Greek Grammars on the Articulation of Prepositional Objects

[Stephen Carlson]

Next up is Stanley E. Porter, Idioms of the Greek New Testament, BLG 2 (Sheffield, JSOT Press, 1992). It is an intermediate grammar, and like many intermediate grammars, we should not expect much. (Wallace is an exception because of his intense interest in the article.) As a result, I'm not sure how many intermediate grammars I'll look, since Porter's is exemplary of the genre in terms of its detail. But because Porter's doctrine of the article differs from other grammarians, let's take a closer look:

2.8.4 Omission of article in prepositional phrases. The article of a substantive in a prepositional phrase is often omitted. Moulton describes this use as entirely normal, stating further, "There is nothing indefinite about the anarthrous non there.'²

Mk 2.1: ἐν οἴκῳ ἐστίν (he is at home), literally 'in house'.

1 Pet. 1.5: τοὺς ἐν δυνάμει θεοῦ φρουρουμένους διὰ πίστεως εἰς σωτηρίαν ἑτοίμην ἀποκαλυφθῆναι ἐν καιρῷ ἐστάχῳ (those being guarded for in [the] power of God through [the] faith for [the] prepared salvation to be revealed in [the] last time), where English articles are demanded for sense.

2. Moulton, Prolegomena, p. 82.

There is a bit of a "move along, nothing to see here" quality to the discussion. We're told that the omission of the article with a substantive in a prepositional phrase is "entirely normal," but this section and the discussion employs the word "omission" which suggests to me that its presence is somehow expected. Moreover, the discussion is not entirely clear that the problem is the "omission" is with those substantives that would ordinarily take the article in other contexts.

Now, others grammarians would identify those substantives as "definite" (or "identifiable" in more recent literature), but Porter rejects definiteness as a way of understanding the article, largely on the grounds that "Use of the article in Greek is not like use of the definite article in English, not least because Greek does not have the same choice of forms" (p. 103). In particular, (NT) Greek lacks an indefinite article and so the systems are too different to be comparable. (Note how opposite Porter's approach is to von Siebenthal's.)

What Porter does is as follows:

Code: Select all

Substantive                  Use 1               Use 2
===========================  ==================  ===============
Articular (with article)     (a) particular      (c) categorical

Anarthrous (without article) (b) non-particular  (d) individual
                                 (qualitative)

When the article is used, the substantive may refer to a particular item, or it may represent a category of items. When the article is not used, the substantive may refer to the non-particular or qualitative character of an item, or it may refer to an individual item. One immediately notices that the presence or absence of the article may affect the sense of the substantive in two related ways. Uses (a) and (d) are similar in meaning, as are uses (b) and (c) (translations may well reflect this similarity). Matters of particularity and individuality are established not on the basis of whether the article is present, but on the basis of the wider context.

This is not a helpful explication of the article. As far as I can tell, "particularity" and "individuality" basically mean the same thing, and their definitions are synonymous: "The article may particularize a substantive." (p.104) and "Individual items may be specified without use of the article." What Porter does is set up a distinction between particularity=specificity and non-particularity/qualitativeness=categorical as important and then states that the Greek article does not encode this distinction!

This formulation illustrates that in Greek the presence or absence of the article does not determine whether the substantive is particular or non-particular, categorical, or individual.

Now, my expectation from a grammar that I would prefer to know what a linguistic form means, rather than what it does not mean. At any rate, here is his treatment of the "individual" use that we see in prepositional phrases:

1.2.2 'Individual' use without the article. Individual items may be specified without use of the article.

Jn 4.27:μετὰ γυναικὸς ἐλάλει (he was speaking with [the] woman)

Col. 2.20: ἐν κόσμῳ (in [the] world); Jn 1.1: ἐν ἀρχῇ (in [the] beginning; 1 Jn 1.1: ἀπ' ἀρχῆς (from [the] beginning). See section 2.8.4 below.

If you look closely, all of these examples involve prepositional objects. But it is not clear whether this is the basis for erecting the category.

[Eeli Kaikkonen]

What Porter does is set up a distinction between particularity=specificity and non-particularity/qualitativeness=categorical as important and then states that the Greek article does not encode this distinction!

Porter 1992:105 wrote:
This formulation illustrates that in Greek the presence or absence of the article does not determine whether the substantive is particular or non-particular, categorical, or individual.

Now, my expectation from a grammar that I would prefer to know what a linguistic form means, rather than what it does not mean.

I wonder if he partly falls into his own trap, namely considering what an English speaker could assume and then denying it. But because I'm not so good at the English articles I can't tell if this is the case.

[Stirling Bartholomew]

The value of the Patristic evidence is providing a background of data to smoke out fallacies in the NT textbooks, intermediate grammars and discourse studies of the Greek article. A lot of ideas fall on the cutting floor after massive exposure to raw data. Some of my pet notions about the greek article from reading monographs published in the last 20 years are no longer viable.

Using a large collection of Patristic Theologians as target, if you compare ἐκ νεκρῶν to ἐκ τῶν νεκρῶν you find a notable preference for ἐκ νεκρῶν 939/138. But before we run off and draw some silly conclusion about this we need to look at other prepositions. For example ἀπὸ νεκρῶν compared to ἀπὸ τῶν νεκρῶν 24/28 and μετὰ τῶν νεκρῶν to μετὰ νεκρῶν 12/8 . This suggests caution when talking about the article with νεκρῶν as the object of a preposition.

νεκρός has a special status as a cultural universal. When used in the plural it generally doesn't require referential disambiguation. This has implications which need to be taken into account when trying to figure out what the lack of the article means. Don't suspect that Wallace or Porter will be much help with this question. I also turn a deaf ear to discussions of salience marking. In my opinion, salience marking has past its expiration date and needs to be pulled from the shelf.

[Eeli Kaikkonen]

νεκρός has a special status as a cultural universal. When used in the plural it generally doesn't require referential disambiguation. This has implications which need to be taken into account when trying to figure out what the lack of the article means.

By saying this you have given a possible explanation for the lack of the article. In a way it restates what was said previously: maybe these prepositional phrases can't be lumped together. There's of course still the difference between ἐκ and ἀπὸ to be explained. So, are you sure you don't contradict yourself?

I think everyone can agree that salience isn't an explanation. But could it still have something to do with information status?

This big question is anyways "if the article marks identifiability, under what circumstances lack of the article marks non-identifiability, and why it doesn't mark it always? Why can the article be left off?"

This reminds me about the discussion about subject-affectedness and asymmetric markedness of the mediopassive. Asymmetric markedness feels natural when you accept it, and we haven't been able to explain why some lexemes are "deponent" in form and some others are not. Maybe identifiability/article behaves in the same way? The article marks it, but isn't needed, and we can't explain when it's not needed?

The article is so ubiquitous and important (well, at least for exegetes who try to find gold nuggets) that it's a wonder we don't yet know more about it. Not even in descriptive level.

[Barry Hofstetter]

νεκρός has a special status as a cultural universal. When used in the plural it generally doesn't require referential disambiguation. This has implications which need to be taken into account when trying to figure out what the lack of the article means.

The article is so ubiquitous and important (well, at least for exegetes who try to find gold nuggets) that it's a wonder we don't yet know more about it. Not even in descriptive level.

It may be because it is so ubiquitous. Latin was mentioned above, a language which does not have any article. When students first learn this, they are like "What?" How do you know when to put them in when you translate into English?" Yet in a few months they are regularly adding the articles where necessary in their English translations, because fluent English speakers just know when to do it. The same with the article in Greek. Often, it seemingly corresponds sufficiently to the definite article in English that we don't even notice it (except in so far as it might help with parsing a difficult form). When it doesn't, we easily adjust. We pick up a "feel" for the article that is sufficient to the reason we are reading the Greek.

[Stephen Carlson]

What Porter does is set up a distinction between particularity=specificity and non-particularity/qualitativeness=categorical as important and then states that the Greek article does not encode this distinction!

Porter 1992:105 wrote:
This formulation illustrates that in Greek the presence or absence of the article does not determine whether the substantive is particular or non-particular, categorical, or individual.

Now, my expectation from a grammar that I would prefer to know what a linguistic form means, rather than what it does not mean.

I wonder if he partly falls into his own trap, namely considering what an English speaker could assume and then denying it. But because I'm not so good at the English articles I can't tell if this is the case.

That might be what's going on here, but he kind of does this to every grammatical category. At any rate, I suppose it's a valid question and other linguists have studied it to the same (or similar) conclusion. For example, Stephanie Bakker's study of the Greek article concludes that the article does not mark specificity, nor does its absence mark non-specificity, but this is part of a much larger treatment where she does explicate a positive theory of the article.

[Stephen Carlson]

The value of the Patristic evidence is providing a background of data to smoke out fallacies in the NT textbooks, intermediate grammars and discourse studies of the Greek article. A lot of ideas fall on the cutting floor after massive exposure to raw data. Some of my pet notions about the greek article from reading monographs published in the last 20 years are no longer viable.

I'm with you on the general point; it's just that I don't get what your specific searches are intending to prove on the issue of the articulation of prepositional objects.

Using a large collection of Patristic Theologians as target, if you compare ἐκ νεκρῶν to ἐκ τῶν νεκρῶν you find a notable preference for ἐκ νεκρῶν 939/138. But before we run off and draw some silly conclusion about this we need to look at other prepositions. For example ἀπὸ νεκρῶν compared to ἀπὸ τῶν νεκρῶν 24/28 and μετὰ τῶν νεκρῶν to μετὰ νεκρῶν 12/8 . This suggests caution when talking about the article with νεκρῶν as the object of a preposition.

I have to admit that my eyes glaze over when I see uninterpreted, raw statistics. Language use is heavily dependent on context, and there are so many context-related variables that these statistics could be controlling for but aren't. These numbers tell me nothing except that there is variation to be explained.

νεκρός has a special status as a cultural universal. When used in the plural it generally doesn't require referential disambiguation. This has implications which need to be taken into account when trying to figure out what the lack of the article means. Don't suspect that Wallace or Porter will be much help with this question. I also turn a deaf ear to discussions of salience marking. In my opinion, salience marking has past its expiration date and needs to be pulled from the shelf.

Given your statistics of 939 ἐκ νεκρῶν to 138 ἐκ τῶν νεκρῶν, shouldn't we be trying to figure out what the presence of the article means?

[Stephen Carlson]

This big question is anyways "if the article marks identifiability, under what circumstances lack of the article marks non-identifiability, and why it doesn't mark it always? Why can the article be left off?"

That's a great way to pose the question.

[Stephen Carlson]

The article is so ubiquitous and important (well, at least for exegetes who try to find gold nuggets) that it's a wonder we don't yet know more about it. Not even in descriptive level.

It may be because it is so ubiquitous. Latin was mentioned above, a language which does not have any article. When students first learn this, they are like "What?" How do you know when to put them in when you translate into English?" Yet in a few months they are regularly adding the articles where necessary in their English translations, because fluent English speakers just know when to do it. The same with the article in Greek. Often, it seemingly corresponds sufficiently to the definite article in English that we don't even notice it (except in so far as it might help with parsing a difficult form). When it doesn't, we easily adjust. We pick up a "feel" for the article that is sufficient to the reason we are reading the Greek.

The people with the best "feel" for a language, of course, are the native speakers. Unfortunately, native speakers are notoriously poor at describing how their own language works. We expect more out of grammarians, linguistics, and language teachers, but as this ongoing survey of grammars is showing, their "feel" for the Greek article isn't paying off in their descriptions.

[Stephen Carlson]

Our next grammar to look at is Stephen H. Levinsohn, Discourse Features of New Testament Greek: A Coursebook on the Information Structure of New Testament Greek, 2nd ed. (SIL, 2000). Levinsohn is best known for bringing information structure considerations to the study of NT Greek, and although he has many publications some more recent, the one that I am citing here is the most influential. As far as prepositional objects go, he does not focus specifically on them, though his approach is applicable to them as a special case and his examples often include them.

With no further ado, I will cite some relevant excerpts:

This chapter is concerned with such purpose [scil. of the article]. It claim that, if the referent of an anarthrous noun phrase is known and particular (or, to be more exact, if the author assumes that the reader will be able to assign it unique referential identity--see sec. 91), this gives it prominence. It is marked as prominent because it is of particular importance.

As noted in the introduction to this chapter, my claim is that, if a reader can assign unique referential identity to an anarthrous noun, then it is prominent (the center of attention or focal). My interest is therefore in a two-way division of anarthrous substantives, between those that Wallace would classify as "definite" and those that Porter (loc. cit.) would call "non-particular," whether Wallace would view them as "indefinite" or "qualitative."

According to the principle of sec. 9.2.1. [scil. the default rule], further references to activated participants are normally articular. Anarthrous references to activated participants are therefore of particular significance. In particular, they make the participant and/or his or her initiative or speech prominent, because it is of particular importance.

The principle of section 9.2.3 that anarthrous references to activated participants indicate prominence does not apply just to proper names for people. It can be applied to any noun with unique referential identity. If an anarthrous substantive has a unique referent and is activated, then its referent is prominent.

***

The problem with an anarthrous substantive that is not a proper name is that, even if it has already been used in the passage (i.e., it has been activated), the new reference could still have an indefinite force. For example, various commentators have suggested that anarthrous instances of law (νόμος) in Romans and Galatians refer to law in general (i.e., it has indefinite force) to to "its quality as law" (Sanday and Headlam 1895:58), even after νόμος has been activated. However, Cranfield (1975:154 fn. 2) maintains, "The view . . . that it was Paul's custom to place the article before νόμος when he was using it with reference to OT law and that, when he omits the article, he is using the word in a general sense, cannot be sustained."

Now, one cannot completely exclude interpretations such as that of Sanday & Headlam. Nevertheless, once a concept like law or faith has been activated, further anarthrous references to the concept may well have "definite force" if the reference occurs in a potentially focal position in the clause or sentence. This position is greatly strengthened if articular references to the concept are found in the same passage.

Levinsohn's approach is very different from that of traditional Greek grammars and it takes some getting used to for those raised on such grammars. One benefit is that it makes a testable proposal for why certain nouns that are inherently definite (or having "unique referential identity"), such as proper names, may lack the article. The traditional grammars are useless on this point.

Nevertheless, I have a couple of demurrals to Levinsohn's proposal:

The first one is that it takes as its starting point that the noun is definite, when the main exegetical question is often whether an anarthrous noun is definite or something else. All Levinsohn's proposal boils down to in this case is that the exegete simply cannot assume that an anarthrous noun is not definite when it is prominent. This is not really helpful for Greek to English interpreters, but those producing Greek sentences might find it useful.

The second one is that non-marking for prominence is backwards, counter-intuitive, contrary to how other linguists think about it, and probably wrong. I think it is better to see the article as marking a presupposition that the noun is identifiable, and that the reason that informationally prominent definite nouns are not marked is because presuppositions are not marked in assertions, or more precisely, presuppositions are marked in order to exclude them from the assertion, in a bid to limit the assertion to a single atomic proposition (see Barbara Abbott, "Presuppositions as nonassertions," Journal of Pragmatics 32 [2000]: 1419-1437). In other words, a prominent definite noun is anarthrous, because everything else around it has been marked somehow as less prominent. It remains an open question for me what an article for a focal constituent is doing, however.