I: Turn-taking system: A review
I.1: Turn Constructional Component
I.2: Turn Allocational Component.
I.3 The signalling view of turn-taking
I.4: Further clarifications on “turn”
II: A Descriptive study of Turn-Taking in Moroccan Arabic
II.1.2: Next-Speaker Selection
II.1.2: Next–Speaker Selection in Dyads
TURN TAKING IN MOROCCAN ARABIC CONVERSATIONS
Sultan MoulaySlimane University
Faculty of Arts and Humanities
In this article, I shall consider the organisational principle of turn taking in Moroccan Arabic conversational materials. Granted that the structure of a conversation is essentially interactional, the turn-taking principle is thus- the most likely determinant of this structure. Considering the fact that conversation, by definition, involves two or more persons, the distribution of talk among the participants is not merely random. It is governed by turn-taking norms which determine who talks, when, and for how long. Therefore, it is the aim of this chapter to examine the different aspects of turn-taking.
I: Turn-taking system: A review
The temporal positioning of items into prior items within sequences and the specific format of items in a conversation give us the impression that conversation is organised in terms of coordinated actions. Participants accomplish their interactional business through the “taking” and “leaving of the floor” which implies that turn-taking is a major determinant principle of conversational organisation. It has been shown by Sacks et al(.1974) that conversations are organised on the basis that speaker take turns, and turns are allocated according to a rule that says only one speaker at time has the right to talk. One of the reasons for studying turn taking in Conversation Analysis, as a “machinery” which is at the same time context free and context sensitive, is to find one way of approaching the problem of how such “models of speaking” are constructed. ( See Sacks, Schegloff and Jefferson (1974)).
The norm of reciprocity, according to Roloff and Campion (1985:174), is defined as the shared expectation that the recipient of a resource is obligated to, and at same time will return to the giver a resource roughly equivalent to that which was received. It follows that the study of the conversational management per se requires different methods and modes of analysis. Conversational interaction is viewed as a complex social activity which entails expectation about possible goals and outcomes about relevant aspects of interpersonal relations and about what count as normal behaviour (Gumperz (1982)).
The latent implications of Gumperz’ claim induces us further to argue that consideration of “politeness phenomena” (Brown and Levinson 1978) and the notion of “face” (Goffman (1967)) are likely to be among the determining factors in rendering a conversational activity as goal-oriented. Schegloff (1972) notices that the sequencing in two-party conversation is alternating, which means in Sacks et al (1974) terms, a motivated basis for the “status of listenership which is independent of the whims and the passions of the participants. However it should be stressed, as pointed in Benson and Hughes (1983:177) that politeness and the face work will have to be displayed using the machinery which organizes the structure of conversation. The strategic conversational play by participants in conversation seems to be also determined by power, social distance and rank imposition (See Brown and Levinson (1978), Grimshaw (1981), Downes (1984), Bull and Roger (1989)).
Among the organisational principles oriented to bythe participants through this strategic play is the “turn-taking”. The nonfixedness of the conversational role (i,e a speaker is not speaker all the time, and neither is the auditor) is a clear evidence of the existence of such a principle. What is more, it is clear evidence that talk in general and conversation in particular is accomplished through a series of turns (for similar claim see S.S.J. (1974-1987), McLaughlin (1984:92)) as a consequence of the coordinated enterprise in which participants are engaged.
Turn-taking, Goffman (1972: 65) explains, refers to the mechanism that generates the “taking and leaving of the floor” in a conversation, which is sort of “intimate collaboration where one turn at talk neither overlaps the previous nor wants for inoffensive conversational supply. For Power and Martello (1986:29), “turn taking” is “the coordination of the activities of speaking and listening in conversation so that at any given moment there is only one speaker”. This systematic view of the succession of participants at holding the floor, in a conversation, can be better understood in the light of the findings of Sacks et al (1974-1978) Schegloff and Sacks (1973), and Schegloff (1972-1982).
According to Sacks et al (1974-1978:7-8), the existence of turns implies that conversation is an economic system, and turns are goods, possession of which entails certain costs and rewards or rights and obligations. The manner, therefore, in which turns are taken in conversation, plays a potentially important role in the development and maintenance of our relationships throughout our lives. The mechanism by which we accomplish turn –transition, however, is not completely understood, though it is this mechanism which gives talk in conversation a sequential patterning. As Cappella (1983) and Wiemann (1985) stress, the system of turn-taking built into the nature of conversation; that is a product of the structure of conversation.
Further, turn taking as formulated by Sacks et al (1974) is a “Local management system” based on a turn-by-turn procedure. Such a system however, is not a system as such but a mechanism that that governs the turn-taking process( i,e change of speaker) and accounts for the properties noted as a set of rules with ordered options which operate on a turn-by-turn basis (see Levinson 1983: 297) it is a mechanism for achieving coordination of the participants’ actions. It appears, thus, that the sequential production of conversation is the mutual responsibility of the participants and allows conversation to proceed in a step-wise fashion.
The system that Sacks et al (1974) describe has two components: a turn-conversational component and a turn-allocational component with a set of rules.
I.1: Turn Constructional Component
Sacks et al suggest that a speaker can use various, syntactically defined, unit types to construct a turn - these may include sentential, clausal, phrasal and lexical constructions. Instances of the unit types allow a projection of the unit type under, and what it will roughly for the unit type to be complicated. Sacks et al state that: “as for the unit-types which a speaker employs in starting the conversation of a turn’s talk, the speaker is initially entitled, in having a “turn”, to one such unit. The first possible completion of a first such unit constitutes an initial Transition Relevance Place (henceforth TRP). Transfer of speakership is coordinated by reference to such transition relevance places, which any unit-type instance will reach” (1974:703). The recognition of possible completion points is licensed then, by the concept of projectability.
In spite of its strength and comprehensiveness, the S.S.J.’s model has one major shortcoming: that is the vagueness in the description of the feature which defines a transition-relevance place (see Sacks et al p.721). In this model, each speaker is allowed one turn-constructional unit whose end is called transition-relevance place, at which point a change of speaker may, but need not, take place. However, these points at which transition of speaker may occur are not always taken as intended opportunities for such transition. A speaker may extend his turn to more than one unit-type for interactional purposes. In attempt to remedy the model, Sacks et al, realizing that speakers do not see all unit-type endings as turn endings, note that interactional features and the sequential or social context my disambiguate the status of unit-types ending. Still, the problem is to state what counts as a possibly complete utterance.
In this connection, Power and Martello (1986) claim that the S.S.J.’S model is practically unclear about T.R.P; their proof is the issues’ raised by the “possible completion” of utterances:
A. Is there a clear cut distinction between complete and incomplete utterances, or are there merely varying degrees of completion?
B. Do judgements of utterance completion depend on the utterance itself, or on the nonverbal behaviour of the speaker, or both?
C. Which features of an utterance do judgements of completion depend upon: prosodic, syntactic, semantic, or pragmatic?
Though one may be tempted to answer these questions, yet the answers seem to be far from agreed upon because of the lack of adequate descriptive machinery which takes into account the different components of linguistic actions. The syntactic criterion for completion seems to be inadequate as a final solution. In his experimental study of dyadic turn-taking behaviour, Roger (1989:76) reports that the judgement of completeness and incompleteness is based largely on grammatical criteria and that in ordinary conversation grammatical rules are not necessarily observed.
Given that conversational interaction is communication of meanings, Power and Martello (1986:33) suggest that “the speaker’s message interaction may be used as one criterion of utterance completion, instead of speculating about how such a message is of pragmatic nature (i,e an illocutionary force” (Austin (1962))).
Moreover, the incidence of overlap and silence witnessed in the data (see the Appendix) is a clear evidence of the structural intricacy of completion points. Considering interruption, one can say that this behaviour stems sometimes from the participant’s mistaking the end of constructional units (i,e T.R.P.s). S.S.J. report in their study that instance of overlap derives sometimes-from the projectability of possible completion of the transition-relevance places.
1- P: Yeh alright dear 2- A: what’s your name again, please
J: okayB: F.T. Galloway
3- A: well if you knew my argument who did you bother to a:sk
B: because I’d like to defend my argument.
In the first case, the overlap is caused by J’s mistaking the T.RP, for P could have stopped after the second word of his utterance. S.S.J explain that in such a case “address” terms or etiquette terms may be productive of structural overlaps. As for the second case, S.S.J. attributes the overlap incidence to “variation in the articulation of the projected last part of a projectably last component of a turn’s talk, which is … a consequential locus of articulatory variation” (P 16).
In fact, the recognisability of the units (out of which conversation is made) as either possibly complete or incomplete and the change of speakers at possible completion are situationally coped with (see Coulthard (1977:70)). The argument may go further to stress that even in cases where speaker-selection is explicit, it does not mean that speakers will change at that point but simply they may do so (see Levinson 1983:297-8), Grimshaw (1983:170)).
Labov and Fanshel (1977) claim that in any overall-view, it is obvious that actions are more important than utterances, since it is actions that have consequences and affects others’ lives and that action is what is intended in that it expresses how the speaker means to affect the listener to move him, to cause him to respond. Considering the functional aspect of the utterances performed in a conversation, one is bound to claim that the turn taking system is functionally motivated since speaker-change can be shown to depend on the type of moves being made (see Owen (1983)). For example, speaker-change should occur immediately after a question, whereas following an assessment or evaluation, the requirement for a response still operates but is less powerful.
Sacks at el (1974) further the impression of functional motivation by their assertion that “all turns have retrospective, a here-and-now, and a prospective aspect”, and further by their recourse to the adjacency pair as a device for implementing speaker selection. Functional claims, however, are not to be included in C.A of turn-taking not in the realm of discourse which, according to Schegloff (1982), should be treated as a structural achievement. He points out, “there is real, recurrent contingency concerning” who should talk now, the fact that someone continues is an outcome co-ordinately achieved out of that contingency” (p 89).
I.2: Turn Allocational Component.
The second component of the turn-taking process is the turn-allocation component; that is, the selection of the next speaker. It involves two rules which organize speaker change, reported herein in a simplified version (C: current speaker; N: next speaker)
Rule 1: applies initially at the first T.R.P. of any turn.
A) -if C selected N in current turn, then C must stop speaking, and N must speak next, transition occurring at the first T.R.P. after N. selection.
B) - if C does not select N, then any (other) party may self- select, first speaker gaining rights to the next turn.
C) - if C has not selected N, and no other party self selects under option (B), then C may ( but need not) continue.
Rule2: applies at all subsequent T.R.Ps when Rule 1 has been applied by C, then at the next T.R.P. rules 1 (A, C) apply, and recursively at the next TRP, until speaker change is affected.
The problem of overlaps and interruption mentioned earlier (the preceding section) can be seen, then, as a capital consequence of the violation of the rules. This system is, in fact, intended to “minimize gape and overlap” (S.S.J.( 1978:12)) because it is organised around transition relevance places which, S.S.J. claim, can be projected and are thus predictable. Two major advantageous features of the system can be detected. First, one speaker will generally be speaking at any time in a single conversation regardless of the number of potential next speakers (See Levinson (1983:300)). Second, the set of rules is cyclical in its application: it reapplies at each transition relevance place for each speaking turn (Houtkoop and Mazeland (1985:596), see also S.S.J. (1978:15)).
It would be a relevant fact to point out that the S.S.J. rules can be summarized in two points: those in which the current speaker selects next speaker and also the next speaker self-selects. The selection of the next speaker may be achieved by the use of various techniques (e.g. form of address, tag questions, request-grant, and special adjacency pairs which require from next speaker to produce an answer). The following transcripts illustrate the point:
Sara: Bill, you want some ?
Sharon: you didn’t come tuh talk tuh Karen?
Mark: No, Karen –Karen I’ve having a fight (4) after she went out with Keith an’ not with (me).
A: I’m glad I have you for a friend.
B: that’s because you don’t have any others.
A: It is not break time yet.
B: I finished my box, so shut up.
However, the turn-allocation component as proposed by S.S.J. suggests that the transfer of speakership is coordinated by reference to projectable transition-relevance places that arise at each possible completion point of the turn-units used in speaker turns. This coordination is methodologically unclear given that turn-taking is organised around T.R.Ps, the recognisability of which, if based on syntactic criteria, seems problematic in cases of longer structural unit-types (e.g. “stories” (Jefferson (1978), also Polanyi (1985), Sacks (1972), Greatbach (1988).
Equally important is the problem of obligation to speak next (Rule 1(a)). Power and Martello (1986) argue that the S.S.J’s claim is false on the ground that the selected next speaker my not feel any obligation to take a turn. For instance, the following example (supplied by the authors)) suggest that rule (1(a)) it not operative.
John: do you have the time, Mary?
Bill: where are you going this summer, Mary?
According to S.S.J, Bill’s utterance would be judged inappropriate, though sequences of this kind frequently occur in conversation. Edmondson (1981:41)) argues that turn-taking procedures are subject to the control of the speaker and/or hearer, such that turn-selection or assignment is different from turn-taking. This means that one cannot predicate at any one point in time that a change of speaker role will occur, though one may well be able to distinguish on the basis of what is said, how it is said and the concomitant behaviours such as eye movement and body shift between, for example, different types of silence (See S.S.J (1974:715) and Edmondson (1982:41)).
In summary, the suggested mechanism organising turn-taking is not quite plausible although the phenomenon of the turn-taking is obvious as argued by Levinson (1983). He reports that psychologists think turn-taking is regulated primarily by single and not only by structural opportunity rules as suggested by Sacks et al (1974: 1978). I shall mention briefly in the following section the signalling view of turn-taking and some problems related to turn definition.
I.3 The signalling view of turn-taking
Surely there can be no ready-made solution for turn-taking and no claim will be made to propose cut-and-dried solutions. But in view of seriousness of such an issue, it is evident that speakers use signals for indicating the appropriateness of a change in speaker.
The works of Duncan and Fiske (1977:1985) dealt explicitly and specifically with the signals which cue speaker-auditor changes as well as the within turn cuing process (i,e back channels)). It is assumed that participants manage turn-exchanges by the display of / and response to cues displayed by other interactants. These signals are composed of behaviours which are discrete and independent of one another. Speaker turn signals include: intonation pattern, a sociocentricsequency, completion of grammatical clause, drawl on a final syllable, and termination of gesticulation or relaxation of the hand (See Duncan and Fiske (1977: 169)).(For a comprehensive review of the signalling approach, see also Wiemann (1985:89-91), Cappella (1985:195-199), also Ellis and Beattie (1986:181-195)).
Argyle (1983) argues that a signal (such as gaze) is of central importance in human social behaviours. It acts as non verbal signal showing, for example, the direction of the gazer’s attention. At the same time it opens a channel, so that another person’s non verbal signals can be received. “Gaze is a signal for the person looked at, but it provides a channel for the person doing the looking “(Argyle (1983:80)). This signalling view seems to suggest that non verbal behaviour as culturally functional component in social interaction is the essential organisational basis for turn-taking. In fact, the way in which the organisation of turn-taking may interrelate with other specific cultural systems (such as culturally significant signals) are not quite clear. According to Schegloff (1982), the methodological impossibility of obtaining illustrative data is a reasonable motive for refuting the signalling approach as a non-practical step in the descriptive process of conversational turn-taking (p77-79).
To know when to take or leave the floor is something that belongs to the communicative competence of speakers. Linguistic, sociocultural and contextual knowledge is basic to this competence. In fact, communicative competence “ extends to both knowledge and expectation of who may or may not speak in certain setting, when to speak and when to remain silent, whom one my speak to, how one may talk to persons of status and roles, what appropriate non-verbal behaviours are in various contexts, what the routines for turn-taking are in conversation …” (Saville-troike (1982:22)), (See also Gumperz (1987) for a full discussion of the relationship between a theory of communicative competence and social interaction process as a part of the communicative competence of native speakers, the know-how and know-when to take turns are implicitly performed in the course of conversational activities. Speakers are constantly signalling their intentions to the partners through signals of different kinds (See Keller 1979:227).
The signalling behaviours that occur near or at points of speaker change seem to have also relational implications. For Wiemann (1985), Duncan and Fiske’s turn system represents one way in which regulation of conversation can be extended to describe how control is enacted at the relational level. In terms of control as manifested in the regulation of turns, Wiemann points out, the signalling model pictures interactants as competent to the extent that they process signals appropriately- both the ones displayed by their interlocutors and the ones they might display themselves.