Language Learning Across Boundaries - Negotiating Classroom RitualsRoger Nunn
When opposed to communicative language teaching, teacher-fronted classroom discourse is sometimes devalued, represented as a rigid and ritualistic process of eliciting pre-planned responses. The analysis of a data sample of 12 teacher-fronted language lessons revealed highly repetitive structuring in the flow of discourse, frequent verbatim repetition in the collective reconstruction of texts, and restrictive turn-taking norms. However these "ritual" features of discourse need to be considered alongside the essential negotiation which takes place during the continuous process of adjustment and re-adjustment between participants. This paper will use discourse analysis to redefine the relationship between ritual and negotiation in "lock-step" teaching in the light of research findings and recent re-evaluation of the notion of ritual in educational settings.
Teacher-fronted classroom interaction, while frequently opposed to the more devolved interaction of communicative teaching, has remained the backbone of much institutional language learning worldwide. The way we represent classroom discourse is crucial to the direction of classroom innovation. This paper will argue that the models commonly used to describe teacher-fronted classroom interaction have a built-in assumption that such discourse is outmoded, rigidly structured and ritualistic and that this assumption neither reflects reality nor assists curriculum development. This discussion will confront abstract, generalized knowledge of exchange structure with findings rooted in the classroom data from a specific context.
In the 1970s, a large scale textbook and teacher training project was carried out in several Gulf States with the aim of reforming English language teaching classroom methodology. The Teacher Training Guidelines for the Crescent English Course characterized the discourse pattern in local classrooms as follows:
The questions are repeated and so are the answers but nothing happens as a result. It must seem a strange ritual for the learner for it bears little resemblance to the way in which he uses his own language. (Bratten, 1978, p. 1-d,) [-1-]
Sample one below is an example of this kind of discourse. Sinclair and Coulthard (1975) call this kind of exchange an "elicit exchange." Their representation of a three-part exchange structure --"initiation," "response," and "follow up," or IRF-- is commonly cited.
The negative characterization of this kind of teacher-fronted classroom discourse, based on classroom observation in a specific context, is echoed in more recent theoretical discussion of the structure of classroom discourse. Andersen et al. (1999, p. 372) refer to teacher-fronted discourse as "classroom ritual" which is "antithetical to classroom discussion". McDonough and Shaw (1993) refer to "breaking" the "traditional lock-step of the classroom" (p. 243), representing the notion of lock-step teaching as a three-part structure, a "traditional pattern of teacher question-student answer-teacher comment" (p. 231). McCarthy also refers to a "rigid" three-part structure, in the "traditional classroom, where roles are rigidly defined and patterns of initiation, response and follow-up in exchanges are relatively easy to perceive . . ." (1991, p. 19).
This paper argues that representing teacher-fronted discourse as a predominantly three-part structure is unjustified. This is important because the choices available to participants and the way participants distribute and use these theoretically available choices determine the roles of teachers and students. The way we represent interactive classroom roles adopted by participants is closely linked to the way we represent the teaching method actually being enacted.
In a previous project, many aspects of discourse were analysed, including turn-taking, topic control, use of the textbook, different levels of discourse, etc. (cf. Nunn 1996a, 1996b, 1999). This paper will consider only the Sinclair and Coulthard (1975) IRF model and only at the level of exchange structure.
The theoretical model will be confronted with data findings from a project in Qatari Boys Secondary School English Language classes. The conclusion of the investigation was that, in spite of a ten-year teacher-training programme aimed at introducing so-called communicative methodology, there was little evidence of communicative teaching in the classroom. Indeed, all the discourse recorded or observed 15 years after the start of the innovation was teacher-fronted classroom discourse of the kind that had been criticized as "meaningless ritual" at the start of the project.
This paper has identified terms such as "rigid" or "meaningless" as being commonly associated with the so-called "ritual" structuring of teacher-fronted discourse. The concept of "ritual" is used to impose a negative connotation on the classroom discourse of others. Rather than seeing all regularly performed institutional events as examples of ritual, including those we ourselves are commonly involved in, we prefer to see "ritual" as external, performed by others but never by ourselves. McClaren (1999, pp. 36-50), however, in his detailed discussion defining ritual, demystifies the notion, stating that "rituals are constitutive of everyday human life, including secular activities" (p. 36) adding that "rituals thrive in the world of lived experience" (p. 38) and are "integrally related to everyday action." To McClaren, even activities such as eating at McDonald's are ritual events. He suggests that "to engage in ritual is, for men and women, a human necessity." (p. 36) Rituals for McClaren are not limited to formal settings. He suggests that "rituals tend to sprout anywhere people gather in groups" (p. 39) and that in different contexts, rituals "transform participants into different social statuses." (p. 38)
McClaren also rejects the assumption that ritual "is only--or mainly--a bounded circumscribed and somewhat frozen act." (p. 46) He maintains that rituals "may oscillate between randomness and formality" (p. 40) and again that a ritual is a "series of encoded movements that must oscillate between exclusive randomness . . . and rigid structure." (p. 42) Schiffer, (1997, pp. 152-153) also challenges the devaluing of the concept of ritual in terms of so-called rigidity. Schiffer distinguishes between "ritual identity" and rigid attitudes towards ritual performance. "Identity is a matter of definitive specification of the rite itself; rigidity concerns rather the attitude taken towards the rite as defined." (p. 153) Schiffer considers it "especially important to separate the formality of a ritual . . . from the attitude of compulsive adherence to its performance." (p. 153) In terms of the classroom, the relationship between performance and formal structure is central. [-2-]
This paper will exploit McClaren's view of a ritual as a "framing device" which "embodies a repertoire of choices." (p. 48) This allows for more relativity than the view of ritual as a rigidly limited and highly restricted code. This view is reflected in Carbaugh (1989, p. 105) who refers to the "relative degree of structuring of the code," expressing the degree of structuring in terms of clines from "fixed to flexible" or "restricted to elaborate." Griffin and Mehan (1981, p. 205) make a slightly different distinction in terms of classroom discourse referring to "spontaneous improvisations on basic patterns of interaction." When the pattern of teacher-fronted classroom discourse is represented as an inflexible, three-part structure, the model itself creates an assumption of inflexibility rather than of spontaneity and impedes adequate description.
In considering ritual in conversation, Goffman (1983, p. 17) does not propose a rigidly limited structure, stating that "ritual considerations help to produce many naturally bounded interchanges that have, for example, three or four parts, not merely two." The way "choice" is integrated into formal representations of structure will be central to this discussion.  The "exchange" in Sinclair and Coulthard's rank-scale model is the most immediate and fundamental point of reciprocal interaction. It is the smallest essentially interactive unit of analysis at which interlocutors adapt to each others' immediate contributions. It is revealing to examine how we represent choice at this level of analysis . Any characterization is necessarily a simplification of a very complex phenomenon, but it is still important to see how the immediate give and take of interaction is represented.
The developing literature on the 1975 model has presented many elaborations that represent the IRF exchange as a more complex structure. The present proposal, drawing on Coulthard and Brazil (1981), Sinclair (1992) and Hoey (1993), emphasizes the way the developing exchange can be represented as providing an interactive choice. In this way, interlocutors in teacher-fronted classroom discourse are shown to be involved in a process that leads to a negotiated outcome. However, the model does not represent interactive structure as being without sequential constraints or without limitations on divergence. Proposing the interactive choice of a negotiating element "N" simply reflects essential possibilities that are available to participants.a. Choices after an "R."
When a speaker provides a response, this response may terminate the unit. Alternatively, the choice of maintaining the negotiation open by providing an "N" is available. The response could also be followed up by just one further element "T" which would then, by definition, terminate the exchange. These options are illustrated in Figure 2 above. They represent a common pattern in the data corpus. (It may be noted that the sample above is referred to as a "unit" of interactive structure. I prefer to consider it an "exchange", but am aware that others may see it as a complex of bound exchanges. The point is that it is difficult not to see it as a single coherent unit.) [-3-]
While the "N" may be seen as a bound initiation (Sinclair & Coulthard, 1975), or an "R/I" element (Coulthard & Brazil, 1981), its sequential position--central in one coherent sequential unit--makes it misleading to represent it as an "initiating" element. The "N" in the sample above is intimately linked to preceding elements in one coherent unit which it "encapsulates," simultaneously maintaining the unit of interaction open. The choice of negotiation is structurally central within a unit that has already been initiated. Whenever the option of using an "N" after a response is not taken up, the unit ends. This means that an "elicit exchange" can terminate in two ways; firstly, with an "R" when a response is not followed up at all, or secondly, with a "T."
After the initiation, there is always a choice until the "T" has been accepted as such. At the rank of exchange structure, one of these choices is always an "N." These choices are available at each point within an exchange at which there is a change of speaker. The choices are as follows:b. Choices after an "I."
Similarly, whenever a speaker produces a negotiation, the next speaker seems to have the option of further negotiation. Both options are illustrated in Figure 4 below.c. Choices after an "N" itself.
The "N" element illustrated in the four samples above can be defined in terms of what Sinclair (1992, p.86) refers to as the two main "mechanisms of coherence in discourse structure": "encapsulation" and "prospection." Encapsulation involves the retrieval of all previous contributions within a coherent unit. Any element of structure that makes a retrospective reference encompassing the preceding contributions within the unit is said to encapsulate them. When an element is exclusively encapsulating, it terminates the unit. "Prospection" is a feature of discourse by means of which a speaker may attempt to manage or influence the future direction of the discourse. Each contribution to the discourse is then said to provide a framework for the ensuing discourse. In an exchange, the initiation creates prospections that influence what Sinclair calls the "minimum extent of the exchange" (Sinclair 1992, p. 84). This does not rule out the possibility of participants using subsequent elements to influence the prospection set up by the initiation. For example, the "N" can be used to modify or even to take over the prospection. An "N" not only maintains or modifies the prospection set up by the initiation, but also encapsulates the preceding discourse contributions within the exchange. [-4-]
Four criteria of structural efficiency (based on Sinclair & Coulthard,1975, p. 15/1; Coulthard,1977, pp.98-99; and Sinclair, 1992) have been applied to the model: sequential position, encapsulation, prospection, and the obligatory or optional nature of an element. Applying these criteria, all elements can be clearly distinguished from each other and impossible combinations of elements can easily be found. All participants in teacher-fronted classroom discourse always have an interactive choice, unless social restrictions intervene to prevent certain participants from making use of available choices.
In the Qatari data corpus, conclusive evidence that students almost exclusively produced only response elements is presented. In this classroom context, an "N" was normally an option only available to the teacher for some pedagogical purpose such as negotiating for a more satisfactory response. In another institutional context (a language school), students frequently produced negotiations, but almost never produced initiations.
It should be emphasized that a variety of models of analysis are needed to describe any data corpus. This very brief summary of the exchange can only give an indicative flavour of the data. It is just one essential part of a baseline description at one level of analysis. (See Chapter 4 of Nunn 1996b for a detailed review .) From 877 exchanges analysed in eleven lessons (entire class periods), only 53% (467) of the exchanges were easily identifiable as three-element units. More than one student was always potentially involved in an exchange. The average percentage of three-element exchanges per lesson was 50%. Three-element exchanges ranged from a minimum average of 29% to a maximum of 81% of the total number of exchanges, although 7 lessons were very close to the average. 8% (73) of the total were two-element exchanges. Two-element exchanges ranged from 0% to 22% of the total number of exchanges.
38% (337) of the total were coherent units of four elements or more. These "exchanges" contained at least one negotiating element. The average was 40% per lesson. If we look at the number of elements involved, we must conclude that the units involving negotiation actually account for a greater proportion of the data than two and three-element units in terms of contributions to the discourse (62%). (On average these longer units contained seven elements.)
The number of words per exchange has not been counted, so it is not possible to draw clear conclusions about raw quantity. This means that quantity is defined in terms of the discourse model only ("exchanges" and elements).
One use of this kind of analysis is to interpret the distribution of elements between participants. [-5-]
The distribution figures do tell us something about roles--who controls the discourse and to what extent. Using this and other information, it is then possible to interpret descriptive evidence pedagogically. Qatari students mainly produced response elements only. They produced only 22 non-response elements over all eleven lessons, only (0.92%) of all non-response elements. The response element is the least significant element for directing or structuring the discourse; the "I," "N," and "T" elements allow the speaker to initiate, negotiate or terminate exchanges. In addition, of the 22 (0.92%) non-response elements produced by students from the total of 2,386, all except three were questions about procedure. Ten of these non-response elements (only 2 of which were not about classroom procedure) were produced in one lesson containing only 29 exchanges and 117 elements.
I have argued that representing teacher-fronted classroom discourse as automatically excluding spontaneity or choice would be a misrepresentation. Multiple re-analysis of the data did, however, reveal highly repetitive features of structuring which are relevant to this discussion on ritual and choice.
In 54% of all exchanges that a teacher terminates (more than 90% of all exchanges), the teacher repeats the student's or students' response verbatim, whether it was grammatically well-formed or not. In Figure 1 (reproduced below for convenience) we may note that the wrongly formed student response is repeated verbatim by the teacher. There is more than just the confirmation of a "correct" answer involved.
Terminations where the student's response was slightly moderated are not included in these figures, so 54% is a minimum figure for a common discourse pattern. Sometimes a student's response is not merely repeated by the teacher, but also echoed in a whole exchange, doubling the amount of repetitive structuring.
The overall conclusion is that the data sample reveals high levels of repetitive patterning. In terms of "ritual" it is on the "fixed" end of the relative degree of structuring. Nevertheless, the constant negotiation with the students represents highly-skilled activity in which all participants engage (often enthusiastically) in lively interaction. Whatever the intended aims of the "communicative" textbook, the result of the classroom approach is often the verbatim reconstruction and memorization of extracts from the texts both in class and in examinations. The importance of repetition in discourse can be seen as social. As McClaren (1999, p. 219) points out, "instructional rituals" can be experienced "as part of the socially accepted framework of schooling". On the other hand, it is important to consider, without rejecting teacher-fronted interaction per se, whether verbatim repetition of a text without any emphasis on meaningful processing is a pedagogically useful activity. [-6-]
This paper has represented teacher-fronted discourse structure as a "repertoire of limited choices." If it is acceptable to present discourse structure in this way, this may lead to different conclusions about the value of a very common form of classroom interaction. I have further considered (after McClaren, 1999) the concept of "ritual" as a necessary feature of identity and performance of all classroom discourse. In terms of identity, it concerns everyone (others, but especially ourselves) and is arguably necessary, meaningful, symbolic (pointing beyond itself and shaping mental dispositions, values and habits). In terms of performance, "ritual" refers to regular, not single, performances. Even highly repetitive performances are potentially energetic, rich and lively group experiences. Performers of interaction rituals have established discourse roles. There are "right and wrong" ways of performing, the "rights" and "wrongs" being negotiated locally. By representing the structure of interaction as a repertoire of choices we admit that there are "zones of flexibility" (Stoller, 1992) and relative degrees of structuring (from fixed to more spontaneous) which permit improvisation on basic patterns.
Rituals--whether viewed positively or negatively--are features of classroom discourse in institutions and are not obstacles to change per se. An adequate description of the regular roles and status of teachers and students has normally been seen as an essential prerequisite for innovation. An adequate description requires us to acknowledge that teacher-fronted classroom rituals do not exclude negotiated choices. Reducing a complex rank-scale model to a three-part IRF structure misrepresents teacher-fronted classroom discourse as being more rigid and meaningless than it is. The characterization of rituals and exchanges as repertoires of limited choices is intended to redress the balance. In teacher-fronted class discussion, students have easier access to negotiating elements, "N", than to initiating elements, "I". This suggests that the distinction made in the present model is highly relevant to curriculum decisions.
The roles and relative status of performers in classrooms may be inferred from the variable distribution of different choices that would be equally available to participants of similar status, such as the right to initiate or negotiate in an exchange, to initiate or develop a topic or the means of influencing the selection of the speaker. The status of performers' roles is reflected in different rights and obligations resulting in the variable distribution of available choices between participants during their interaction. It is also susceptible to negotiation between participants. Variable social status need not necessarily be understood only in terms of power. One Arab colleague perceived the relationship as one in which "each respects the role of the other." Once an adequate understanding of current practice has been achieved, the next stage is to decide what might need changing and how feasible such a change would be within a complex social structure. In a post-communicative era, representing teacher-fronted discourse as a potentially useful and flexible educational tool can provide a more adequate baseline description, resulting in a more balanced and compatible plan for innovation.
 The paper assumes some knowledge of the considerable literature on exchange structure since 1975 which is also fully discussed in Nunn 1996b, but which will not be referred to in detail here. (See Coulthard 1977; Burton 1980; Berry 1980; Coulthard & Montgomery (Eds.) 1981; Francis, F. & S. Hunston (1987 & 1992); Sinclair & Brazil 1982; Stubbs 1983; Coulthard (Ed.) 1987 & 1992; McCarthy 1991; Willis, D 1992; Willis, J (1987 & 1992) Hoey 1993; Tsui 1994.) [-7-]
 What is at issue here is not so much the whole Sinclair and Coulthard model itself, which is a rich and complex model allowing intricate analysis at several ranks. It is more the way the exchange level three-part IRF paradigm is extracted from the model and used as a common representation of teacher-fronted discourse. [-8-]
 Copy available on request by attachment file from email@example.com.
Andersen, J., Nussbaum, J. & Grant, J. (1999). Interaction skills in instructional settings. In A. Vangelisti, J. Daly, & G. Friedrich (Eds.), Teaching communication theory and research methods (pp.359-374). London: LEA.
Berry, M. (1980) Layers of Exchange Structure Discourse Analysis Monographs No 7: University of Birmingham, English Language Research.
Bratten, N. (1978) Foreign Language Learning--Why do we Fail ? In Teacher Training Guidlines for the Arab World. Beirut: ELTA OUP
Burton, D. (1980) Dialogue and Discourse. London: Routledge and Kegan Paul.
Carbaugh, D. (1989) Fifty Terms for Talk: A Cross-Cultural Study. In Korzenny, F & S. Ting-Toomey Language Communication & Culture. (pp. 93-120) Newbury Park: Sage.
Coulthard, M. (1977 & 1985) An Introduction to Discourse Analysis. London: Longman.
Coulthard, M. (1981) Developing the Description. In Coulthard, M. & M. Montgomery (Eds.) Studies in Discourse Analysis. (pp. 13-30) London, Routledge
Coulthard, M. (Ed.) (1987) Discussing Discourse. Birmingham.: E L Research.
Coulthard, M. (Ed.) (1992) Advances in Spoken Discourse Analysis. London: Routledge.
Coulthard, M. & D. Brazil. (1981) Exchange Structure In Coulthard, M. and M, Montgomery (Eds.) Studies in Discourse Analysis. (pp. 82-106) London, Routledge.
Coulthard, M. & D. Brazil. (1992) Exchange Structure In Coulthard, M. (Ed.) Advances in Spoken Discourse Analysis. (pp. 50-78) London, Routledge.
Coulthard, M. & M. Montgomery. (Eds.) (1981) Studies in Discourse Analysis. London: Routledge & Kegan Paul. [-9-]
Coulthard, M., M, Montgomery & D. Brazil. (1981) Developing a Description of Spoken Discourse. In Coulthard, M. and M. Montgomery(Eds) Studies in Discourse Analysis. (pp. 1-50) London, Routledge.
Francis, F. & S. Hunston. (1987 & 1992) Analysing Everyday Conversation. In Coulthard M. & M. Montgomery(Eds.) Studies in Discourse Analysis. (pp. 107-148) London, Routledge and Coulthard, M (Ed.) Advances in Spoken Discourse Analysis. (pp. 123-161) London, Routledge.
Goffman, E. (1983) Replies and Responses. In Goffman, E. Forms of Talk, 5-77. Philadelphia: University of Pennsylvania Press
Hoey, M. (ed.) (1993) Data, Description, Discourse. London: Harper Collins.
Hoey, M. (1993) The Case for the Exchange Complex. In Hoey, M. (Ed.) Data, Description, Discourse. (pp.115-138) London: Harper Collins.
McCarthy, M. (1991) Discourse Analysis for Language Teachers Cambridge CUP
McClaren, P. (1999) Schooling as Ritual Performance New York Rowman and Littlefield Publishers Inc. 3rd Ed.) New York
McDonough, J. & C. Shaw (1993) Materials and Methods in ELT. Oxford: Blackwell Griffin, P. & H. Mehan. (1981) Sense & Ritual in Classroom Discourse. In Coulmas, F. (Ed.) Conversational Routines. The Hague: Mouton.
Nunn, R. (1996a) The Method-in-Use in Qatari Secondary Schools (Selected Papers in the TESOL ARABIA (AL-Ain UAE, March, 1995) Conference Proceedings 1995 & 1996 -- In Holbrook, A., C. Coombes and S. Troudi (Eds.) pp. 1-29)
Nunn, R. (1996b) Describing Classroom Interaction in Intercultural Curricular Research and Development. Unpublished PhD Thesis University of Reading. Centre for Applied Language Studies.
Nunn, R. (1999) The Purposes of Language Teachers' Questions. International Review of Applied Linguistics in Language Teaching (Vol 37, Issue 1: Feb. 1999: 23-42).
Sacks, H., E. Schegloff. & G. Jefferson (1974) A Simplest Systematics for the Organization of Turntaking for Conversation. Language: 50.4 (pp. 696-735)
Schiffer, I. (1997) Symbolic Worlds Art, Science, Language, Ritual New York CUP
Sinclair, J. McH. (1992) Priorities in Discourse Analysis. In Coulthard (Ed.) Advances in Spoken Discourse Analysis (pp. 79-88) London: Routledge.
Sinclair, J.McH. & D. Brazil. (1982) Teacher Talk. London: OUP.
Sinclair, J.McH. & M.R. Coulthard. (1975) Towards an Analysis of Discourse: the English Used by Teachers and Pupils. London: OUP. [-10-]
Stoller, F.L. 1992 Analysis of Innovations in Selected Higher Education Intensive English Programs: a Focus on Administrators' Perceptions. North Arizona University. Unpublished PhD thesis.
Stubbs, M. (1983a) Language, Schools & Classrooms. London: Methuen.
Tsui, A. (1994) English Conversation. Oxford: OUP.
Willis, D. (1992) Caught in the Act: Using Rank Scale to Address Problems of Delicacy. In Coulthard, M (Ed.) Advances in Spoken Discourse Analysis. (pp. 111-122) London: Routledge.
Willis, J. (1992) Inner & Outer: Spoken Discourse in the Language Classroom. In Coulthard (Ed.) Advances in Spoken Discourse Analysis (pp. 162-182) London: Routledge.
Roger Nunn has a PhD from the Centre of Applied Language Studies at Reading University in England. He is currently a Professor in the Department of International Studies at the University of Kochi in Southern Japan. He has been an itinerant EFL teacher and lecturer since leaving Britain for Germany in 1979. After a brief stay in Ethiopia he moved to the Middle East where The Gulf State of Qatar became a home from home for almost 12 years. He is currently Professor of Applied Language Studies at Kochi University in Southern Japan where he has taught in the department of International Studies since 1995. His main academic interests are the analysis of approaches to language teaching from an intercultural perspective and curriculum development.
© Copyright rests with authors. Please cite TESL-EJ appropriately.
Editor's Note: Dashed numbers in square brackets indicate the end of each page for purposes of citation.