Reading: Hellerman, J. 2005. Syntactic and prosodic practices for cohesion in series of three-part sequences in classroom talk. Research on Language and Social Interaction, 38(1), 105–130.
I’ve been working on a review of the literature on classroom discourse, and one of the main things I’ve noticed is that a lot of people have been foregrounding the importance of theory and methodology. The idea is that whichever theoretical approach you choose will allow you to see certain kinds of patterns in the data while leading your attention away from other things. So, like the blind men and the elephant, you need to listen to a lot of different perspectives.
Since I’m taking one class on systemic functional linguistics and another with a lot of conversation analysis readings, those are the two main areas where my attention has been lately. With some papers that I read, when I look over the excerpts of the data that they offer, it’s like looking at a Necker cube: I see adjacency pairs out of one eye and logico-semantic relations out of the other.
Here’s what I mean, with a segment of talk from a U.S. history lesson (Hellerman, p. 108):
| 3 | T: | okay the CLASS STRUCTURE:.(.) strict? or mi:ld. what | |
| 4 | idea did you get yesterday. | ||
| 5 | Jin: | mild. | |
| 6 | T: | → | mild. could you move among the classes. |
| 7 | Jin: | (social mobility). | |
| 8 | Aly: | yes. | |
| 9 | Tina: | yes. | |
| 10 | T: | → | yes. by:? |
| 11 | (1) | ||
| 12 | Jin: | plantation. | |
| 13 | Tina: | owning a plan↑tation. | |
Conversation analysis says: This is a pretty clear example of the “initiation—response—feedback” (IRF) pattern that we see all the time in classroom discourse. The teacher asks a question in 3–4, the student answers in 5, and the teacher ratifies the response by repeating this answer in 6. Then, still in 6, the teacher asks a new question, students answer in 7–9, and the teacher repeats the “correct” answer in 10. Lines 10–12 begin a third IRF cycle. (I didn’t reproduce the feedback here because it gets a little complicated.) It’s key that these are “known information” or “display” questions; that is, the teacher knows the answer and is testing whether the students know it too.
The question in line 10 is especially interesting, as Hellerman points out, because it’s not a question at all, in the syntactic sense. Instead, it’s the unfinished beginning of a sentence — what Irene Koshik calls a designedly incomplete utterance — which prompts the student to respond with the remainder of the sentence.
At the same time, systemic-functional analysis says: There is a relationship between each pair of successive IRF cycles. Line 6 creates a logico-semantic relationship of elaboration, restating or further specifying the preceding idea: “The class structure was mild, in other words, you could move among the classes.” Line 10 is an example of enhancement, in this case, causal enhancement marked by by, which asks for the students to name the means by which the process described in 6–9 could happen. So, the teacher chooses a sequence of display questions that provide cohesion to the activity sequence overall.
I wonder if this sort of joint CA+SFL (or interactional sociolinguistic + SFL) analytic approach could be sustained over longer stretches of talk. There are potential correspondences all over the place: phases of interaction might match up with genre stages; “reach” (in the Goffman sense) could be realized through reference chains; stance and modality are closely related.
This is probably what I was thinking when I wrote this drive-by of a post.
School for Linguists 
From an SFL perspective, rather than a cube, don’t you want an onion?
Incidentally I am excessively flattered to have my blog listed on yours alongside blogs of such calibre, expertise and repute! Especially given its level of neglect.
It would have to be something that looks like an onion from one eye, and (what can stand for CA?) a maze, maybe, from the other eye (as in, Schegloff, “Why that now?”). I have no idea what that would look like, but Christian Matthiessen could probably pull it off.