Re: review of Opening Dialogue

Gordon Wells (gwells who-is-at oise.utoronto.ca)
Wed, 7 Jan 1998 22:59:16 -0500 (EST)

To answer the (relatively) easy question first:

On Wed, 7 Jan 1998, Mike Cole wrote:

> ------
> Gordon: What is the implications of using the phrase IRF versus
> IRE (Initiation-Response-Feedback) vs (Initiation-Response-
> Evaluation) that comes from Mehan et al work?

Very simply put, there are more options in the third, follow-up, slot
than making an evaluation of the contribution in the previous slot.

The triadic exchange structure - Initiate, Respond, Follow-up - occurs in
many contexts and is the normal structure when the Initiating move is a
Demand of some kind. Here are some (invented) examples:

a) A: What's the time, please? b) A: What's the time in New York?

B: Three o'clock B: Three o'clock

A: Thanks A: No. New York is 3 hours ahead

c) A: Who d'you think will win the d) A: Would you like a drink?
election?
B: The greens, I expect B: Yes, a beer if you have one

A: Really, why's that? A: Labatt or Heineken?

B: Well, ecology's the hot issue B: Heineken, please
this time
A: Here you are
A: Yes, but do you think that'll
override traditional loyalties?
e) A: What's the name of your boss?
B: I don't see why not
B: Ms Devere
A: I doubt it
A: I'll be sure to get it right
when I next meet her

f) A: What's the name of this shape? g) A: How are you feeling today?

B: A rhombus B: Not very well, I've got a
dreadful sore throat
A: That's right. Well done
A: I'm sorry to hear that. I hope
you'll soon feel better

It's not difficult to decide which of these examples (b and f) come from
an instructional setting. It's not that there is a third move, as all
the examples contain a third move. Rather it is the nature of the
initiating demand and the way in which the response is treated. Only in
instructional activities do people ask questions in order to evaluate the
response.

However, even in classrooms, that doesn't have to be the case. We are
currently coding transcripts of whole class discussions from DICEP
classrooms. Although we haven't begun the analysis yet, my guess is that
less than 25% of triadic exchanges have an evaluation as the third move.
As I argued in my paper in MCA ("Using the toolkit of discourse ...)
there are many other options that can occur in the third move, such as
comments that add to or offer an example or explanation of the student
response, or invite the student to make such a move. When the teacher
frequently chooses these options, students feel free to check that
they've understood before answering, to ask questions of their own and to
offer contributions without waiting to be asked a question. In fact, the
conditions are in place for real dialogue to develop.

What follows is the first few minutes of a science discussion in a
culturally diverse combined grade 3 and 4 class that followed some
practical work in which the children, in groups, had estimated the mass
of certain materials and predicted what would happen when they changed
state (e.g. from ice to water). On that previous occasion the teacher had
invited the groups to discuss their predictions; she had also asked
speakers, when appropriate, to relate their contributions to those that
had gone before. In the following, they discuss the relationship between
estimating and predicting.

526 T: When I say `estimate' and `predict' or `predict, don't
527 bother to estimate', what am I saying? Are they two
528 different activities - mental activities that you have to
529 do? Are they two different mental activities? How many
530 people think they're two different mental activities?
531
532 [Most children put up their hands]
533
534 T: How many people think they're NOT two different mental
535 activities?
536
537 [Philips raises his hand]
538
539 T: Why not, Philips?
540
541 P: Cos like they're more- more or less the same.
542
543 T: How are they the same?
544
545 P: `Estimate' is like- sort of like guess and `predict' is
546 like um- guess- yes, guess too .. [**]
547
548 T: [OK]
549 You heard somebody who says they're not the same.
550 Now there's a whole bunch of people who say they're two
551 different mental activities.
552 What do YOU think, Emily?
553
554 E: I think that- well I don't agree with Philips because
555 I think that they are two different things . because
556 `predict' is sort of like guess what will happen . and
557 then `estimate' is like you estimate the mass . using
558 a form of weight, centimeters . and it's not just with
559 mass, you estimate other things.
560
561 T: OK
562
563 AU: I don't agree with Philips either because `predict'
564 sort of means like what WILL happen and `estimate' is the
565 er- do it- estimating something that's already there, but
566 taking it further.
567
568 T: Now, listen to both answers. None of the answers are
569 right or wrong. Will someone make a distinction?
570 Auritro has made a little- even a more- a greater
571 distinction. OK?
572
573 J: I don't agree with Philips (laughs) because he said that
574 `estimate' is guessing . and `predicting' is ALSO
575 guessing but . um- actually guessing is also different
576 from those two because when you guess you don't have very
577 much information about the object or the thing (T: uh-huh)
578 and so you're just making a- like a wild guess . but
579 when you predict you're- you're actually you're maybe
580 doing an experiment . and you are trying- using the
581 information, you are trying to find out what would
582 happen-
583
584 T: Mm
585
586 J: - and estimating is um different from guess because . you
587 have um certain information, for instance if you estimate
588 the mass, you get the object in your hand and you . and
589 you have the weights in the other hand and you can sort
590 of . like estimate the . mass, so it's not guessing.
591
592 T: OK. Christopher.
593
594 C: Well, I don't agree with Philips either because . um
595 `estimating' is sort of . um . when you get- you get er
596 you at least know a bit of what it is but `predicting'
597 is just making . like a guess.
598
599 T: So you think `predicting' is a guess, `estimating' is
600 not? OK? Benjamin?
601
602 B: Um- I-I don't agree with Philips . um as well and I
603 think that `predicting' . is . if you predict then you're
604 saying that . um . I'm predicting what's going to happen
605 to me tomorrow - what I'm going to do tomorrow and with
606 `estimating' you would- it would just be something like
607 um . if . um . you would- you would est- you would
608 estimate um . um .. estimate how heavy er <something is>.
609
610 T: OK, that's a good attempt again
611
612 ?: <Yeh>
613
614 T: [I'll come to you] (to Emily, who has her hand up)
615
616 AM: [Well, I ..] I don't agree with Philips because
617 `estimating' is like you look at the object and you
618 think about how much it weighs or how tall it is .
619 and `predicting' is . like you predict if it'll be-
620 like . predicting is like- you predict whether . you
621 can say it like what we did yester- the other day
622 when we predicted whether it'll change or not change
623 or it'll be the same ..
624
625 T: OK .. Emily?
626
627 E: I'd really like to <revise it> a little but I started by
628 <changing a little> and discuss everything so we get a
629 little information and then you go further see what will
630 happen next. I think it's true and . for estimating we
631 also . like- <for anything> you look at the object . and
632 then you guess- well you DON'T guess but then you try to-
633 like you have a- some weights and then you . like try to
634 feel the um- see what it weighs or that's how I <think it
635 is (trailing off)
636
637 T: OK, I'll come to you (to Wilson). Emily H, then Wilson.
638
639 EH: I think `estimate' is **** something when you try and
640 guess how much it weighs and `predicting' is sort of .
641 um when *** you predict that **
642
643 T: OK . Wilson?
644
645 W: I don't agree with um . Philips because . um in our math
646 book it says `estimate to the nearest tenth' but it
647 didn't- it doesn't say `PREDICT to the nearest tenth'.
648
649 T: That's right. So what's the distinction?
650 Good, you're using- <you're> using experience in math .
651 to help you make a distinction . and do you ever see in
652 your science that I ask you to predict? ...
653
654 [One or two children nod in agreement]
655
656 T: In your science activity it also asks you to estimate.
657 So the very fact that sometimes for certain things you
658 use the word `predict' and for certain you will have to
659 use `estimate', therefore you think there's a distinction?
660 Just from usage, OK?
661 Any more?
(from Wells, 1997)

The discussion goes on for another 500 lines, as the class progressively
refines the relationship between the three terms. There are several
exchanges that can be described as IRF, but there is hardly a single case
of an evaluation in the third move.

My main point, then, is that triadic exchanges are not intrinsically either
good or bad. It depends on the activity or task that they mediate.
Where a whole class is involved in inquiry, it may well be necessary for
the teacher to manage the turn-taking through a form of IRF, but as in
the above example, this format doesn't necessarily limit the
discussion. In fact, it can help to make it 'progressive'.

Reference
Wells, G. From guessing to predicting: Progressive discourse in the
learning and teaching of science. In C. Coll & D. Edwards (Eds.)
Teaching, Learning and Classroom Discourse. Madrid: Fundacion Infancia y
Aprendizaje, 1997.

Gordon Wells, gwells who-is-at oise.utoronto.ca
OISE/University of Toronto
http://www.oise.utoronto.ca/~ctd/DICEP/