Last week was Chuseok, the rice harvest festival, here in Korea. It¡¯s a time when families play Yut, a traditional game which is a little like parchesi played with divination sticks.
About two weeks ago, one of my grads gathered this data. Like a lot of data transcribed from the teacher¡¯s point of view, it is written up as a kind of dialogue between one teacher called T and a monster that has forty bodies, but only two ears and one mouth, which appears as "S". Here's some fresh data, then:
T: Children, look at this.[C] This is Yut. [S] OK? What's this?[Q]
S: Yut!
T: Yut.
Two participants, three turns, six utterances (if we define the utterance as a real or potential change of speaker). Three of the utterances are grammatical sentences: those marked C for command (imperative), S for sentence (indicative declarative), and Q (indicative interrogative). Three utterances are not (which in my experience is pretty normal for classroom discourse).
But how to explain the fact that we consider this to be a unit? I guess the usual explanation (Mehan, Sinclair and Coulthard, Nassaji and Wells) is that it IS a unit of triadic dialogue; it is one exchange of IRE (or IRF) and as such it consists of one initiate turn, one response turn, and one feedback/follow up/evaluate turn.
This is a little hard to maintain when we look at the rest of the data.
T: Cha, yucheul deonchyeosseyo (Now, we throw the divination sticks). What is this?
S: Mo! ( A combination of four sticks)
T: Mo! Then, what's this?
As you can see, things are not so simple. We can¡¯t simply map Sinclair and Coulthard¡¯s moves onto physical turns on a one to one basis. There are two reasons we can¡¯t:
a) The teacher¡¯s turns almost always consist of THREE moves rather than one. One is a REACTING move (Cha and Mo). One is an COMMENTING move (We throw the sticks, then). One is a QUERYING move (What¡¯s this?)
b) The ¡°stitch¡± between one exchange and the next occurs in the MIDDLE of the teacher¡¯s turn, not the end. (after Mo and before Then).
Why are the teacher¡¯s turns so very complex? And why don¡¯t the structural boundaries of the exchanges correspond to the physical boundaries (the changes in speaker)?
One of Mehan¡¯s brilliant insights was to show that the traditional I-R-E exchange is not a modular unit. Exchanges do NOT go in any order. They are carefully ordered.
But they are NOT ordered beforehand, at the beginning of the lesson. Instead, they are ordered in a post hoc fashion. The teacher takes a step, checks to see that the kids are following, and then takes the next one.
Some of this checking is visual (the teacher checks eye contact and facial expression instead of asking a question), and that¡¯s why the teacher¡¯s moves look so complex. That three part turn is really THREE turns:
T: T: Cha (teacher checks eye contact), yucheul deonchyeossoyeo (Now, we throw the devination sticks). (teacher checks facial expressions) What is this? (teacher checks verbal understanding)
T: Mo! (Teacher smiles to show this is the right answer and children smile back) Then, (Teacher pauses to show that a new set of devination rods is under construction. Teacher checks that all children are looking at the devination rods before asking¡¦) what¡¯s this?
But some of the checking is not visual. A lot of it happens, believe it or not, is through that nasty, monologic, dictatorial EVALUATE move.
Evaluate is NOT, contrary to what Sinclair and Coulthard felt, an evaluate of the student¡¯s response. Instead, it¡¯s an evaluate of the felicity or satisfactoriness of the WHOLE EXCHANGE. If that¡¯s okay (from the point of view of action), the teacher goes on. But if it¡¯s NOT satisfactory, here¡¯s what happens.
T: Cha, OK. Look at this. Look at the screen. (shows a chart on the video screen)
S: Cancel!
T: This ¡°Do¡± means [Cancel!].[S]
S: Candy!
T: Oh! No, no. Ha ha. This ¡®µµ¡¯ means ¡®Cancel!¡¯. Teonjin Geosi Modu, Modu Yeoreo beon Deonjyeodo Muhyoga dwineun geoyeyo! (The throw is all, all every time you throw this, it¡¯s all cancelled.) Repeat after me. (Gestures ¡°X¡± with both hands)
S: Cancel! Candy!
T: OK. Look at this sheet. This will help you.
The exchange is extended. The way it is extended is by the Evaluate move, of course.
I think Mehan's great insight means that there is a kind of RANKSHIFT in the evaluate move. It¡¯s not just an operation, not just a response to operational conditions. But it¡¯s not an ACTIVITY, either; if it were that, it would occur at the very beginning of the lesson, laying out the motives for the whole lesson in quite general terms (¡°This is an English lesson.¡± Or ¡°Today we¡¯re going to learn X¡±).
It¡¯s an action: it¡¯s motivated by the goal of successfully completing one exchange (operation) and moving on to the next. But it's an action that is embedded in the ongoing operation.
I can guess that I will catch hell from Tony for saying this. And Gordon Wells himself will probably object tool (because for Gordon it¡¯s the SEQUENCE and not the EXCHANGE that is operational). But there¡¯s an advantage to thinking about it this way.
The payoff is that we can consider the ¡°Evaluate¡± move as a form of RANKSHIFT—it¡¯s BOTH an operation (because it¡¯s a response to the response and so part of the exchange) and an action (commentary on the whole exchange and so part of the sequence).
It¡¯s the same as a grammatical rankshift. When we say:
That he won is not surprising!
we get a clause "That he won" rankshifting down, like God made flesh, an immortal clause pretending to be a mortal noun subject. In the same way, the Evaluate is an action incarnating itself as an operation.
In Gordon¡¯s 1996 article in MCA (Using the Toolkit of Discourse in the Activity of Learning and Teaching, MCA 3(2), pp. 74-101) he tries to articulate activity theory with systemic functional linguistics (Hallidayan grammar).
I think the earlier article doesn¡¯t work very well because
a) activity theory works WELL above the level of exchange and Hallidayan grammar for the most part is concerned with clauses.
b) The activity theory that Gordon presents is REDUCTIONIST: activities are NOTHING but their constituent actions, and actions are ONLY made up of operations.
It seems to me that if we go back to Mehan and understand the Evaluate move as a bit of rankshift, we get around both of these problems:
a) The Evaluate move takes part in BOTH the operation (the exchange) and the action (the sequence) (Mo! That¡¯s right! So¡¦)
b) The action is more than just the sum of its parts, because the Evaluate move is not simply reactive but also pro-active, and even proleptic, bringing the next exchange into focus for the children (Mo! Then¡¦).
I think we also might get a little closer to the key problem that Nassaji and Wells tackle: why are SOME tripartite dialogues so much better than others?
A friend of mine wrote (partly in response to the Craig business) that the difference between lesbian and straight has very little to do with whom you actually sleep with; it¡¯s to do with whose love really helps you get things done and get on with your life.
In the same way, we might say that some tripartite dialogues look forward, and others look back. Some dialogues teach, and others merely test. (And in the same way, we might say that it is not the same for everybody!)
David Kellogg
Seoul National University of Education
PS: A native speaker thought on the data: Hmmmm....must tell Hojin that "then" is not really the same as "so".
Second thoughts: Says who?
dk
---------------------------------
Luggage? GPS? Comic books?
Check out fitting gifts for grads at Yahoo! Search.
_______________________________________________
xmca mailing list
xmca@weber.ucsd.edu
http://dss.ucsd.edu/mailman/listinfo/xmca
Received on Thu Oct 4 16:43 PDT 2007
This archive was generated by hypermail 2.1.8 : Tue Nov 20 2007 - 14:25:43 PST