The Word Order of Clitic Pronouns

A few days ago I posted a few texts where we had the very same clause used six times and a seventh time nearly identical. In that post, I simply asked a question: What makes them tick? At that point in time, I didn’t give much explanation. Well, actually, I didn’t give any explanation. I want to do that now. It would have come sooner, but I’ve spent the past several days examining roughly a thousand clitic pronouns (the accusative, dative, and genitives enclitic forms of ἐγώ). Incidentally, looking through those occurrences is how I found the examples from my previous post.

My basic claim is that these pronominal clitics attach to the right of the most pragmatically salient phonological phrase possible. That is to say: where the clitic attaches says something about what the author/speaker considers the most important part of his clause.

For this reason, clitics are regularly pulled forward by fronted Focal constituents (Salient, asserted information – italicized in my translation), as seen below:

(1) Matt 11:27 Πάντα μοι παρεδόθη ὑπὸ τοῦ πατρός μου
All things have been given to me by my father.

(2) Luke 12:14 τίς με κατέστησεν κριτὴν ἢ μεριστὴν ἐφʼ ὑμᾶς;
Who appointed me judge or arbitrator over you?

(3) John 1:38 πόθεν με γινώσκεις;
How do you know me?

(4) 1 Tim 1:12 ὅτι πιστόν με ἡγήσατο…
because he considered me faithful

Because focal constituents received a clause’s central phonological stress (just like in English), the clitic is pulled forward.

Examples (1-4) provide extremely helpful parallels to what we see in Mark 14:30, 72; and Luke 22:34. In these verses, Mark and Luke have chosen to mark τρίς as the dominate focal element. By doing so, these two authors create a contrast between Peter’s claim of zero denials and his actual three denials. That is to say, the central focus of these clauses is the number of denials, rather than the act of denial.

(5) Mark 14:30 καὶ λέγει αὐτῷ ὁ Ἰησοῦς· ἀμὴν λέγω σοι ὅτι σὺ σήμερον ταύτῃ τῇ νυκτὶ πρὶν ἢ δὶς ἀλέκτορα φωνῆσαι τρίς με ἀπαρνήσῃ.
“Truly I tell you,” Jesus answered, “today—yes, tonight—before the rooster crows twice you yourself will  three times you’ll disown me.”

(6) Mark 14:72 καὶ ἀνεμνήσθη ὁ Πέτρος τὸ ῥῆμα ὡς εἶπεν αὐτῷ ὁ Ἰησοῦς ὅτι πρὶν ἀλέκτορα φωνῆσαι δὶς τρίς με ἀπαρνήσῃ·
Then Peter remembered the word Jesus had spoken to him: “Before the rooster crows twice you three times you’ll disown me..”

(7) Luke 22:34 ὁ δὲ εἶπεν· λέγω σοι, Πέτρε, οὐ φωνήσει σήμερον ἀλέκτωρ ἕως τρίς με ἀπαρνήσῃ εἰδέναι.
Jesus answered, “I tell you, Peter, before the rooster crows today, three times you’ll deny knowing me.”

In translation, I have represented this by italicizing three times and contracting the English auxillary in order to make clear where the main focus of the statements is.

In clauses where the central phonological stress appears on the verb, the clitic follows the verb – assuming that the clitic in question is working at the clause level rather than the phrase level (e.g. an accusative clitic functioning as an object in a clause vs. a genitive clitic functioning as a possessive in a noun phrase). In a sense, this post-verbal position might be considered the default, but I suggest that this is merely the result of principle I’ve delineated above: that clitics attach to the most pragmatically salient phonological phrase possible. The verb, being the central unit of a clause, by default receives the primary phonological stress. This is especially clear in reduce clauses with simplified information structure, such as participial clauses (especially substantival and attributive ones). Complex structure in participial clauses creates processing difficulties for a speaker’s audience, detracting one’s attention from the matrix clause the participle supports. For this reason, phonological stress in participial clauses more often than not follows the “default” structure: the verb receives the primary phonological stress. And thus, pronominal clitics in participial clauses will connect directly to the participle.

Imperative clauses function similarly. The verbal event commanded in the imperative tends to be the central element of the clause unless there is some other element in the clause which is being restricted or contrasted. Consider the English clauses below.

(8) Give me the ball.

(9) Give me the ball.

(10) Give me the ball.

(11) Give me the green ball.

(12) Give me the green ball.

(13) Give me the ball.

Example (8) in the unmarked form will always place the primary phonological stress on the verb: give. In contexts where there are no other mitigating circumstances where there is no one else to give a ball to and there is only one object for someone to give, examples (9-12) are not felicitous. But example (13) is. This sentence could either be used in a situation where the hearer of the statement chose to disregard the speaker when example (8) was uttered.

Likewise, in contexts where the speaker needs to specify a particular recipient or object, example (8) would not be felicitous – and also any of the other examples that do not fit the phonological-information structure requirements of the speaker. But the point is that in the majority of cases, the phonological structure of (8) will be used with a single primary stress on the verb.

Thus in Greek we find that in the vast majority of commands, the pronominal clitic directly follows the verb:

(14) Acts 12:8 ἀκολούθει μοι.
Follow me.

(15) 2 Cor 12:13 χαρίσασθέ μοι τὴν ἀδικίαν ταύτην.
Forgive me this wrong.

If the pronominal receives its own discourse function within the clause, the speaker uses the accented form. Compare example (14) above with example (16):

(16) John 12:26  ἐὰν ἐμοί τις διακονῇ, ἐμοὶ ἀκολουθείτω.
If anyone should be my disciple, then me, he must follow.

In John 12:26, Jesus (the reference of ἐμοί) is established as the Focal element in the protasis, where the force of the clause is: “If anyone should be my (as opposed to anyone else’s) disciple….” Then in the matrix clause (the apodosis), ἐμοί is fronted with that same force: following Jesus and only Jesus is the the prerequisite for being his and only his disciple.*

This discussion of post-verbal clitics does well to explain Luke 22:61 and John 13:38, show below as examples (17-18).

(17) Luke 22:61 καὶ ὑπεμνήσθη ὁ Πέτρος τοῦ ῥήματος τοῦ κυρίου ὡς εἶπεν αὐτῷ ὅτι πρὶν ἀλέκτορα φωνῆσαι σήμερον ἀπαρνήσῃ με τρίς.
Then Peter remembered the word the Lord had spoken to him: “Before the rooster crows today, you will disown me three times.”

(18) John 13:38 ἀμὴν ἀμὴν λέγω σοι, οὐ μὴ ἀλέκτωρ φωνήσῃ ἕως οὗ ἀρνήσῃ με τρίς.
I tell you the truth, before the rooster crows, you will disown me three times!

On the basis of my claims above that clitics following an initial verb constituent a more basic clause type, I propose that in these two instances, the primary stress falls only on the verb. I hold that here Luke and John make their central focus the denial itself rather than the number of denials or who is being denied/disowned. For Luke, this draws attention to Peter’s realization of the fact of his denial. He wasn’t going to disown Jesus at all, but he did it. The act of denial itself is the central focus. Conversely, John’s presentation of the Last Supper remove’s Peter’s claim that he will never deny Jesus. Thus, when Jesus makes the statement in (18), there is no three to contrast with zero. It is again the act of denial itself that in focus contrasted with Peter’s exclamation that he would lay down his life (13:37).

Now, we have examined examples where the pronominal clitic associates itself with the verb of a clause or with a fronted constituent. In none of these examples have we seen a clitic attaching itself to the verb while there is also a fronted, focal constituent. And that’s the kind of construction that we find ourselves dealing with the last word order variation in my previous post: τρὶς ἀπαρνήσῃ με.

But before we can focus on this example in particular, we need to look at the constituent-verb-pronominal clitic ordering more broadly, particularly as it relates to what we saw in examples (1-4). As noted above, pronominal clitics tend to be pulled forward by focal constituents. The question is: under what circumstances are they pulled forward and under what circumstances do they remain in place?

I propose the key to answering this question is found our English examples (12-13), where major phonological stress appears on two phrases rather than only one. These sentences have been repeated below as (17-18).

(17) Give me the green ball.

(18) Give me the ball.

Each of these clauses has more than one major phonological phrase. Example (17) has major stress on me and green and (18) has major stress on give and the ball. Because of previous conclusions about the difference between clitic and non-clitic personal pronouns, it is clear that (14) does not fit the bill for dealing with τρὶς ἀπαρνήσῃ με. In Greek, we would expect the equivalent of the stressed me to appear as a non-clitic form. This leaves us with (18), where stress appears on the verb and one of its constituents.

I propose that this is the best way to approach and understand the final two variations found in Matthew 26:34, 75.

Matthew 26:34 ἔφη αὐτῷ ὁ Ἰησοῦς· ἀμὴν λέγω σοι ὅτι ἐν ταύτῃ τῇ νυκτὶ πρὶν ἀλέκτορα φωνῆσαι τρὶς ἀπαρνήσῃ με.
“Truly I tell you,” Jesus answered, “this very night, before the rooster crows, three times, you will disown me.”

Matthew 26:75 καὶ ἐμνήσθη ὁ Πέτρος τοῦ ῥήματος Ἰησοῦ εἰρηκότος ὅτι πρὶν ἀλέκτορα φωνῆσαι τρὶς ἀπαρνήσῃ με·
Then Peter remembered the word Jesus had spoken: “Before the rooster crows, three times, you will disown me.” And he went outside and wept bitterly.

In this case, Jesus words and their order presented by Matthew, if read out loud would receive two major phonological stress marks: one on τρὶς and one on ἀπαρνήσῃ. In order to represent this in written translation, I fronted and separated with a comma three times and then also removed the contracted auxiliary and put all the stressed words in italics. Like example (13) and its respective scenario where the hearer of the command initially refused to accept it, Matthew presents Jesus as correcting Peter’s entire response with force.

So then, the variation in the presentation of Peter disowning Jesus three times can be explained by a single clause phonological principle which interacts with the information structure of the clause and discourse. Because of the close relationship between phonological phrasing in a clause and information structure, personal pronouns function, in a way, as markers for the discourse functions Topic and Focus. The non-clitic forms receive a discourse function, while their clitic forms attach to a discourse function,** most often Focus. In fact, the parallel between clitic attachment and Focal information in a clause is great enough to deserve further study – something which I hope to dig into in the coming months.

*Note: In Steve Runge’s LDGNT analysis, he treats the second pronoun as the Topic of the clause rather than the Focus. This is a side issue which is irrelevant here. When either discourse function is used with a pronoun, the accented form is always used.

**Note: That is, unless they are used with a preposition.

About these ads

11 thoughts on “The Word Order of Clitic Pronouns”

  1. I’m reminded of an item in a little book on Bavarian dialect that explained the nuanced distinctions between several expressions which superficially mean the same thing (which meaning you can readily guess):
    1. Leck mi am Arsch! (mild contempt)
    2. Am Arsch sollst mi lecka! (strong contempt)
    3. Du leckst mi am Arsch! (something like, “Well, I’ll be damned!”)

  2. Thanks, Mike. That makes excellent sense. Of the examples you used in the first post, the Luke examples become particularly interesting, since the word order changes from the prediction to the remembrance.

    Luke 22:34 ὁ δὲ εἶπεν· λέγω σοι, Πέτρε, οὐ φωνήσει σήμερον ἀλέκτωρ ἕως τρίς με ἀπαρνήσῃ εἰδέναι.
    Jesus answered, “I tell you, Peter, before the rooster crows today, THREE TIMES you will deny that you know me.”

    Luke 22:61 καὶ ὑπεμνήσθη ὁ Πέτρος τοῦ ῥήματος τοῦ κυρίου ὡς εἶπεν αὐτῷ ὅτι πρὶν ἀλέκτορα φωνῆσαι σήμερον ἀπαρνήσῃ με τρίς.
    Then Peter remembered the word the Lord had spoken to him: “Before the rooster crows today, you will deny me three times.”

    I.e, in 22.34, the emphasis is on the number of times Peter will deny Jesus. In 22.61, the focus shifts more to the simple of fact of the denial. (Though 22.61 is the ‘default’ order…)

    1. Sorry this post was a little slower coming that I had intended. Its been a busy week.

      These clitics get more interesting with negated clauses & ἵνα clauses. I’ll be writing more.

  3. If my records are correct, this post completes your service project for the Junior Jedi Discourse Badge. This means you change rank from Advanced Geek to 1st Class Nerd. Remember to wear your sash and uniform to the meeting so that we can award you.

  4. I’ve just been reading Anderson’s book on clitics (2005), and haven’t had much experience with these kinds of clitics, so it was a cool co-incidence that I just read your blog post!

    With your last two examples of Matthew 26:34,75, are you proposing that they have two major but distinct pragmatic emphases? That seems quite odd to me considering that the words are adjacent.

    1. I’m proposing that there are two major phonological phrases that receive a pragmatic function, yes. Is it the possibility of double focus that is odd to you or the fact that such a thing could occur in two adjacent constituents?

      Either way, I’m not entirely sure why you would find it odd. If you have any more thoughts on that, I’d be curious to hear more.

      (Incidentally, I’ve been interested in Anderson’s volume — though transformational grammar has never been that attractive to me. In any case, it is on my Amazon wishlist. I’ve been using Zwicky & Halpern (2006) and Halpern’s article in Zwicky & Spencer (2001) primarily in what I’ve been working on here.)

  5. What seems odd is (if I’m understanding you correctly) that there are two phonological adjacent phrases with the same stress receiving distinct pragmatic emphases. Or do they combine to make one pragmatic emphasis? Or based on Greek syntax would τρὶς ἀπαρνήσῃ never form one constituent, so it must necessarily be two?

    I’m accessing Anderson (2005) via Oxford Scholarship Online, which is quite horrid (it has a silly 5 page limit and no pdfs) but as my uni library doesn’t have a paper copy it will have to do. I haven’t got to his chapters on syntax yet, but it doesn’t look like it will have much transformational grammar. Instead he proposes an optimality theory approach to syntax.

    1. Ahh, that’s where we’re missing each other. No, I’m not claiming that the two phrases have the same stress. I think the verb’s phonological phrase is the bigger stress phrase than the fronted constituent. But I do consider both stress phrases to be salient — though the one with the clitic is more salient than the other.

      As for Anderson, as I understand it he’s using Optimality Theory within either Minimalist Program or Government & Binding. I’ll take a look at it online where you’re reading it. Thanks for that.

Leave a Reply

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

You are commenting using your 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