Abstract

This study analyzes the agenda-setting done by anchors using topical management in the interactional aspect of the institutional speech event; the Political News Interview (henceforth Political talk-show). In the current scenario of 24/7 news, audiences turn to political talk-shows to understand the events of the day and judge the news. The discussion on topics in the political talk-shows is managed and developed by the anchor. This study carries out a discourse analysis of the recorded talk-shows for topical shifts and perspectival shifts to check if the topic is set and controlled by the anchor, by not allowing participants to bring any information to the discussion. The qualitative analysis of topical shifts reveals that the anchor keeps tight control over the topic set in the openings and using the technique of topical management the anchor keeps the agenda set through the development of perspectives and recycling of topics.

Key Words

Discourse Analysis, Discourse Access Profile, Topical management, Agenda Setting,           Pakistani Political talk-shows, Topical shifts, Recycling of Topics

Introduction

Agenda Setting as Interaction

The topics represented in the openings of a political talk-show/interview assume the significance of topical macro structures that initiates the discussion which is then followed by an elaborate discussion on these same topics from diverse perspectives. This study analyzes this interactional aspect and checks it, for the specific power displayed by the anchor and explores how it relates to agenda setting done in the openings of the programs selected as sample for the study. The analysis carried out is guided by the following theoretical underpinnings that inform this analysis of topical management.




Topical Management

Topics, speaking generally are the matters or subjects under discussion in discourse or conversation. The notion of a topic is very difficult to put down theoretically it can be understood with the notion of “speaking topically” or “the speakers’ topic” within a topical framework (Brown and Yule, 1983). Coulthard (1986) views a topic to be a dynamic feature of a conversation that progresses with the flow of topics. In an everyday conversation, both participants equally control the flow of topics they are at liberty to develop previous topics or introduce newer ones. According to Brown and Yule (1983) in discourse analysis, a researcher has to intuitively decide about the notion of a topic for a chunk of talk, because in a conversation a speaker change does not signal a topic change as well. It follows then that a chunk of conversational discourse can be treated as a unit of some kind because it pertains to a particular ‘topic’.

Topic as a discourse unit has received much attention from linguists (Gundel, 1977, Halliday, 1994 and Lyons, 1995), discourse analysis (Van Dijk, 1977; Brown and Yule, 1983; Bublitzs, 1988 and Yule and Mathis, 1992), and conversational analysts (Schegloff and Sacks, 1973; Schegloff 1979). All these different research approaches have helped forward the understanding of how discourse, in general, is structured, yet with all this attention there is still no consensus about what constitutes a topic or at which level can a topic is studied. Linguists have focused on the notion of the topic at the clause level and identified it as a sentential subject. Keenan and Schieffelin (1976: 380) propose a better approach to the topic from the perspective of discourse as a proposition and not just a noun phrase. Their concept of the propositional topic is still a semantic notion and as yet has not broken the boundary of the sentence level (Brown and Yule, 1983). Yet another level at which discoursal topics have been looked at is as an underlying framework for a whole text (Van Dijk, 1977), much like a title of a story might function.




Agenda Setting and Discourse Access Profile of Anchor

In the news interview the topic is highlighted at the beginning by the anchor who announces it in the first segment of the program; the opening. After having introduced the topic of the program it is further developed, analyzed or elaborated by the different perspectives from which the anchor questions his guests and asks them to comment upon it. Sometimes the openings only contain the headlining of only one topic; however, it might be that the anchor sometimes headlines a couple of topics that might be taken up later on in the program. When more than one topic has been mentioned in the program opening the most important one is selected and formulated into a question by the anchor.

A news interview is a form of institutional discourse and as such, the anchorperson holds an institutional role, in that role he has the power to select the topic to be discussed. Van Dijk (1993) in his analysis of various modes of discourse access draws a parallelism between social power and discourse access. He proposes that for every group, position or institution a distinct ‘discourse access profile’ can be mapped out. It follows, that any position that entails the management of a discourse participation event such as a news interview would allow access to control over turn allocation, topic selection, important decisions and management of other consequential dimensions of this institutional talk. In keeping with the discourse access profile of the anchor he selects and manages the topics of the interview. This selection choice can be guided by many factors such as emergent news events, significance in relation to current affairs, or it could also be an editorial choice. 

Thus, it is in the discourse access profile of the anchor to select a particular topic, develop it through various perspectives and shift this topic when he feels he wants to move to another topic the understanding of the selection process and the shift and negotiation of topical shifts have been developed by keeping the discourse access profile of the anchor as an institutional role in mind.

The reality of the process of how the selection and the choice of a topic are made can only be ascertained by having access to the production process that takes place behind the scenes. However, the setting of the agenda for the talk and then the progressive control of that topic and its progression can be charted out by carrying out an analysis that maps the setting of the topic and its progression, negotiation and recycling in the process of the program. Also, important insight can be gained into the comparison of topic choices across programs and channels to highlight what type of topics are chosen in relation to the events taking place, are they reflective of just presenting events taking place or are they news generative events? Topical development in a news interview takes place by analyzing it from many perspectives; therefore the topics considered important by the anchor would be analyzed from many different and varied perspectives and would be recycled during and across programs. 

Method

The overarching objective of the current study is to explore the discursive power manifestations (Agenda Setting) present in the anchor discourse of Pakistani Political talk shows. The researcher collected data from PTV World. This English news channel is state-owned and comes within the public sector. PTV World provides news updates around the hour and in-depth political analysis to not only national audiences but also to the global audience. The program selected from PTV World is ‘The World Tonight’ (WT), it is aired during the week from Monday to Thursday. A snapshot of a week is taken as the data for this research.  

The political talk shows are large chunks of data that cannot thematically be analyzed in one go; therefore, the researcher devised a framework whereby boundary elements; Openings, Closings and pseudo-openings and pseudo-closings, informed by the CA tradition could be analyzed in the first section of analysis for the representational aspect of power in setting the agenda, while the interactional aspect of power was analyzed in the next section of analysis. This framework can be graphically visualized as given in the diagram below:

Figure 1

Analytical framework developed for the analysis of Political Talk-Show



Figure 2

Agenda Setting through Interactional Discourse



The division of analysis on the level of representation and interaction gave rise to different types of analyses that were carried out to get the answers to the questions posed. The three interactional aspects of a dialogic news interview are: 

The analysis was carried out to see how often the anchor uses his role related power to control the talk through topical management of the topic under discussion.

The process of selection for this analysis was justified and the detail about it is given below with examples. The selection of what the topic of the talk was and what was meant by perspective; the researcher had to read the transcription and make choices as to what the topic was. In most cases, for example, the analysis of the topics in the main openings it was most clearly mentioned what the participants and anchors would be talking about for example:

[1] 23rd June 2014 WT

IR: …today the subject that we are talking about is the operation Zarb e Azb…

The above example is a clear topic selection by the anchor of what the topic of the talk was going to be it was going to be ‘the Operation ZA (OZA)’, yet at the point of formulation of the first question, the anchor selected an aspect related to this topic an asked a question related to the topic but not directly the topic itself this kind of questioning from a certain aspect has been called a ‘perspective’ in the context of this study. An example of perspective is given below from the same program as above:

[2] 23rd June 2014 WT

IR:… in your opinion a what do u think that the role of the media has been in projecting the operation …

So the discussion is on the topic of OZA yet it is from the perspective of the ‘role of media’, then when the anchor felt he had covered the role of media in projecting the OZA enough he would change that aspect or angle yet still be asking about OZA just from a different aspect, that would be considered a shift in the aspect or perspective and would be counted as such. Sometimes, this shift of a perspective would emerge from the answer that the participant gave and the anchor would use and forward it as a legitimate aspect to be covered and this was counted as perspective shift initiated by the other. The following is an example of this kind of aspect shift.

[3] 23rd June 2014 WT

IE: <G1.1.1> but what kind of support and strength are signaled or messaged are you sending to the armed forces when an all these there is internal squabbling the I feel that

IR: <A1.1.1> Sir we’ll raise that question, in fact, we’ll take that question to Owais Tauheed sahib who is joining us on the line he’s a senior analyst Owais Tauheed sahib a very interesting question has been raised by one of our guests in the studio major general Naeem sahib sir what kind of a message are we sending to our armed forces when we have a circus being created on the streets of Lahore and Islamabad and a Rawalpindi what is your opinion of that sir.

In the above example 3 it can be seen that the guest is giving his opinion on the previous aspect the anchor has proposed and in answering it he reached a conclusion ‘what kind of a message are we sending to our armed forces’ so this is then a perspective shift where the topic is still OZA, we have not started talking about the political issues or the IDPs issue we are still talking about the operation, yet we are not trying to answer the aspect that what kind of a message is being sent to the armed forces involved in the operation. As the above shift in perspective of ‘role of media’ to the ‘message being sent to forces’ is a perspective shift brought on by a guest hence the ‘other’ initiated shift.

All the above examples are taken from the same program to give a coherent idea about the selection process and also the choice is made from the set of data to clarify the selection process. This is not to say that there were not areas where the idea of topic or perspective was blurred and the researcher had to make intuitive choices as to what is being talked about and from which perspective. The following analysis is developed from this selection process to see what kind of control and setting of agenda is carried out by the anchors of the programs.

Result and Discussion

Agenda Setting as Interaction: Topic Management

In the CA tradition, the topic is initiated in the opening segment of the news interview and its progress can be traced across the type of questions asked on it from different perspectives. The anchor/presenter, as the initiator and manager of the talk, is at liberty to make a topic shift and initiate a different topic. Sometimes this topic shift is initiated by a guest while commenting upon a previous topic. The anchor, in this case, has the authority to accept this topic change and allow it by asking a question upon it from the same guest or another guest or he can contest it and dis-allow it by opposing it and not forwarding it by asking a question upon it. Some topics which are considered really important and significant keep getting recycled by the anchor again and again in the program. The topic shifts and the initiators of this shift should lend an understanding into the kind of control the anchor has in deciding to shift the topic to a new one or recycle a previous one or to allow negotiation from a participant guest. Therefore a mapping of the topical shifts has been done to show the patterns and types of topical shifts made. Also, the topics of the main openings are analyzed to see if they reflect any insight into how agendas are set for the talk.

 

Topics Initiated in Main Openings of WT

The main openings of political news interviews are functionally an important point in the program as they set the agenda for the talk, establish important background information, introduce the guests and formulate the first question. In all these functions, an important stage is to introduce the topic in the headline segment of the opening. The topic as taken in a broader thematic sense proposed by Van Dijk (1977) can provide a cohesive framework for the whole talk. In this sense, such a discoursal topic functions like the title of a story. In this segment, a meta-textual analysis of the topics as announced in the main openings is carried out. Table 1 gives the range of the main topics as analyzed in the main corpus:

Table 1. Topics Initiated in the Main Openings

prog	Topic 1	Topic 2	Topic 3	Topic 4
P1.1.	Operation Zarb e Azb (OZA)	-	-	-
P2.2	Operation Zarb e Azb (OZA)	efforts for the IDPs	ban on television channels	-
P3.2	political parties and their agenda	Impact on IDPs	the operation OZA	Qadri Sahab is talking about a revolution
P4.1	operation Zarb E Azb (OZA)	Crisis of IDPs	maneuvers made to harm democracy	-

The above table shows that as far as the main topics of the political news interviews are concerned there seems a topical consensus across the programs. The above 11 topics highlighted in the main openings can be categorized under five broad topic heads. The above table 1 of main openings shows that there are a total of five main events considered important enough to include in the openings of programs of the main corpus, the list and their percentages are given in table 2:

Table 2: List of Topic heads and Percentages

Topic Heads	Percentages
Crisis of IDPs	25%
Operation Zarb e Azb (OZA)	37.5%
Dr. Tahir ul Qadri	12.5%
Political parties and their agenda	12.5%
Media/ PEMRA	12.5%

From this table we can say that the most important topic as highlighted in the main openings is the Operation Zarb e Azb as it accounts for 37.5% of the main openings, the second-highest percentage of the topic set in the main openings is the Crisis of the IDPs as it accounted for the main openings 25%. We can say that this depiction of the main openings though important as far as setting the agenda of the talk is concerned only shows half the picture, these topics re-appear as recycled topics across the news interviews and are initiated mainly by the anchor and we can say to some extent conclusively that the anchors set the agenda and control the topic throughout the program. Table 1 highlights the topics mentioned in the headline by the anchor and as can be seen from the table, sometimes more than one topic is being highlighted by the anchor. Only one topic can be carried forward by the anchor as the topic for local discussion and this involves a further selection of one topic from the ones mentioned which are used to formulate the first question of the show. The analysis of the selection process and first question shows that for program P2.2 out of the three topics OZA gets to be forwarded as the first topic to be formulated into a discussion question, for P2.3 it is the political parties and their agenda that is selected first and gets formulated into a question. This selection is in keeping with the percentages that we have got from the analysis of the main openings and the topics headlined in them.

 

Topical Range and Initiation Pattern in WT

The above section shows the topics highlighted in the opening, in the course of the program these topics are discussed by guest practitioners (politicians) and commentators (analysts) from varying perspectives. During these discussions, a particular topic from a certain perspective (aspect) is exhausted or new topics emerge from the talk ensuing, in such a situation the anchor generally initiates a shift from the previous topic to a new topic. A range of topics was covered in the programs making up the main corpus. This range is achieved by different types of shifts in topics, this pattern of shifts and initiation is captured in the quantitative pattern of this table is presented below in table 3 to give an idea of how many times topic shifts occurred, it also shows the initiation pattern as to who was this shift initiated by, finally it gives us the total number of perspectives that were used to develop these topics and also again who was the initiator of these perspectives. The idea of the topic head is like the title of a story that gives thematic structure to a story, similarly, a topic head would be the main title topic for example ‘The IDP crisis’, the subsequent questions or comments initiating an answer on that aspect for the topic head are considered to be perspectives. When the anchor feels that a particular aspect of the topic has been covered he puts forward a different aspect in his next question or comment, this shift is generally signaled by an affirmative discourse marker such as ‘right’. Sometimes the discourse marker used at the beginning of a perspective shift can be ‘but’, which is not a signal of disagreement generally the anchor has found a comparative point from the discussion or would like to add a comment to the answer given by the participant.

Table 3. Topic shifts and Initiation Pattern

	Total no. of Topic Shifts	Presenter Initiated topic shifts	Other Initiated topic shifts	Perspectives in the main turn introducing the topic shift	Perspective shifts	Presenter Initiated  perspective shifts	Other Initiated perspective shifts
Total	37	30	7	30	69	63	6

Table 3 gives a clear indication that the topic shifts are controlled and initiated by the anchor. Topic shifts are negotiated by the guests only 19% of the time while the anchors' initiation of topic shift occurs 83% through the course of the discussion. The development of these topics is done by elaborating about them from different perspectives the development of these 37 topics across the 3 programs was achieved through 69 different perspectives out of these perspectives only 6 times the perspective was initiated by the guest and 63 times by the anchor, giving us a ratio of 9% is to 91%. The above table clearly shows that even though the guests are called to the programs as expert analysts or practitioners of politics still their voice is not seen in the initiation of a topic perspective or a shift.

The above pattern is translated into a range of topic and it is analyzed how the shift takes place is it a new topic that is introduced or a previous topic thought important enough to be recycled by the anchor. There are instances of topic negotiations by the guests where they introduce a shift in the topic in their turn and that is allowed or disallowed by the anchor. A topic shift if allowed by the anchor will be forwarded by the anchor by a perspective shift. There is only one instance of the anchor dis-allowing a topic negotiation and it resulted in a disagreement between the anchor and the political guest. Table 4 shows the range of topics covered in these programs and can also show in which significant way are these initial topics headlined in the main openings developed or held on to in the course of the program.

Table 4. Range of Topics and type of appearance and reappearance in the show

Topics	Appearance in Main Openings	Appearing as fresh topic shifts	Negotiated topics initiated by guests and allowed by anchor	Recycled topics brought in again	Negotiated topics initiated by guests dis-allowed by anchor
Operation ZA	3	-	-	3	-
The IDP issue	2	1	-	10	1
Tahir ul Qadri and his demands	1	2	-	2	-
Media’s Role	-	2	-	1	-
Fascination with Messiah	-	-	1	-	-
Ban on TV channels/PEMRA	1	-	-	-	-
Constitution of Pakistan	-	1	-	-	-
Role of private sector	-	1	-	-	-
Pakistan Gov. presence in Waziristan	-	-	1	-	-
Lahore Incident		1			-
Political parties and their agendas	1	1			-
Strategic debt		1			-
Political agenda of PTI		1		2	-
Presence of PM	-	-	1	1	-
Future Plans	-	1	-	-	-
News Reporting	-	-	1	-	-
Terrorist group and their resources	-	-	1	-	-
Annexation of  FATA	-	1	-	-	-
Political manoeuvres to harm democracy	1	-	-	1	-
Media coverage of OZA	-	-	1	-	-

The above table 4 does show that the topical agenda set in the opening segment of the political talk-show gets to be the main topic throughout the show as the anchor regularly recycles it back into the discussion done by the guests. On the other hand, the topics brought in by the guests the negotiated topics are not recycled or followed up by the anchor. Most of the topics brought in and recycled in the above table can fall under the main topics highlighted in table 1 with just the addition of two additional head topics. For the purpose of checking the topics from the main corpus, this grouping is done to get a percentage idea about the main area the topics pertaining to. This collapsed topic range under topic heads can be seen in table 5

Table 5. Collapsed topic range

Topics	Appearing in the main opening	As fresh topic shifts	Negotiated topic shifts	Recycled topics	Negotiated topics disallowed	Total topics under topic head	Percentage %
The Operation ZA	3	-	-	3	-	6	13%
The IDP crisis	2	1	-	10	1	14	33%
Tahir ul Qadri	1	2	-	2	-	5	12%
Media’s Role	1	2	1	1	-	5	12%
Political parties and issues	2	1	1	3	-	7	16%
Government issues	-	3	3	-	-	6	13%

 

Table 5 while reaffirming the previous claim of the topics headlined in the main openings as setting the agenda of the talk, also points to another interesting insight that the other important issues not highlighted in the main openings are brought in as fresh topical shifts and get some representation in the programs. The negotiated topic shifts offer some space for the bringing in of topics not raised by the anchorperson in the course of the program, yet the anchors still have the control of allowing these topics to become legitimate topics. Above table 4.11 shows that the political parties and political issues are an important part of the events taking place every day get captured in fresh topics introduced by the anchors and sometimes by the guests too they are not introduced as main topics in the main openings.

Topic Progression in WT

The previous segments have highlighted the topics that are present and kept reappearing in the program. These topics are not discussed directly they are maintained over a period of time in the discussion by the anchor presenting the questions pertaining from different perspectives to the guests. Therefore, topic progression can be seen by the development of it across many perspectives.

The topic progression across each program has been mapped out to see what kind of insight towards topic management can be seen across the individual programs as a whole.

Table 6. Topic Progression across Programmes Programme P1.1

Topics	No. of perspective through which they are developed	Presenter initiated perspective	Other initiated perspective
Operation ZA	3	2	1
IDP crisis	7	7	-
Tahir ul Qadri	2	2	-
Role of Media	4	4	-
Fascination with messiah	1	-	1

Program P2.2

Topics	No. of perspective through which they are developed	Presenter initiated	Other initiated perspective
Operation ZA	4	3	1
IDP crisis	9	5	4
Tahir ul Qadri	2	2	-
Gov. presence in Waziristan	5	4	1
Role of media	5	5	-
Constitution of Pakistan	1	1	-
Lahore incident	2	2	-
Role of the private sector	1	1	-

Program P3.2

Topics	No. of perspective through which they are developed	Presenter initiated	Other initiated perspective
Political parties agenda	1	1	-
Strategic debt	4	4	-
Political agenda of PTI	8	8	-
IDP crisis	10	9	1
Tahir ul Qadri	2	2	-
Where is the PM	1	1	-
Future plans	1	1	-
News Reporting	2	1	1
Terrorist and their resources	2	2	-
Annexation of FATA	1	1	-

Program P4.1

Topics	No. of perspective through which they are developed	Presenter initiated	Other initiated perspective
Operation ZA	8	-	-
IDP crisis	2	-	-
Political will	2	2	-
PTI political agenda	5	5	-
Political maneuvering	4	-	-

The first insight that we can gather from the above development and progression of the topics is concerned to the aspect of controlling the topical agenda setting, the topics are largely managed by the anchor who clearly seems to be the one in position to make the topical and perspective shifts significantly more often than the participant guests politicians and analyst. Out of the above perspectives, only 6 perspectives are given by the guests while 63 perspective shifts were made by the anchor/presenter. In terms of percentages, we would get a percentage of all most 91% perspectives from the anchor and only 9% by the guest participants. It is interesting that these guest participants have such a low input towards the perspectives, considering they are invited to speak upon a topic as analyst and expert practitioners of the political field or political party. The strongest voice then is that of the anchorperson who seems to do the major development of the topic that he himself has suggested.

The above mapping of topic progression suggests that the topics that are developed across the most number of perspectives are the ones that have the prime focus of the anchor as he invests a lot in developing them. The charted programs in the above tables for the political talk show ‘The World Tonight’ (WT) shows a more balanced focus across topics of importance and topics of less significance to the channel. It can be observed that the most progressed and developed topics are that of ‘The Operation ZA’ and the ‘IDP crisis’ in third place by the ‘Role of Media’, finally, in a considerably lesser degree by the emergent event of the particular days i.e. Tahir Ul Qadri’s  (TUQ) return and the subsequent issues. If we were to contrast the emergent event of the day and its development we see that on the 23^rd of June 2014, the day that TUQ arrived back in Pakistan and it led to a media event, is developed across only 2 shifts in the program of WT aired on 23^rd, June 2014. As an emergent media event that took place on that day, it would have had a little more representation in the talk show, yet it was not even developed in subsequent shows of WT. In this manner, we can see a selection process that either highlights or downplays a topic. This agenda-setting aspect of the interactional power of anchor becomes clear from the above discussion and examples. It also suggests that the political talk shows can overplay topics they want to highlight and underplay the topics they do not want to become significant to the masses.

 

Topical Recycling in Main English Data

While charting the topical development it was noticed that topics get recycled during the course of a program and it seems to be a convenient way for bringing the discussion back to the topic that the anchor wants the focus to remain upon. Keeping this reasoning in mind table 4.12 below shows the most recycled topics and their percentages to see if we can relate them to the topics proposed in the main openings.

Table 7. The pattern of Topic Recycling

Topics	Number to times recycled	Percentages
Operation ZA	4	18%
The IDP issue	11	50%
Tahir ul Qadri and his demands	3	14%
Political agenda of PTI	2	9%
Media’s Role	2	9%

The above figures for topic recycling display the hierarchy or the exact same topics that were found to be highlighted in the main openings of the talk shows of the main corpora. A comparative chart is developed from the percentages in table 2 appearing as blue bars and those in table 6 represented in red bars.

Figure 3

Recycling of topics in WT

The recycling process consolidated the setting of agenda by the anchors and they seem to recycle the same topics as highlighted in the main openings of the talk shows.

Conclusion

This analysis has been carried out to understand how the anchor keeps the topic on point while interacting with the participants. The above discussion and analysis reflect that the anchor keeps the control of the topic even after they have represented it in the main openings. Another finding of the analysis is that the topics highlighted in the main openings get developed through a lot of perspective shifts. Finally, the topics that are considered most important get represented in the main openings and then they are the ones that get recycled most by the anchor during the course of the programs. 

The analysis shows how the anchor sets the agenda even in the interactional part of the political talk-show by using topical management to restrict the participant guest politicians to keep to the topic brought in by the anchor. These aspects of topics suggest a tight control over the topical agenda set in the openings which are kept on point by using all the techniques the anchor has of keeping the topic set through the development of perspectives and the recycling.

References

• Brown, G. & G. Yule (1983). Discourse Analysis. Cambridge: Cambridge University Pres
• Bublitz, W. (1988). Supportive fellow-speakers and cooperative conversation: Discourse topics and topical actions, participant roles and
• Coulthard, M. (1986). An Introduction to Discourse Analysis. London: Longman
• Gundel, J. K. (1977). Role of topic and comment in linguistic theory. Mimeo: Indiana University Press
• Halliday, M. A. K., (1994). Introduction to functional grammar. London: Arnold
• Lyons, J. (1995). Linguistic semantics. Cambridge: Cambridge University Press
• Van Dijk, T. (1977). Text and context. London: Longman
• Van Dijk, T. (1993). Principles of Critical Discourse Analysis. Discourse and Society. 4(2): 249-28
• Schegloff, E.A. &H. Sacks, (1973). Opening up closings. Semiotica8(4),289-327
• Yule, G. and Mathis, T. (1992). The role of staging and constructed dialogue in establishingspeaker's topic. Linguistics,30,199-215
• Shegloff, E. A. (1979) TheRelevance of Repair to Syntax-for-Conversation in Discourse and Syntax. Syntax and Semantics 12: 261-286
• Keenan, E. &Schieffelin, B. (1976). Topic as discourse notion: A study of topic in theconversation of children and adults. In Subject and topicLi, C.N. (ed). 335-384. New York: Academic Press