A word order tendency: old information to the front

Started by wm.annis, July 09, 2023, 08:27:19 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.


I've been dissatisfied with the word order discussion in the Horen for a while now, and in doing some background research on that, I noticed a tendency in the dialog in the first Avatar film. So, a while back I sent Paul a long question, which included a lot of examples of that dialog he produced for A1. In it I suggested that definite noun phrases (including pronouns) tended to move to the front of a sentence. After some conversational back and forth, Paul has endorsed this analysis. I posted a copy of the email I sent him a while back: Feb 11, 2023 forum post.

Paul's comments from early this morning:

Quote from: Karyu PawlHere's what I would do at this point for the new version of the Horen:

First, I think your Korentsyìp Amip is eminently worthy of inclusion, since it expresses a genuine tendency that people are probably unaware of. I mean, I was unaware of it, at least consciously! Your examples from A1 are convincing.

In enunciating and explaining the rule, I'd suggest that instead of saying "Definite constituents are often fronted," you modify that slightly to "Definite constituents tend to be fronted." The second version is stronger, since "often" is notoriously vague. "Tend to," however, implies that the phenomenon occurs more than 50 percent of the time. (Would you agree?) Also, be sure that readers understand we're talking about stylistics here rather than grammaticality.

I think my explanation involving the tendency to start a sentence with "old information" and move on to new would be interesting to readers, provided you felt it would fit in. If so, a condensed version of what I originally wrote to you, revised as necessary so it's in your voice, not mine, would be appropriate, although I'm not sure how much space you'd want to give it.

As for the very intriguing idea of a relationship between fronted definites and obligatory initial topicals (in Forest Na'vi, that is!), let's punt on that for the time being. My hunch at this point is that the two phenomena, while resulting in the same word order, have different motivations. But I need to think more deeply about that.

I have a more radical analysis involving the topical, but he has passed on that for now.

His discussion about old information and how it works into conversations:

Quote from: Karyu PawlThe reason, I think, has to do with "old" vs. "new" information. This is something I may be overly sensitive to, since it was a major topic in my Advanced Writing for Business course at USC, and it's discussed extensively in the book I used. The idea is that for smooth conversational flow that's easy for the listener/reader to follow, information in a sentence should move from "old"--that is, material that has already been referred to—to "new"--material being introduced into the discourse for the first time. This means sentences in a paragraph ideally link up in this form:

S1: [ . . . new 1]
S2: [old1 . . . new2]
S3: [old2 . . . new3]   etc.

That's overly simplistic, of course, but it illustrates the idea that for smooth information flow, the tail of one sentence links up with the head of the following sentence, and so on.

For example, which of the following is to be preferred?

A. One of the most difficult poems I've ever read is "Tom's Garland." The Jesuit priest G.M. Hopkins wrote it.

B. One of the most difficult poems I've ever read is "Tom's Garland." It was written by the Jesuit priest G.M. Hopkins.

It's clearly B, despite the fact that B's second sentence is in the much-maligned passive voice. In A, the listener is whipped around in the second sentence by something out of the blue that hasn't been referred to previously—"What does this priest have to do with what you just said?" The connection to the previous sentence only comes with the "it" at the very end. In B, the initial "it" immediately establishes the link to what's already been mentioned and leads the listener smoothly to the new information.


Since definite constituents generally refer to old information, I think the old-to-new idea may explain their frequency in initial position in Na'vi, along with the complementary idea that final position is the place for the most impact. So in "fìpoti oel tspìyang," fìpoti is "old," having been referred to previously, while tspìyang comes at the end of the clause, thereby increasing its impact.

Note that in this the lower need for a passive in Na'vi is obliquely explained away. We can just front old information in Na'vi's more flexible word order, without needed a new verb construction to handle that movement.