voice) represents a person who experiences the emotion we call love or a person who acts in a way characterized by good will. Does ἀγαπάω function like the English word “love,” or do you also think it is different?

Greek Argument Structure

I agree with Mike Aubrey that the category Verb Phrase is not particularly helpful at this point in the discussion of Hellenistic Greek syntax. I would like to propose that we talk instead about the “Argument Structure” of Hellenistic Greek verbs.

Each verb requires, or clearly implies certain elements, such as a subject, and frequently one or more objects as well. Where these elements are essential to the meaning of the verb, we can say they are part of the verb’s “Argument Structure.” Modifiers that are optional, in the sense that they are not demanded by the meaning of the verb, we can say are not part of the verb’s argument structure.

Let’s take the verb δίδωμι as an example. In Matthew 4:9 we find

ταῦτά σοι πάντα δώσω
All of these I give to you

Here the verb is accompanied by three arguments: ταῦτα (these), σοι (to you), and -σω (I). One of these arguments (the subject) is attached to the verb itself and need not be expressed separately unless the context demands it. The other two we can call “complements.”

In certain discourse contexts, one or more of the complements may be left unexpressed. In Matthew 5:42, for example, we find

τῷ αἰτοῦντί σε δός
Give to the one who asks you
Give to the one who begs from you

Here the Recipient is expressed  explicitly: τῷ αἰτοῦντί σε (the one who asks you, the one who begs from you), but the Patient (that is, the thing that is given) is not expressed explicitly. Crucially, the cultural and discourse contexts make it clear that sustenance, in the form of food or money, is what is expected. The meaning of δίδωμι itself asserts a Patient role (the object given), and when the context clearly implies what must fill that role, it may be left implicit rather than directly expressed.

It is my contention that a reference grammar for Hellenistic Greek, if it is to serve the interests of both language learners and exegetes, needs to include this kind of information. Optimally,  argument structure information should be included in the lexical entry for every verb. The grammar would simply need to explain argument structure, and refer to an accompanying lexicon for details of specific verbs.

Some strides have been made toward this goal in recent research. Simon Wong provided a great deal of relevant data in his A Classification of Semantic Case-Relations in the Pauline Epistles (1999). What he calls semantic case-relations I would call arguments to avoid confusion with morphological case (nominative, accusative, etc.), but the data he provides could be very useful.

If you know of other research on this topic, please include it in your comments.

What would you like to see about the argument structure of verbs in a reference grammar? In a lexicon?

Greek Verb Phrase?

I would like to thank Michael Aubrey for his comments on the the lack of usefulness of the category VP (Verb Phrase) for describing Ancient Greek. In particular, he challenged some comments that I made in Levels of Constituent Structure for New Testament Greek (1995).

This is, of course, the way to advance the field. As we each examine the claims of our colleagues and submit them to scrutiny, we move the discussion forward.

The comments I made about the Greek VP in 1995 were part of a larger argument for the existence of phrase-level categories in general. While I am still firmly committed to the usefulness of the syntactic category Phrase in general, I have never been particularly committed to the usefulness of VP in particular for addressing the phrase structure of Ancient Greek.

In the years since 1995 I have come increasingly to view the syntax of Ancient Greek as determined by the argument structure of verbs—verbs and the phrases demanded by their lexico-semantic properties. This view does not necessarily require the category Verb Phrase, though postulating the existence of such phrases may eventually prove useful.

As for Aubrey’s objection that the subject appearing between the verb and one of its subsequent arguments (specifically a PP in the examples he cites), the objection works only if you discount the possibility of verb movement. He is probably right, but it’s not as obvious as it might seem.

Thank you, Michael, for your thoughts and research on this issue. I look forward to reading more.