DISCOURSE ANALYSIS FINAL ASSIGNMENT

Analysis of turn-taking in a scripted conversation

2014

1. INTRODUCTION
This paper is a discourse analysis of a food and drink dyad from a student
course book (appendix 1). The analytical approach used is Conversation Analysis
(CA). The study focuses on turn-taking, and shows how a scripted audio recording is
an unnatural representation of what happens in natural conversational discourse.
In order to put the study in context, a summary of the history of CA is given
and a brief definition of the approach is outlined in 1.2. Next an overview of the
discourse being analysed in the study is given in 2.0. The discourse is then considered
with reference to turn-taking and a more detailed look at adjacency pairs is discussed
in 2.1 and 2.2. In 3.0 contextual and cultural factors are analysed leading to a brief
discussion about scripted discourse and authenticity.
\Due to the nature of the scripted dyad, and word count limitations, the
analysis focuses primarily on adjacency pairs and omits direct discussion about presequences, insertion sequences and turn keeping.
2. CRITICAL FRAMEWORK
2.1 What is Conversation Analysis?
CA is the study of natural talk in interaction. Its aim is to see how people
communicate, whether casual or institutional (in the school, government office or any
other establishment). CA developed from ethnomethodology in the late 1960s and
early 1970s principally from work by the sociologist Harold Garfinkel, who was
interested in the question of how people interpreted daily life and developed a sense
of social order. In 1974 work by Harvey Sacks and his close associates Emanuel A.
Schegloff and Gail Jefferson on turn-taking in conversation further developed what is
known today as CA. In short CA,
Focuses on detailed recorded conversations, analysing them for specific
features of their moment by moment production, and interpreting the
significance of the utterances in the light of their environments of action.
(Heritage, 1984.p.292)

The central and interdependent issues of CA are contextual sensitivity,

accountability and the sequential organisation of conversation.
\Contextual sensitivity in CA is concerned with the expectations of language
use. How people interact in different contexts is analysed and interpretations of
language use are made. Particular consideration is given to whether interactants meet
or fail expectations of what is considered appropriate in the context in which they
occur. Additionally the social identities of speakers are considered and how there
verbal actions maintain or break perceived social conventions.
CA realises interactants understand social codes and rules of conduct and are
accountable for their verbal actions. The approach recognises that interactants are
aware of the consequences of not conforming and that utterances are shaped as much
by what reaction is expected as by what they actually want to say.
Early observations in CA studies with natural data, (Sacks et al., 1974)
revealed the approach was useful for looking at how patterns emerge as interaction
unfolds. This research led to the developments of theoretical systems about turntaking and adjacency pairs. From this came recognition of a general principle
underlying conversational organisation: that of the sequential implicativeness of talk.
Eggins and Slade (1997) assert that this is the idea that conversation turns make sense
because they are interpreted in sequence. They note, CA suggests that the most
significant placement consideration in conversation is that of adjacency. Thus
wherever possible the speakers current turn will be interpreted as implicating some
action by the responder in the next turn (P29,1997) Adjacency pairs are seen as
prototypical elements of a broader conversational principle of sequential relevance.
This paper now applies CA theories about turn-taking and adjacency pairs to the
scripted food and drink dyad (appendix 1).

CA suggests that the most significant placement consideration in conversation

is that of adjacency. Thus wherever possible the speakers current turn will be
interpreted as implicating some action by the responder in the next turn
(P29,1997).
Adjacency pairs are seen as prototypical elements of a broader conversational
principle of sequential relevance. This paper now applies CA theories about turntaking and adjacency pairs to the scripted food and drink dyad (appendix 1).
3. DATA PRESENTATION
The food and drink dyad is taken from Reward Starter, a beginners English
course book. The conversation is a scripted discourse used as a listening activity, in
which students are asked to listen to Finn and Selina talking about their favourite food
and drink. While listening learners are asked to tick the things they like and cross
the things they do not like from a list of food given in an earlier exercise in the text
book.
3.1 Turn Taking
A central feature of CA is the study of turn-taking. The analysis of turn-taking is
concerned with how interactants take turns speaking and how who speaks when is
determined. Early conversation analysts Sacks et al.1974) suggested that at any
given point of viable turn transfer two possibilities exist. Firstly, that the current
speaker self-selects and continues speaking. Secondly, the current speaker selects the
next speaker, which would mean the selected speaker could take the next turn and
speak or select another speaker to take the turn.
An example of this can be given by looking at appendix 1. Turn 1 shows Finn (the
speaker) selects Selina to take a turn speaking by asking her a question. Turn 2 shows
Selina cooperating with Finn and deciding to answer thus taking a turn speaking. This
particular kind of turn taking is called an adjacency pair and is discussed in more
detail in 2.2 Prior to this it is useful to look at other features that affect turn-taking. In

natural conversation overlap occurs. This is when an interactant starts speaking before
the turn of the other speaker has finished. Ervin-Tripp (1979) observes that overlap,
(interruptions or simultaneous speech), in turns happens in approximately 5 per cent
of conversation. This suggests that speakers usually know, through complex verbal
and nonverbal signalling, when a turn is ending.
McCarthy (1991) notes overlap is significant in conversation as it normally
signals, urgency, annoyance, or the desire to correct what is being said. The food and
drink dyad contains no interruptions or overlaps. This implies the turn-taking is
cooperative rather than competitive. Overlap in conversation can also be a sign of
heightened emotional involvement. Tannen (1989) defines co-operative overlap as a
sign of a high-involvement style in which interactants show concord, enthusiasm and
interest in each others talk. An example of this is shown in turns (13) and (14) at the
end of the food and drink dyad. However, the lack of involvement shown by the
speakers in the preceding turns makes the sudden interest sound forced and unnatural.
Further indicators for turn taking are highlighted by Wardhaugh (2002) who points
out that a change in pitch level often signals closure or the end of a turn. This happens
in the food and drink dyad. Pitch changes can be heard when listening to the last
words of turns (4), (9), (11) and (12) (see appendix 1 and CD enclosed). The words
often, tea, mayonnaise and though have fall in tone. Pauses are also used to indicate
the end of a turn in conversations. This is illustrated in the turn change between turns
(9) and (10). Here the short pause, of approximately 0.8 of a second, after the word
mayonnaise signals the end of the turn.
In conversation listeners often like to show the speaker they are interested,
understanding, agreeing or simply attending to the message. This is usually done by
inserting a word such as sure, right, or yeah . In addition back- channel vocalisations
such as hmm, ahh, and eh-eh are commonly used by the listener as non-interrupting
signals to the speaker.

Other utterances that typically signal the end of a turn are phrases like you know,
or something and do you know what I mean? Appendix 1 has no examples of backchannel or any short words or phrases to show the speaker they are attending to the
message. As McCarthy notes,
Natural conversational data can often seem chaotic because of back
channel, utterance-completions and overlaps (1991.p127)
The food and drink dyad lacks many natural conversation elements. This paper
now looks atthe adjacency pairs in the dyad to determine whether they are authentic
representations of adjacency pairs that occur in natural conversation.
3.2 Adjacency Pairs
Adjacency pairs allow speakers to allocate and give up turns. Sacks et al
(1974) identify adjacency pairs as major turn types. As with the example above
the most common is the question/answer adjacency pair. The food and drink dyad
contains six question/answer adjacency pairs. There is no talk (or pre-sequence)
prior to the opening sequence and no insertion sequences Pre-sequences and
insertion sequences are linguistic elements ofnatural conversations that are not
included in the scripted dyad.
Different adjacency pairs occur in spoken discourse. For example:
(1) offer/accept: A) Ill make dinner (2) request/grant: A) Can I sit here?
B) Thanks
B) Yeah, no problem
The adjacency pairs above show that different utterances are likely to be followed
by particular responses. An offer is followed by acceptance and so on. Writers
such as Guy Cook identify further patterns,
In an adjacency pair there is often a choice of two likely responses. A
request is most likely to be followed by an acceptance or a refusal. (1989
p.53)

Linguists refer to the two response choices as preferred and dispreferred.

3.3 Preferred and dispreferred responses
Eggins and Slade (1997) note preferred responses tend to be shorter and
linguistically simpler than dispreferred responses. They suggest that this is
because preferred responses are usually supportive, compliant and orientate
towards closure. Whilst dispreferred responses are generally longer as respondents
are more likely to want to justify or explain themselves. Adjacency pair 1
(appendix 2) illustrates this point about dispreferred responses. When asked what
her favourite food is the respondent explains she is a vegetarian, thus eliminating
meat foods from her thoughts about favourite foods. The preferred response
adjacency pairs 2/4/5 and 6 (appendix 2) also confirms Eggins and Slades
observations; each response is short, supportive and compliant.
McCarthy (1991) observes another characteristic of dispreferred responses is
that they tend to require some sort of linguistic softener. Linguistic softeners are
elements of informal but polite conversations that are usually used to smooth over
potentially embarrassing situations. They help people save face in responses and
are often necessary for speakers not to appear impolite or rude. Cook (1989) also
points out that in an adjacency pair any response that is not a preferred or
dispreferred could be interpreted as rudeness or a lack of attention. Examples of
softeners are in bold text below,
(5) I dont like coffee, its too bitter for me (6) (A) Fancy a drink?
(B) I Cant gotta get back to work
Turns 2/4/ and 6, (appendix 1), show dislikes without any kind of linguistic
softening (or politeness). No attempt has been made to explain the dislikes and
consequently the dyad appears more like an interview than a conversation. It is as

if the speaker is responding to a questionnaire or does not care that the speaker
might like some of the things that s/he dislikes.
In authentic conversation, interruptions and overlaps are part of the
negotiation of turns. Smooth speaker switches are negotiated with no simultaneous
speech or perceptive pauses. This signifies the speakers are negotiating turns
cooperatively, rather than competitively where interruptions would feature more.
Apart from the speaker switch between Selina and Finn (shown in turns (9) and
(10) of appendix 1) where a perceptive pause occurs between mayonnaise and what
the other speaker switches are smooth.
However, the smooth speaker switches intuitively sound odd and unnatural. If
a person likes coffee and somebody says to them, I dont like coffee, a probable
response might be an interruption to find out why the person does not like coffee.
Alternatively a person might seek to defend their position of liking coffee and
interrupt the speaker to hail the benefits of an early morning cuppa. Either way,
for someone to remain quiet and listen attentively to someone listing their likes and
dislikes without any reasons or linguistic softening for their choices appears
unusual.
3.4 Scripted Discourse and Authenticity
Turn-taking signifiers are complex and range from drops in intonation, to the
response expectations inherent in adjacency pairs. Even in casual conversation
speakers have different roles whether they are friends, co-workers or strangers.
These different roles determine how people interact with each other. Guy Cook
illustrates this point saying, students fall silent when the professor speaks - in the
bar as well as the seminar (1989.p53). Without a context a layer of meaning is
removed from the message and the listener may not personalise the language, and
misunderstand its grammatical function.

A context is not provided for the food and drink dyad. The listener is given no
clues as to where the conversation is taking place or what the relationship between
the speakers might be. The lack of context provides an unnatural situation for the
listener. Different roles and settings generate different structures, without any
context or any idea of the relationship between the two speakers an interpreter of
the conversation is left only with the structures which provide few clues to how,
when and with who the language might be used. Even the pace of the conversation
is uniform and in received pronunciation without the variability of authentic
conversation. As Porter and Roberts (1981) note regarding scripted dyads in ELT,
the intonation has sounds like a mother talking to a child with unusually wide and
unusually frequent pitch movement. (p37)

English brings with it its own cultural conventions to turn-taking, shorter

pauses and overlaps are more common in English than in Japanese for example.
McCarthy (1991) notes that silence is more acceptable in Japan and where a short
pause might indicate a point of possible turn transfer in English it may well not in
Japanese. These cultural rules can cause all sorts of cultural misunderstandings.
The inclusion of more authentic elements such as overlaps in scripted
conversations in classroom listening activities can help the learner appreciate the
different conventions English has compared with their native tongue. Backchannel sounds vary from language to language; the Japanese tend to oohh rather
than uh-hu for example.
However, as Guariento and Morley (2001) note with lower level students,
unless the content of authentic material is very carefully selected with due
attention paid to the materials familiarity, predictability and syntactic simplicity,
the use of authentic texts,

may lead them to feel frustrated, confused, and, more importantly,

demotivated. And this would seem to undermine one of the main
reasons for using authentic texts in the first place (p.348)
4. CONCLUTION
Through its focus on turn-taking, this study has shown the scripted discourse
in appendix 1 is unnatural. It has done this by applying CA observations made by
linguists, about turn-taking in natural speech, to the scripted- dyad and discussing the
differences. The analysis of conversation is complex and this paper has merely
touched the surface of one school of thought, namely CA. However, it has shown
authenticity is an issue in scripted dyads used in English learning.

Appendix 1

1. F: Whats your favourite food, Selina?=

2. S = Well, I dont eat meat, so I like lots of salad things: lettuce and tomatoes etc, and
3. vegetables. I like pizza and pasta very much, too, and I eat a lot of cheese. I like
4. potatoes, but I dont eat themoften.=
5. F: And whats your favourite drink?=
6. S: = I like mineral water and cola. I dont like coffee, but I liketea. What about you?
7. Whats your favourite food?.=
8. F: Well, I like chicken but I dont like beef. I like tomatoes and lettuce as well. I like

9. potatoes, especially baked potatoes with tuna andmayonnaise

10. S: What do you like to drink?=
11. F: = Oh, I like tea.=
12. F: Do you know what my favourite food is, though?.=
13. S: No, what?=
14. F: = [p] Chocolate cake!=
15. S: = [f] Oh yes! I like chocolate cake, too.

Appendix 2

REFERENCES
Burns, A. and Coffin,S. (2001) Analysing English in a Global Context: A Reader,
London
& New York: Routledge.
Cook, G. (1989), Discourse, Oxford: Oxford University Press.
Eggins, S. and Slade, D. (1997), Analysing casual conversation, London: Cassell
Ervin-Tripp, S. M. (1979). Children's verbal turn-taking. In Ochs, E. and Schieffelin, B.
(Eds.), Developmental pragmatics. New York: Academic Press, pp. 391-414
Greenhall, S. (1997), Reward Starter Students Book CD1, London: Macmillan
Heinemann
Guariento, G and Morley, J. (2001) Text and task authenticity in the EFL Classroom.
ELT Journal 55/4. Oxford University Press. England
Heritage, J. (1984). Garfinkel and Ethnomethodology. Oxford: Blackwell.
McCarthy, M. (1991). Discourse Analysis for language Teachers, Cambridge:
Cambridge University Press.

Porter, D. and Roberts, J. (1981). Authentic Listening Activities ELT Journal Volume
36/1. Oxford University Press
Sacks, H., Schegloff, E.A. and Jefferson, G. (1974), A simplest systematics for the
organisation of turn-taking in conversation,
Language 50 (4), 696-735.
Tannen, D (1989), Talking Voices, Cambridge: Cambridge University Press
Wardhaugh, R. (2002, 4th ed.), An Introduction to Sociolinguistics, Oxford: Blackwell.