Second Language Learning and Language Teaching

Touchstone (McCarthy et al., 2005); students match questions and answers in a
dialogue, ‘Complete the conversations’ in fill-in sentences and then practise cor-
rect responses to ‘Thank you’ and ‘I’m sorry’.
This view of conversation relates to the speech act theory derived from philoso-
phy or linguistics, which assigns functions to utterances: ‘Open the door’ is mak-
ing a command; ‘Why is the door open?’ is making a question, and so on. This is
closer to Lang
4
, the social side of language. Such functions of speech formed the
basis for the communicative teaching approach, seen in the functional/notional
syllabus advocated by David Wilkins (1972). Its influence can be seen in almost
any coursebook to this day. Just Right (Harmer, 2004), for instance, teaches func-
tions such as ‘making promises’, ‘paying compliments’ and ‘giving opinions’.
The difficulty with teaching functions has often been the disconnection from
the structure of conversation involved in teaching one function at a time; how do
you practise paying compliments without knowing when to pay a compliment or
how to reply to it? Hatch’s conversational structure provided one way of connect-
ing functions to conversational moves. Hence People and Places taught compli-
menting as part of a three-move interaction:
stating:
Simon: This is my new jacket.
exclaiming:
Helen: What a smart jacket!
complimenting:
Helen: It suits you.
Students had to continue in this vein by commenting on the other things that
Simon and Helen were wearing in their pictures and then describing the clothes
of other students.
Paul Seedhouse (2004) points out how this type of approach differs from the
discipline of Conversation Analysis (CA) that it superficially resembles. CA does
not try to establish categories and units in a fixed structure; instead it looks at a
slice of conversational interaction and tries to work out what is going on from the
point of view of the participants; ‘For those trying to understand a bit of talk, the
key question about any of its aspects is – why that now?’ (Schegloff et al., 2002).
Classroom interaction and Conversation Analysis
166

---

The most obvious feature of interaction is that people take turns to speak. One
exchange of turns is the adjacency pair:
A: What’s the time?
B: Five o’clock.
The move by the first speaker is followed by a related move by the second speaker,
chosen out of a limited range of acceptable options. Sometimes, as in question
and answer, the second speaker has little choice; after compliments, there may be
a less conventional range of responses; after stating opinion:
‘I love Picasso’s blue period.’
there may be an obvious agreeing/disagreeing move:
‘I can’t stand it.’
or a more nebulous range of options. While the second speaker may indeed decide to
say nothing at all, this is a highly marked option: deliberately failing to respond 
to ‘Good morning’ would be the height of rudeness, in CA amounting to a refusal to
accept social solidarity.
The two parts of the adjacency pair, however, do not necessarily follow each
other:
A: What’s the time?
B: Why do you want to know?
A: So I can put this letter in the post.
B: Five o’clock. You’re too late.
The speakers keep an ongoing idea of the adjacency pair in their minds even when
they are diverting onto side issues. In an early experiment (Cook, 1981), I tried to
see the extent to which the concept of the adjacency pair was established in L2
users’ minds by getting them to supply first or second moves, finding that the
adjacency pair indeed had psychological reality for them.
Central to the idea of interaction is what happens when it goes wrong – the
organization of repair. According to Emanuel Schegloff et al. (2002), this is not the
same as the failure to communicate covered by the communication strategies
described in Chapter 6, but is an interruption, after which interaction is restored.
Usually a distinction is made between self-initiated repair by the same person:
A: Where’s the saucepan? Sorry, the frying pan.
and other-initiated repair by the other speaker:
A: Where’s the saucepan?
B: Where’s the what?
A: The saucepan.
B: Oh, it’s on the bottom shelf.
For the classroom this occurs at two levels; one is the repair of the classroom inter-
action itself, where the teacher or students have to make clear what is going on,
which may well be in the first language; the other is at the level of the interaction
sequence of the language learning activity, which will normally be in the second
language. Schegloff et al. (2002) point out that repair is the essence of the L2 
Describing conversation 167

---

classroom interaction and that much depends on how people understand and
produce self-repair in a second language.
While CA has often been concerned with interaction in constrained institutional
settings, this is seen as related to wider settings rather than unique. The language
teaching classroom has its own characteristic forms of turn-taking, adjacency pair,
repair, and so on. Paul Seedhouse (2004) shows how turn-taking depends on the task
involved, particularly crucial in task-based learning. The problem with applying
CA to language teaching, however, is that its aim is to describe conversational
interaction as it happens, rather like a Lang
3
sense of language as a set of external
sentences. But it does not say how the participants acquire the ability to interact
and so help with how to teach it. It may be possible to deduce how the learner is
proceeding and what the teacher should do, but this depends largely on other
learning theories and approaches, such as the interaction hypothesis dealt with in
Chapter 12, not on Conversation Analysis itself. A CA analysis can tell us whether
a repair occurred and whether it was successful, but it cannot in itself say whether
anything was learnt.
Classroom interaction and Conversation Analysis