Turn taking mechanisms in conversation.
Turn taking mechanisms in conversation.
From the amount of conversations we witness on a daily basis we can see that they are governed by some sort of mechanism or rules. From these observations, it becomes clear that turn taking is a major constituent of conversation, with the arrangement of talk across two participants. Levinson (1983: 296) explains that, despite the 'obvious' nature of turn taking (i.e. A speaks, then B speaks, then A speaks again) the way in which distribution is achieved is "Anything but obvious". He states that "Less (and often considerably less) then 5 per cent of the speech stream is delivered in overlap, yet gaps between one person speaking and another starting are frequently measurable in just a few micro-seconds". This phenomenon is of interest to pragmaticians who, through the practise of conversational analysis have studied conversation on the micro-pragmatic level and have sought to theorise the mechanisms responsible.
In order to study the turn taking system operating in conversation I transcribed three brief conversations from Big Brother 2 (Appendix). Big Brother is a popular 'reality T.V' game show where contestants are invited to live in a house for up to 8 weeks where they are constantly monitored and filmed. The public evicts each week one contestant, with the winner being the last contestant left. I decided to use conversation from Big Brother for several reasons. Firstly, the conversation was easily accessible and could be replayed repeatedly to study the conversation in detail. Another advantage was that I was able to see facial expressions and body language of the participants. Knowing the context in which conversation occurred and the participant's reactions enabled me to give a more accurate analysis. One major disadvantage is that 'natural conversation' may have been effected as a result of the program being broadcast to a national audience. Participants must have been aware of the presence of the cameras and may have altered their speech accordingly. However, the observers' paradox would occur in any other situation where participants knew they were being recorded so I did not consider this to be a major problem.
I thought that it would be of interest to use the transcripts to study the Turn-taking system operating in the speech of the participants and to observe how the participants are involved in sequencing, or in some cases how they fail to correctly observe the TRP of other participants I decided to analyse the conversation with reference to previous analysis completed in this field of pragmatics.
Sacks, Schegloff and Jefferson (1974: 700) proposed a model for the organisation of turn taking in conversation. They identified how turns in conversation were systematic and observed the following rules applicable to the conversation system.
) Speaker change recurs, or at least occurs
2) Overwhelmingly, one party talks at a time
3) Occurrences of more than one speaker at a time are common, but brief
4) Transitions (from one turn to a next) with no gap and no overlap are common. Together with transitions characterized by slight gap or slight overlap, they make up the vast majority of transitions
5) Turn order is not fixed but varies
6) Turn size is not fixed but varies
7) Length of conversation is not specified in advance
8) What parties say is not specified in advance
9) ...
This is a preview of the whole essay
2) Overwhelmingly, one party talks at a time
3) Occurrences of more than one speaker at a time are common, but brief
4) Transitions (from one turn to a next) with no gap and no overlap are common. Together with transitions characterized by slight gap or slight overlap, they make up the vast majority of transitions
5) Turn order is not fixed but varies
6) Turn size is not fixed but varies
7) Length of conversation is not specified in advance
8) What parties say is not specified in advance
9) Relative distribution of turns is not specified in advance
0) Number of parties can vary
1) Talk can be continuous or discontinuous
These rules can be seen as a 'local management system' in that they produce an allocational system operating on a turn-by-turn basis. Levinson (1983: 297) suggests that the turn-taking system can be viewed as an economy, where the control of the 'floor' in conversation constitutes the scarce resources. 'Turns' consist of syntactic units, which range from one-word utterances to complex sentences. They are often identified as 'turn units' through prosodic and intonational means and, according to Sacks, Schegloff and Jefferson can be assigned through three main processes.
They distributed turn allocation techniques into two groups; selection by the current speaker or self-selection. They devised rules to govern turn construction. These rules provided for the allocation of the next turn to a participant and coordinated the transfers in order to minimize gaps and overlaps in conversation. It was recognised that transfers from one speaker to the next occurs at what they call the Transitional Relevant Point (or TRP). Current speaker selects next techniques may be used at the beginning of the turn, for example by addressing the next speaker by name or the use of gaze. If this occurs the participant is obliged to take his turn but the transfer does not occur until the TRP.
In transcript one N selects the next speaker with a question:
N: What kind of music do you like?
If another speaker is not selected self-selection may take place at the TRP, with the first starter acquiring a right to a turn. A further example from transcript one shows how this mechanism works, where N self selects after B has completed his turn:
B: Abba are superb
N: What's your favourite Abba song?
If nobody self selects then the current speaker may continue his turn until he selects another speaker, or another speaker self selects at the TRP.
Simultaneous starts can occur when one or more speakers attempt to self-select at the same time, but turn is brief for later self-selecting speakers. These issues of overlap will be reviewed later in this essay.
The selection of further speakers can follow reasonably basic patternings such as questioning, with a term of address of gaze cast at the next speaker selected. Yet, the majority of transitions do not occur in this way. Techniques for selecting next speakers can be complex and in this next section I have attempted to use the data to identify how TRPs are recognised by speakers.
In transcript one the participants are questioning each other about their tastes in music. The conversation begins with N asking B 'what kind of music do you like?' This is an example of a 'speaker selects next' mechanism as N signifies that B has been selected through looking at him as she addresses him. This is the first TRP of the conversation, yet when B fails to respond immediately N decides to continue her turn until B decides to answer. .
In studying the transcripts I identified a variety of ways in which a TRP can be signified. Completed syntactic units, such as questioning with a rising intonation often signals that a TRP is approaching. Tag questions such as 'you know';
P: It's alright yknow (.) there's piles on the fire
Are used as an exit technique at TRP of a turn, signalling a turn is over and prompting other participants to take a turn.
Until this point my analysis has concentrated mainly on the mechanisms operating when two speakers are engaged in conversation. However, the mechanism involved in turn taking is also capable of operating in circumstances with varying numbers of participants in a conversation without a specific order of speakers being pre-allocated. Despite this, the numbers of participants in a conversation affects transitions. "In a two party conversation, a current non-speaker can pass any given TRP which is non-obligatory (i.e. where 'current selects next' technique has not been used) with full assurance of being 'next speaker' at some point " (Sacks, Schegloff and Jefferson 1974: 712). The situation changes when there are more than two participants, as current non-speakers must self-select and will therefore be under constraint to self select at the first possible TRP in order to secure their turn.
Where overlaps do occur they appear to do so for a number of reasons. Overlaps may occur where TRPs have been mis-projected or can be used as an intentional interruption. Most overlaps are examples of the former and are brief in nature. Levinson holds this accountable to a "resolution system" (1983: 300) integrated in the turn taking mechanism, ensuring that if more than one speaker has selected themselves one of the speakers (usually the last speaker) drops out of the conversation. In transcription two S overlaps P:
P: Yeah [I'm alright
S: [Good
P: Yeah I'm fine.
It is interesting to note that once the speech of P is cleared of overlap from S the speech that was obscured by S's overlap is repeated (not exactly, but in semantic terms). Transcript one shows how prosodic features can be used to prolong a turn or 'win the floor'. D uses the pre-sequence:
D: I'LL TELL YOU WHAT
Along with increased amplitude to 'drown out' P's speech. Once securing the floor he brings the conversation back to the line of questioning that was being uttered before P interrupted.
Deborah Tannen (1990) identifies a number of different types of overlap in informal speech. Some of the overlaps occurring are an example of back channel, which shows support towards the current speakers turn:
B: Paul Weller (.) Beatles (.) [Jam (.) Donna (.) Carpenters (.) Abba (.)
D: [Cool (2.0) Carpenters wicked mate
As well as interruptions:
BB: You have failed to successfully complete this [week's task
D: [whatever
N: [what a surprise
And pre-empting:
S: Yeah I'm [knackered too
PA: [absolutely knackered
Conversation is natural and therefore the turns are not pre-specified as occurs in other speech exchange systems such as interviews or court cases where turns are often pre-allocated by formula. Although what participants say is not specified in advance but it must be to some extent projectable of the end of each turn to account for the quick transition from speaker to speaker. Where these turns are easily predicted in advance, such as greetings like Good Morning we would expect an answer like 'Good morning' and not a detailed analysis of the weather. These utterances that require such constrained replies are often termed adjacency pairs. In analysing transcript two I found a number of adjacency pairs operating.
The conversation starts with the adjacency pair greeting greeting:
S: Morning=
P: Morning
The following utterances I found amusing. If I were to categorise them I would interpret them as insult-retort. As a result of their disagreement on the previous evening both S and P are on edge with each other. S begins by implying that P is not receiving much help from the rest of the group. P appears to interpret this as an insult, perhaps suggesting nobody wants to help her and the group therefore dislikes her. P retorts by saying 'great hair' when it is obvious that S is having a particularly bad hair day:
S: You've got a lot of help
P: Great hair
I also noted a number of adjacency pairs operating in the transcripts:
N: What kind of music do you like?
B: Paul Weller (.) Beatles ....
This is an example of a question and answer adjacency pair. Another adjacency pair occurs further down the passage with the assessment - agreement pair between N and B
N: Oh I love Abba
B: Abba are superb
Examples of greeting-greeting adjacency pairs are evident:
S: Morning =
P: Morning
S: How you doing (.) alright?
P: Yeah I'm alright
In transcript 2 there appears to be one adjacency pair embedded within another. These take the forms of question - answer
PA: Alright mate (.) how's it going? (Question 1)
S: Alright? (Question 2)
PA: Knackered (Answer 2)
S: Yeah I'm knackered too (Answer 1)
and are termed by Levinson as 'insertion sequences' (1983: 304)
The participants in conversation must listen to and interpret the adjacency pairs to understand if he is being selected as the next speaker and to ensure that his turn is correct. Levinson (1983: 306) explains how the notion of 'conditional relevance' is highly important in adjacency pairs, in that given the first part of adjacency pair the second part is expected. In cases where the second part is nonexistent it is generally prominently absent and usually provokes a repetition of the first pair part. Yet, a problem with this notion of 'relevance' is the amount of potential responses in the second pair. Mey (2001: 151) shows that not all second pair parts are of equal structural complexity, with some being extremely simple (such as 'yes' or OK) whereas others show long build-ups of pauses and explanations. Simple responses can be termed as unmarked or preferred turns, whereas those including delays or hedges are termed as marked or dispreferred turns.
Transcript 3 shows an interesting example of preferred and disprefered turns. Speaker H addresses the group with an assessment, or an admittance of blame:
H: We've failed it and it's all my fault
Levinson (1983) explains how the preferred response, would be something syntactically simple, such as 'yes' whereas the disprefered response would be more lengthy. The opposite, is however true in this case of self-denigration. The preferred response would be an agreement, yet by using this the other speakers would be criticising H. To avoid this the other speakers must give the dispreferred response of disagreement. In this particular conversation the other speakers do not respond, the silence leading H to propose the issue again
H:It's all my fault.
Yet the silence she receives a second time, where B instead of administering her remark announces his disappointment to the rest of the group. Here the silence that H is exposed to appears to signify agreement with her assessment.
To conclude, I have used the transcriptions collected to analyse the system of turn taking in respect of Sacks, Schegloff and Jefferson's theories (1974). I found their rules for the turn taking mechanism to be relevant to the conversations I analysed with sufficient evidence to support their rules. I also identified other points of interest within the transcripts and attempted to account for these using the theories of Mey (2001), Levinson (1983) and Tannen (1990).
Bibliography
Levinson, S.C. (1983) Pragmatics. Cambridge. CUP.
Mey. J (2001) Pragmatics: An introduction 2nd edition. Oxford. Blackwell.
Sacks, H. E.A. Schegloff and G. Jefferson (1974) A simplest systematics for the organization of Turn taking for conversation. Language 50; 696 - 735.
Tannen, D. (1990) You just don't understand: men and women in conversation. London: Virago
Word Count: 2,400