Journal of Pragmatics 16 {1991} 31-58 North-Holland
31
Planning in ordinary conversation Anneke Scholtens* Received Januarr 1988: revised version April 1990
In this article it is sketched how people proceed in making plans in ordinary conversations. The material used consists of so-called 'one-sided planning fragments'. That is to say: planning fragments in which initially only one of the participants develops a plan {and holds the floor) while the other has more or less the role of recipient. On the basis of the recipient reactions it is described which role the recipient plays in the development of the plan. For this purpose concrete recipient reactions which occur in the planning fragments are graded from 'following" to "steering'. Following reactions leave the planning structure of the person who is contributing the plan intact. Steering reactions intervene in the planning structure: they are independent initiatives of the recipient in the form of advice or suggestions. The impact that the recipient has is related to her foreknowledge which is said to consist of: knowledge of plan structures, knowledge of scripts and also personal experience. It is argued that the more knowledge the recipient has of the state of affairs the better she can anticipate the plan and the more initiative she can take. Her reactions can be said to be more steering then. Finally it is demonstrated to what extent the final text of a conversation is a co-production of the participants in the conversation, in this case of the person contributing the plan and of the recipient. "'The central goal of conversation analytic research is the description and explication of the competences that ordinary speakers use and rely on in participating in intelligible, socially organized interaction." (Atkinson and Heritage I984: I)
1. Theoretical background: Planning in conversation 1.1. Introduction The competence on the basis of which language users are capable of participating in social interaction has two components" comprehending the current discourse and making a contribution that fits into the state of the discourse at that moment. * Correspondence address: A. Scholtens, Amsterdamse Academic, Eekholt 32, I l l 2 XH Diemen, The Netherlands. This article was written with the support of ZWO (the Netherlands Organization for Pure Scientific Research) in a project supervised by D. Springorum. 0378-2166/91/$03.50 © 1991 - - Elsevier Science Publishers B.V. (North-Holland)
32
A. Scholtens / Planning in ordinary conversatwn
The first component has been investigated a great deal. Here researchers normally limit themselves to the processing of texts in monologue. The question is posed which knowledge and strategies a language user must possess to be able to interpret those texts. In discourse, however, comprehension does not stand by itself. The recipient is busy in fact with two things at the same time: first of all, with processing what is being said, and secondly with conceiving her own contribution. We cannot say that processing takes place first and then subsequently 'production'; processing takes place in light of considerations pertaining to the planned production. In a monologue text the producer ,s the person who indicates what is the main point in her explanation, how the different components fit into this and how the components relate to each other. This aspect of text production we can refer to with the term 'staging', introduced first by Grimes (1975). Clements (1979:287) applied this term in a slightly different sense: "Staging is a dimension of prose structure which identifies the relative prominence given to various segments of prose discourse'. With dialogues or conversations with more participants, on the other hand, the recipient can influence this 'staging'. (Polanyi 1979). Her contributions to the text production can vary from questions for clarification, by which a certain component gets more attention than the producer judged necessary by herself, to objections or suggestions, by which the car.tent of a compo'-.ent of the text can even be revised. This circamstance, namely that the recipient can participate and influence the conversation and ~s aware of this, makes the question for the recipient whether she understands th- text or whether she is in agreement with the text even more urgent th~a w~h a monologue. This awareness influences the way the text is processed. If comprehension of the discourse is closely tied for the recipient to the thinking about continuation alternatives, research into strategies of processing for participants should be based on material in which comprehension as well as continuation alternatives play a role: not monologues but conversations with more participants. The problem with this type of material, undoubtedly ~he reason why it has not been selected very often in research, is that the concrete moves which the participants in a conversation make always result from collapsing: - their reconstruc"ion of the conversation up to that momen,; their interpretation1 of it; - their comments e.n ;~: and; - their decision to coa6nue in line' with the preceding text or to bring in a new aspect. In other we~ds: the next turn of a recipient rarely provides an unambiguous indication 07 the way in which she has processed the previous text. Still, concrete conversational material seems to be an essential source of informa-
A. Scholtens
Planning in ordinary conversation
33
tion when we are dealing with the description of processing strategies in discourse. In this article I will show how this source of information can be used. For the description of the material I will use existing theories of discourse comprehension of course.
1.2. Description of the corpus The conversational material studied consists of six conversations between pairs of women friends: physical therapists (in training), students of classical languages, women graduated from a teacher training program who want to continue studying, and three pairs of housewives. These conversations lasted approximately half an hour each. From each I have selected fragments in which one of the participants in the conversation provides information, while the other has the role of recipient more or less. A further restriction I made concerns the speech activity engaged in: only those fragments in which a plan was unfolded and discussed were taken into account. The reason for this choice will come up in the discussion below. The result of these selections were 17 fragments of unequal length: the shortest contains 9 turns, the longest 48. Of course the length of these turns is highly variable: sometimes they are short questions, sometimes complete reports.
1.3. Selection procedures 1.3.1. Fragments with one leading participant Fragments in which one of the participants more or less occupies the discourse space (the floor) seem to take an advantageous middle position between, on the one hand, monologue texts, in which there is a clear dividing line between producer and recipient, and on the other hand the complicated situation of 'symmetric' fragments in which both partners in the conversation produce text as well as react to each other. As was stated already, the recipient's task is twofold: comprehending and contributing. It is assumed that in the fragments chosen the verbal reactions of the recipient will display relatively more of the processing side. That is to say: the recipient's contributions can be ch~r~ct~-ised as 'reactions to" the explanation of the plan by the plan i.,ai.*iator. ~inc~ we are interested in the processing side here, this situation is ~dvantagc.:3~,~ for the purposes of the present study. 1.3.2. Fragments in which a plan is unfolded On the other hand it would be of liitle use to work with fragments in which the recipient does not make a contribution at all or limits herself to mere signals of acknowledgement, because that would shed no light on the way the recipient processes the discourse.
34
A. Scholtens
Planning #~ ordinao' conversation
With that in mind I have chosen fragments in which a plan is unfolded (socalled planning fragments), since these fragments are both open-ended and task-oriented. The combination of these two properties facilitates a certain contribution on the recipient's side. This point will be clarified below. 1.3.2.1. Utfolding a plan as an open-ended activity. Speakers" initiatives differ in the extent to which they urge the recipient to participate: questions impose a kind of obligation for the recipient to answer, while assertions do not necessarily have this effect. The same can be said of speakers" explanations. When a speaker reports on some personal experience she can be said to leave the recipient a smaller range of possible reactions than wher~ she starts to recall common experiences. This is partly caused by the fact that in the second case both participants are well informed about exactly what happened while in the first case only the speaker knows the plot. We can say that in the second case the3' have equal access to the relevant information, while in the first case only the speaker has this. Labov and Fanshel even use the term "expert' for a speaker in this circumstance: 'qn ordinary conversation, we are always subject to being contradicted on matters of fact, and we may expect contradiction most often if we speak about areas where the other person is known to be expert and we are not. But a speaker can be confident that there are many areas where he himself is the undisputed expert. These are his personal and private emotions, experience, and all of the events that make up his biography" (1977: 34). They add to this description that the difference in this knowledge will decrease gradually during the speaker's explanation. As soon as the speaker has finished her first version both the participants can examine the reported events. At that time the situation of 'equal access to relevant information" has been installed again. In the act of unfolding a plan the spreading of information is not as unequal as with "reporting on personal experiences' but also not as equal as with "recalling common experiences'. On the one hand the plan initiator knows what she plans to do of course but on the other hand she is talking about the future. So things are open yet and admissible for proposals of change. The recipient can use the knowledge she has from personal experiences with comparable plans to put forward these proposals or to support the plan (as it was) outlined. So they can both be said to have access to relevant knowledge and this is supposed to have a positive influence on the recipient contributions. 1.3.2.2. Unfolding a plan as a task-~riented activity. The expectations about (the structure of) a discourse fragment are more specific to the degree that the fragment is more task-oriented (Brown 1984). This is due to the fact that a task generally has a clear structure: it consists of steps that have to be taken in a certain sequence.
A. Scholtens / Planning in ordinary conversation
35
In more formal studies on planning the act of planning is often described as computing a path from some initial state to a (few) goal state(s). The necessary and/or possible steps of which the path consists are implicationally ordered. The structure of a plan can then be represented as a tree structure. In everyday language, speaking of planning, we would rather use the word 'plotting' a path instead of'computing' one, but in essence we aim at the same process. This idea is worked out by Linde and Goguen (1978) They state that in the text of the planning discourse analysed (the Watergate tapes), one can recognise the underlying tree structure(s). Furthermore they suggest that this tree structure is psychologically relevant for the discourse participants: they make use of a pointer which indicates the participant's focus of attention in the plan tree at a certain moment. 'It points to the part of the plan which they are presently developing and embodies their expectation about where the next development will occur' (1978: 228). Linde and Goguen follow the idea of Grosz (1978) that a task-oriented fragment adopts the structure of the underlying task. The same approach more or less is taken by Cohen (1987) in analyzing the structure of argumentative discourse. Cohen assumes that when some speaker tries to convince a hearer of a particular point of view, this he~,.er at first builds a representation of what it is the speaker wants him to believe (the claim and evidence relations), before judging credibility and eventually responding. This representation takes the form of a tree structure, whereby the 'daughters' of a node are its evidence relations and the "mother' of a node is the claim for which the node itself forms the evidence. Thereby Cohen stresses the point that this expected coherent structure leads to a restricted processing strategy. It is proposed here that the same holds for an expected plan structure. As soon as a recipient understands that her conversational partner intends to unfold a certain plan she expects an exposition composed of certain intended steps (implicationally ordered or not), which will be more or less grounded. This expectation, based on knowledge of plan structures, facilitates the processing of the plan exposition. In addition to this knowledge of structures, which is specially required in processing task-oriented conversations, the recipient uses at least two other types of knowledge: knowledge of so-called "scripts' and personal experiences. These sources of knowledge are more general in the sense that they are also used in dialogues that are not typically task-oriented. They will be discussed below.
1.4. Using script-like knowledge in processing 1.4.1. Sketchy scripts While the plan initiator starts unfolding her plan the recipient tries to imagine what the overall plan will look like. In this connection Van Dijk and Kintsch
36
A, Scholtens / Plann#~g #~ ordinary conversation
(1983) speak of "constructing a mental representation'. They state that 'comprehending' consists for the bigger part of building this representation. It is not suggested here that once the recipient has formed this representation she cannot change it anymore. Of course she makes a prelimina~3' representation at first on the basis of incomplete data, which later can be confronted with information still coming in. We could also mzI'e use of the terms: top-down and bottom-up processing (Van Dijk and Kintsch 1983; Ensink 1986: 10-17; Brown and Yule 1983: 233-236). Bottom-up processing refers to the processing strategy which starts from perceived data, while the term top-down processing is reserved for a processing strategy which starts from a directing principle (for instance a global idea of what is to come). Van Dijk and Kintsch state that a recipient can make use of both processing strategies at the same time. For the planning fragments this means that on the basis of the first set of input data (the data which the plan initiator provides) the recipient forms a global idea in a top-down way of the final complete plan, This preliminary global idea creates an interpretation frame for the incoming data, but conversely the further data processing reshapes and corrects the top-down image already formed. The resemblance with the "expectation structures" mentioned by Tannen (1979), (originally this term stems from Ross 1975) seems clear here: "Thus, structures of expectation make interpretation possible, but in the process they also reflect back on perception of the world to justify that interpretation." (Tannen 1979: 144)
But the recipient herself is not a blank information processor, as Van Dijk and Kintsch make clear in their presuppositionai assumption and this is of special importance here: "'More generally, then, we will assume that understanding involves not only the processing and interpretation of external data, but also the activation and use of internal, cognitive, information. Since this information can be considered as cognitive presupposition of the construction process. we will call this the presuppositional assumption of the model." (Van Dijk and Kintsch 1983: 5)
With respect to the planning fragments this means that we assume that as soon as the recipient knows the subject of the planning, she usually also has a global idea of possible steps in the plan. For example: when the plan initiator exposes a plan to go on holiday, the recipient knows that the plan initiator has certain decisions to make: the destination, the means of transportation, the way to travel, the time to spend, etc. In ordinary conversation people generally discuss ordinary subjects so the chance that the recipient really has this global idea of possible steps is very great. This type of knowledge, these cognitive presuppositions, is often
A. Scholtens / Planning in ordinary conversation
37
referred to in terms of scripts, frames, schemata, etc. In the sense it is used here, within planning fragments, the notion of script wo.~ld fit best since we are dealing with event sequences (Schank and Abelson 1975). Since the recipient knows the components of which the plan consists she can anticipate and infer things that have not been said explicitly (yet) and shorten her processing time this way. In section 1.3.2.2 it was stated that knowledge of task structures on the recipient's side facilitates processing task-oriented fragments. Knowledge of scripts (like going on holiday, renting a house, etc.) facilitates processing discourse in general.
1.4.2. Filled-in scripts Until now we have discerned two sources of knowledge the recipient uses ~r, processing the explanation of a plan: knowledge of plan structures (steps and evidence relations) and knowledge of scripts (standard event sequences). But in ordinary conversation the recipient generally has a third kind of knowledge: her own experiences of a certain standard sequence of events. Presumably she has gone on holiday herself a couple of times for instance. She might have ideas about beautiful holiday destinations, the most excellent and the most boring means of transportation, etc. So besides a global idea of possible steps and possible argumentations the recipient has her own opinion about the most desirable steps and the most desirable argumentation. We shall refer to the first kind of script (possible steps and possible argumentation) with the term 'sketchy script' (Riesbeck 1985) and to the second with "filled-in script'. The effect on processing of this filled-in script is not unambiguous. It depends a lot on whether the conversational partners agree or disagree on the way (the steps of) a plan should be filled in. If the plan initiator tells what she plans to do and that is exactly in line with what the recipient herself would do, it is easy for the recipient to follow this explanation. She will not need much extra information or argumentation; the course of events is self-evident for her. But we could imagine that processing becomes more difficult if the exposed plan and the personal experiences of the recipient diverge a lot. The recipient will be frustated, since she expects to be told something different and she might want to interrupt and ask questions. The influence of diverging recipient ideas on the recipient's contribution will be discussed in section 1.5. 1.4.3. Sketchy scripts versus filled-in scripts This distinction is not really discrete. A sketchy script is said to be a standard sequence of events. The word 'standard' suggests that people judge certain sequences of events as 'normal' and other sequences as 'not normal'. In the first place it is clear that this cannot hold across cultural boundaries.
38
A. Scholtens / Planmng in ordinary conversation
We only have to think of standard events like: making or canceling an appointment, inviting someone, etc. to know what differences occur. But even within cultures there are subcultural pecularities and within these subcultures not all members share the ideas on ~standard event sequences'. Maybe it is rather the case that we all have more idiosyncratic scripts than stereotypical ones (Brown and Yule 1983: 244). So it is not always clear which parts of a script are to be called "sketchy' and which parts "filled-in'.
1.4.4. Using a divergozg script The situation in which conversational partners use diverging scripts does not have to be problematic: if we don't have the same scripts to rely upon for making inferences we can use our inferences to build up the script that is underlying the explanation given. For instance, suppose someone has a habit of riding on the subway without paying and says to his conversational partner: 'This morning when ! was on my way to town there were ticket inspectors in the subway. I was scared but I managed to jump off at Nieuwmarkt station'. The recipient might not have this habit himself but he can infer from the explanation that the other one has it, And he can understand the causal chain that is implicated: riding on the subway without a ticket makes you run away from ticket inspectors so the best thing to do is jump off as soon as these officials are sighted. In the literature these idiosyncratic event sequences with idiosyncratic argumentation structures are called "belief systems' (Abelson 1973) or "belief spaces' (Wilks 1986), The idea is that as long as people are aware of the fact that there can be differences in what they believe and what their conversational partners believe (a so-called apartitioning of assumptions (Wilks)) they can communicate perfectly.
1,5. Using script-like knowledge in contributing to a dialogue When the recipient has little knowledge (in terms of sketchy and filled-in scripts) about the state of affairs, she will be inclined to listen to the plan initiator's account and follow it. The recipient will probably give indications of attention (minimal responses, exclamations, paraphrases, etc.). Her reactions can be called mostly following. To the extent however that the recipient has more knowledge of the state of affairs she can anticipate the plan better. She can ask the plan initiator whether she has done this or that already, whether she has thought about this or that, etc. (see section 2.2.2 example 1). In this last case the recipient takes more initiative. Her reactions can be said to be more steering. At the end of section 1.4.2 it was suggested that a diverging filled-in script does not help the recipient very much in processing. It could even interfere with a smooth processing. In this circumstance the recipient might be inclined
d. Scholtens / Planning in ordinary com,ersation
39
to verbalize this discrepancy between her own filled-in script and that from the plan initiator. She might come up with suggestions or advice and give her own argumentation why she thinks a certain step could be better performed in another way. In this case her reactions can be said to be the most steering: she brings in new points and thereby tries to find an alternative for a certain intended step.
1.6. Factors intervening between recipient knowledge and recipient contribution It is not suggested here, though, that knowledge of sketchy scripts or (deviating) filled-in scripts always automatically results in a certain kind of contributions. There are intervening factors that have to do with differences in power and the rules of politeness (Lakoff 1973). The way relations of power influence the contributions of conversational partners has been described extensively. When we state here that knowledge of scripts (sketchy and filled-in) enables a recipient to contribute actively to the conversation, we must add immediately that this is the case at least in ordinary conversation. It is clear that in formal interactions (between a doctor and a patient or a teacher and a pupil) the participants cannot show and use tt,.ir knowledge in the same way. The one with the least power will sometimes even have to hide knowledge and keep silence. In that case there are other (formal) mechanisms that govern turn-taking. However, I do not claim to give a full account here of how knowledge of task-structures and/or scripts on the side of a recipient relates to recipient-contributions in general, as is done by Linell al. (1988)for instance. The remarks and conclusions made here are strictly related to ordinary conversations. Another factor that intervenes the direct influence from recipient's knowledge on recipient's contribution concerns the rules of politeness. We are particularly interested in the rule: 'Don't impose', which is further worked out by Lakoff as: "Remain aloof, don't intrude into 'other peol:,lo's business'" (1973: 298). It is easy to see how this rule could withdraw a recipient from contributing. We stated in i.5 that especially departing knowledge could urge the recipient to contribute to the exposed plan: to give suggestions and advice or to contest the given argumentation. This rule of politeness asks a recipient to approach this inclination critically or at least to model her formulation in a certain way. It is assumed here that there is a relation between this rule and the kind of contact people have, Obeying this rule on the one hand creates distance between people but on the other hand it also expresses distance. It is not likely that in conversations among friends this rule plays a very central role, since 'not imposing" can be felt here as 'not interested'.
A, Scholtens
40
Plann#zg in ordinary conversation
1.7. The most steerhlg recipwnt reaction: Attaching new hranches in the underl);ng structure
In what follows I demonstrate on the basis of a planning fragment, included in the appendix as sample fragment, what is meant with the most steering recipient reaction: advice or suggestions for the plan initiator's plan (section !.5). In this fragment a student says that she is going to teach in addition to studying. She discusses this with a friend who also combines her studies with work. We will go through this fragment and see how the underlying tree structure is developed in the course of talking. Meanwhile we will come to discuss various recipient reactions, especially the one we are particularly interested in here: giving advice or suggestions. We start ~it',~ lilies i-38: in thi~ t,~ec,, of the sample fragment the recipient starts to recall a plan that the plan initiator must have unfolded some time before. It concerns the schedule of the plan initiator for the coming year. The question seems to be if this schedule is to consist of two components (teaching and studying) or of only one. plan initiator
recipient
SCHEDULE P. [ FOR NEXT YEAR
/% PLAN TO STUDY
PLAN TO TEACH
SCHEDULE P. FOR NEXT YEAR
I
'ol
pr_aN
TO
Fig. !. The recipient questions the branch STUDY and especially the link AND. The plan initiator takes STUDY for granted and defends the branch TEACH and the link AND.
A. Scholtens I Planning bl ordinary conversation
41
The recipient thinks that the plan to teach is definite, while studying on the side is still a matter of discussion. From tile plan initiator's reaction we can learn however that she surely wants to study, but doubts whether there can be thought of some teaching on the side. The main point of interest for both o f them seems to be though whether these two activities can be combined (~ee the link 'AND' in figures I and 2). We can schematize this situation as i~ figure I. In lines 16-38 the recipient at first contests the argumentation given by the plan initiator for the link ' A N D ' and subsequently the plan initiator comes up with an objection to the recipienrs objection and adds a second argument for the branch 'TEACH' and for the !:ink "AND'. The recipient finally agrees. The outcome of this first piece can b,,~ represented as figure 2. ;
ill
SCHEDULE P . / IFOR NEXT YEAR
/i:'l/ / /
Fig. 2. The plan initi~tor and the recipient agree on both branches STUDY and TEACH.
In line 38 the p;,~an initiator starts posing questions to the recipient concerning a certain ~hool (without any explanation at first). The explanation follows soon: t~lere is a vacancy at that school. Strictly speaking the plan initiator does/not give an explicit explanation; but for every cooperative recipient in t}/i~iscase the plan initiator's statement 'since there was a vacancy ...' is heard/as an explanation. The recipient is helped to process the plan initiator's ~'~atement this way by the introductory "since', but most of all by her sketchy script of 'finding a job as a teacher'. This sketchy script can intuitivel/y be outlined as in figure 3. Here/' the main goal is represented as the root of the tree structure. From this nf;de two arcs to "lower' goals emerge. This means that the main goal is deco~nposed in two subgoals which have to be satisfied in order to satisfy the mai~/t goal. These two subgoals are tied by an "AND'-link which means that the.3' both have to be satisfied (see for this representation Rich 1983: 87). /The fact that we deal here with a sequential process (the search for i ,/acancies must be executed before the act of applying can be done) is //implicitly expressed" one processes the branches of a (search) tree from left to • right.
42
A. Scholtens ,~ Planning in ordinary cont'erxation
FINDING A JOB AS A TEACHER
/
] APPLY
SEARCH FOR APPROPRIATE VACANCIES
[
Fig. 3. The main goal FINDING A JOB AS A TEACHER is decomposed in two subgoals SEARCH FOR APPROPRIATE VACANCIES and APPLY which have to be satisfied in order to satisfy the main goal.
Since the outcome of the foregoing piece of discussion was that the plan initiator certainly planned to become a teacher, speaking of vacancies is a logical step. The plan initiator does not have to add "and I was wondering whether I should apply but I am still in doubt since I don't know what kind of school it is and therefore I asked you .... etc.
[13 SEARCH FOR APPROPRIATE VACANCIES
[2]
[3]
SEARCH FOR APPROPRIATE VACANCIES
SEARCH FOR
]
APPROPRIATE
VACANCIES
{
!
S C A N SEV. ISS. O F TEACHER'S JOURNAL
S C A N SEV. ISS. O F T E A C H E R 'S JOURNAL
ASK ACQUAINT. ( in gen.
SCAN
SEV.
ISS.
OF
I ASK IACQUAINT l
TEACHER'S
I(in
gen)
JOURNAL
Fig. 4. The subsequent stages of the recipient's proposal to work on the subgoal SEARCH FOR APPROPRIATE VACANCIES in a more thorough way.
A. Schollens / Planning in ordinary conversation
43
Of course the recipient understands that the plan initiator thinks of applying for the vacancy just mentioned. This is all at once clear from line 69: the recpient asks the plan initiator why she does not wait a little and try some other possibilities. In fact the recipient urges the plan initiator to collect more information about vacancies and the recipient also offers a proposal about the way to do this. To see clearly what happens now we have to give a more detailed version of the sketchy script proposed. From the sample fragment we can learn that the plan initiator tries to satisfy the subgoal on the left in a rather simple way: she just takes the first vacancy that happened to be recommended to her. The recipient proposes to work on that subgoal in a more thorough way. In lines 74-132 of the discourse we can see her gradually fill in this subgoal in a totally different way. In the first place she mentions the activity: 'SCAN SEVERAL ISSUES OF THE TEACHER'S JOURNAL'. Then suddenly she comes with the idea to ask one of their common acquaintances ('ASK H E') whether he knows something [i] and then she generalizes this to 'ASK A C Q U A I N T A N C E S ' (in general) [2] and from there she comes to propose several instances of this general idea ('ASK J', 'ASK M') [3]. In schema this looks as in figure 4. The total underlying plan structure in which both the plan initiator's and the recipient's proposal are represented looks as in figure 5. In the sample fragment we can see that the plan initiator decides to skip her own arc emerging from the goal node: 'SEARCH FOR APPROPRIATE VACANCIES' and take over the arcs proposed by the recipient. Such a recipient proposal we call the most steering, since while the recipient causes a real change of plan, the plan initiator removes a branch proposed by herself and attaches two linked branches proposed by the recipient instead. This kind of reaction differs fundamentally from reactions in which the recipient questions or contests the factual representation of the plan or the argumentation for a proposed step. In these latter cases the plan structure can in principle stay intact. These less steering (or even following) reactions will be discussed in section 2. I choose this point to give a more complete explanation of what was meant in section 1.4.3 with the remark: 'the distinction between sketchy and filled-in scripts is not discrete'. The more global the characterisation of a certain sequence of events is, the more probable it is that this sequence is 'standard', that is to say that people agree upon it. The more details we add, however, the more we move away from standard scripts and the closer we come to idiosyncratic ones. So the sketchy script may consist of a limited amount of real basic steps (see figure 3). Each of these steps (goal nodes) however can be accomplished in different ways. We can expand a certain node and fill in the details thus creating a new tree structure (figure 4: l, 2 and 3). And of course we can
44
,4, Schohens
Planning in ordinary conversation
FINDING A JOB I AS A TEACHER I
AND
APPLY
SEARCH FOR APPROPRIATE VACANCIES
J T A K E THE F I R S T VAC. T H A T IS RECOMMENDED
S C A N SEV. ISS. O F TEACHER'S JOURNAL
ASK ACQUAINT (in gen)
Fig. 5. The total underlying plan structure in which both the plan initiator's and the recipient's proposal are represented.
choose a level under it and expand one of the nodes which were generated just before (figure 4: 3). In fact the sketchy script (figure 3) is so basic that there will be hardly any chance that people disagree upon it. But the further the steps are worked out, the more people come to disagree. Looking at figure 5 we can clearly see that the expansion of the nodes gets
A. Scholtens / Planning in ordinary conversation
45
more personal as we come lower. The first level consists of the two nodes that can be said to be fairly common. But already one level under it the discourse participants came up with different ideas. And, of course, still one level deeper ('ASK H E', 'ASK J', 'ASK M') the expansion is really personal, in this case strictly limited to a certain circle of friends. The expansion of the root node will in most cases belong to the sketchy scripts. Maybe in some cases the same holds for the level under it, but in general we can say that the further we move away from the top the further we move away from the sketchy script and the closer we come to a filled-in one.
2. Analysis of the chosen fragments
2.1. Ranking of the recipient reactions In section 1.5 I mentioned following and steering recipient reactions. I did not mean by that however that all recipient reactions can be divided in purely following and purely steering reactions. It seems that they can be graded from really following to really steering, with different types in between. For the recipient reactions in the 17 fragments (section 1.2) under discussion here I suggest a ranking as follows:
(l) hearer signals; (2) question to the plan initiator to formulate the factual representation of affairs of her plan in a clearer way; (3) question to the plan initiator to clarify the argumentation for her plan; (4) contestation of the factual representation of affairs of the plan initiator's plan; (5) contestation of the argumentation for the plan initiator's plan; (6) advice or suggestions for the plan initiator's plan. It was suggested earlier that this range of reactions from following to steering can be related to the amount and the nature (diverging or not) of the previoh~ knowledge of the recipient (1.5). Within the lop.ger fragments these reactions can be described as pertaining to certain phases in information processing. The following' reactions occur in the earlier stages (where the conversa tional partners don't have equal access to the relevant information), while the more 'steering" reactions 3ccur in later stages (where a balance in access to relevant information has been more or less installed). But sometimes the recipient already possesses much relevant previous information at the beginning of the plan iriitiator's exposition. In that case the more steering reactions are not exclusive for later stages in the conversation; the recipient can come up with these very quickly. Below I will discuss why a certain reaction has been labeled following or
46
,4. Schottens
Planning in ordinary conversation
steerhtg. Here I will represent bits of the planning fragments studied as an illustration. These fragments have been incorporated in the text and numbered as example 1, 2, etc.
2.2. Illustration of the ranking of recipient reactions 2.2.1. Hearer signals This reaction indicates that the recipient is busy collecting data. In addition to minimal responses such as 'yes', "hm', etc, we also have the more extended hearer signals in this category, such as: affirmation, paraphrasing, expansion, summary, conclusion.
2.2.2. Question to the plan imtiator to formulate the factual representation of affah's of her plan #~ a clearer way In the planning fragments it occurs frequently that the recipient asks the plan initiator for extra information. The recipient presumably does not have a clear image yet of the plan and wants clarification on some points.
Example 1 From a conversation between two physical therapists. The plan initiator tells the recipient about her plans to move. The recipient wants more information about the new house. no he has I have sent off that letter, in which I asked whether there was a renovation taking place etc. or at least that I wanted to know first and you don't know yet how much rent you have to pay ? don't know that either and then you must have a residency permit for it? no:, oh god, can't be, I don't have anything //'((laughs))] (.) not the right kind of papers
//((laughs))]
no
A. Schohens / Planning in ordinary conversation
47
The recipient uses her knowledge of the sketchy script renting a house in posing her questions; she knows what occupying a house involves and that 'the level of rent' and 'residency-permit' are relevant matters here. However, requests for extra information of the recipient are not simply indications of the fact that the plan initiator has sketched an incomplete picture. Apart from the fact that the recipient can pose these questions tentatively just to show her eagerness, these requests also say something about the recipient's priorities. The recipient's questions also make clear to which points from the sketchy script the recipient wants to give more weight as recipient. So there is a sketchy script underlying a planning fragment - more or less at the disposal of both the conversational participants - but the choice which of the components from this sketchy script is really verbalised is made in the first place by the plan initiator but also by the recipient. This is exactly what was meant in section 1.1 with the remark that recipients in conversations can influence the staging (as opposed to monologues where only the speaker controls the actual staging).
2.2.3. Question to the plan initiator to clarify the argumentation for her plan Sometimes the recipient poses questions about the background motives of the plan initiator's plan. Apparently it is not clear for the recipient then why the plan initiator is going to execute certain components of the plan or why she is going to do that in that way. Example 2 From a conversation between two housewives. In this part of the conversation the plan initiator recounts that an acquaintance of hers (Axle) is going to cut a wooden board for her table. The recipient asks for the argument to put a wooden board on the table.
//yes but he is a bit bigger] than yours, //isn't he?] (.) why does he put wood on it?
I think, I estimate th/ this one is about 90 cause //mine is a bit smaller] //yes]
I don't need a glass table //no but does he, does he plate any more//I have glue] formica on it? had Patty's splinters] (Patty is the plan initiator's daughter who has broken the previous glass table.)
48
A. Scholtens / Planning #1 ordinao' conversation
In this example the staging by the recipient is heavily influenced by her own filled-in script as we can see very well when we follow the above fragment (Example 2) further on:
Continuation of Example 2 uh? yes 'f course I am going to glue something on it, I can't just have a bare board, that's unsightly, now, if you sandpaper it nicely it's not bad ram? if you sandpaper such a wooden board it is nice as well,//if you polish it ] and then you sandpaper it//smooth] and then with eh lacquer //on top] is nice as well
//what do you mean?] //just] with eh //varnish] is also nice is easy/- you have to be able to wipe it clean with a wet cloth
you can do that also when it is lacquered The recipient does not ask further adstruction for the choice for a wooden board out of sheer anxiousness. It is clear that she has a very exact idea of her own of what should happen and what would be best. This adds up to an advice to the plan initiator. Reactions 2 and 3 are on the one hand an indication of the fact that the recipient does not yet have enough information (or does not consider herself to have enough information). On the other hand, it seems that the recipient often comes to these eliciting remarks on the basis of an already formed preliminary image. The latter can be seen well in the above fragment (continuation of Example 2), but is perhaps clearer still in the sample-fragment: lines 190-215. The plan initiator mentions there at a certain moment that it is not nice to treat 'song texts' in school in the frame of an English class because that really does not go well with the students. The recipient then reacts:
A. Scholtens / Planning in ordinary conversation
49
'What kind of classes? 'cause that may differ from case to case' In the first instance this is a request for further information (what kind of classes?). The addition seems only a motivation in principal for the first question. After the plan initiator has given the answer the recipient adds more however: 'yes (.) no uh I can imagine that that it does not work in a mavo 3 or 4 mayo, I experienced that myself in the GV-school, but maybe it works - it is extremely elitist - but that it does work in 4 or 5 gymnasium.' From what follows it appears that the recipient poses the question (what kind of classes?) from her own idea about this question, partly based on her own experience (I experienced that myself in the GV-school). In certain cases the recipient's own opinion is so transparent in the question that the question is more like a contestation. In that sense the above question of the recipient (what kind of classes?) can be paraphrased: 'You can't say in general that song texts don't go well with students; when you make a claim like that you have to specify for which kinds of classes that claim really holds'.
2.2.4. Contestation of the factual representation of affairs of the plan initiator's plan When the plan initiator is unfolding her plan, the recipient sometimes contests the plan initiator's representation of the state of affairs.
Example 3 From a conversation between two housewives. The plan initiator tells the recipient that her mother would like to have a lean-over door in the weekend cottage, but that uncle Arie is not so keen on that extra job. The plan initiator's mother knows exactly how it should be done and the plan initiator tells the recipient this:
//that's also possible]
//now you don't need you don't ] need a handle in the upper part, 'cause in the lower part you take,
(.) with with a handle above and below //my mother says] that that seems so easy to me I am a woman but //that I would do in a jiffy]
50
A, Scholtens / Planning h~ ordinao' conversation
you take a spring-bolt inside (3 yes such a sliding one "/and above an ordinary handle ]
yes, and in the upper part //an an ordinary handle] with a sliding lock so that you can close it of
//yes]
//course]
The recipient combats the idea that one would need two door handles on a split door; one on the upper part is enough. The knowledge the recipient rests upon here is in fact 'frame knowledge', since we cannot easily use the term 'script knowledge' referring to knowledge about how a split door is constructed. For the line of argument this does not make any difference however,
2.2.5. Contestation of the argumentation for the plan initiator's plan Not only the factua', representation of affairs given by the plan initiator is subject 'o contestations on the part of the recipient. In addition to this sometimes also the plan initiator's argumentation is scrutinized.
Example 4 From a conversation between two physical therapists. The two womeT>, discuss the possibility of going to work abroad for a few years. my god, even if I don't have a stable partner that I leave behind, you do have other friends and so on, //that's also] (.) and your parents, that is also pretty eh, //fairly close] ties
//no: (.)ram:]
//yes of course]
//yes: but] that is something, that that mm (.) that affects you or so, but I don't believe that that should be a//reason for not going]
A. Scholtens / Planning in ordinary conversation
51
then they can say all they want mm
The plan initiator states that the fact that her relation to her parents is still pretty close is a reason for her not to go abroad for a long time. The recipient claims that this circumstance should not be a reason to give up such a plan. Another example of reaction 5 can be found in the sample fragment: lines I !-37. Simplified this sequence looks like this: Plan initiator:
but when once I started with that program then I imagine that you can be teaching on the side for a while in fact you can work also just like that can't you? Recipient: yeah but that's because in our case the Masters' program is so free Plan initiator" yeah but it is like that over there as well: besides I really want to do that (---) when I have to continue the way I do now (---) then I still don't know whether I can teach, then I have to start with it at that point (---) Recipient: yes yes that's true The actions which the plan initiator and the recipient undertake here can be indicated as follows: Plan initiator: t!
Recipient: Plan initiator: |!
Recipient:
plan first argument objection to the plan initiator's first argument objectioa to R's objection second argument agreement with the plan initiator's second argument
The plan initiator uses a very special argument in this example to add support to her position, which is that studying and teaching can be combined. She uses her conversational partner as a point of comparison, as a norm: you can do it so therefore I can do it. The recipient 'contests' the validity of this comparison by explaining in which way her situation is different. Subsequently the plan initiator contests that the recipient's situation is essentially different. She considers the recipient a valid point of comparison. The recipient does not attack the plan initiator's claim here that studying and teaching can be combined. Neither does she try to keep the plan initiator
52
A, Scholtens / Plam~ing in ordinary conversation
from her proposed step. The recipient purely directs herself at the argumentation for this proposed step.
3. Concluding remarks The main point in the sections 1.1-1.7 was the question which kinds of knowledge people need to process conversations and also to contribute to conversations. Subsequently an attempt was made to link the theories about the necessary knowledge and the concrete recipient reactions as they appear in the material studied. In section !.7. was shown, more particularly, how an underlying planning structure, elaborated by both participants, emanates from the concrete moves in the ungoing conversation. This underlying structure appears to consist of elements of a so-called sketchy script (in this case the script 'looking for a job as a teacher'), but also elements of filled-in scripts. In the sample fragment (see appendix) the participants each seem to work with their own filled-in script. Finally the plan initiator takes over the recipient's filled-in script, Such a recipient reaction we call the "most steering" reaction, since the recipient appears to contribute to the plan on the level of content. In section 2 an overview is given of the other recipient reactions, which don't influence the plan structure on such an elementary level. In these cases the recipient only questions or contests the facual representation or the argumentation made by the plan initiator. These reactions are graded from 'following' to 'steering' on the basis of the impact they have on the planning structure. It is often stated in the literature (section 1.4.1) that the recipient's interpretation of discourse results in intermediate interpretation frames on the one hand and on the other hand is steered by such frames. The previous knowledge of the recipient is often seen as a "sketchy script' only, on which the participants agree, and too little attention is given to the fact that this previous knowledge could be departing from the incoming information - that there can be differences in the filled-in scripts. In such a case the recipient's previous knowledge causes a conflict which will not be noticed if this recipient is processing a monologue text but which will be very important if this recipient participates in a dialogue. The several phases a recipient passes through in processing a piece of discourse, in this case the reporting of a plan: collecting information, signalling obscurities or falsities in the information or in the argumentation, signalling departing ideas, etc., are reflected in the concrete recipient reactions. It is argued that a closer examination of these concrete reactions could be useful in theorising about the way participants in discourse process conversation.
A. Scholtens / Planning in ordinary conversation
53
Appendix Sample-fragment Participant I
Participant 2 but you, uh, but, if you are going to teach in August or September,
mm
for a couple of hours, then it is pretty heavy in any case to start something on the side, whatever it may be yes but then I would um that is that is not true, look, what I have to do now, is very much of course 10
mm
but when once I started with that program then ! imagine that-that you can be teaching on the side for a while, in fact you can work 15 also just like that can't you yeah but that's because in our case the Masters' program is so free yeah but it is like that over //there as well] 20 yeah//well they'd like you] to finish it in two and a half years but when, I imagine that if I do it in three years or three and a 25 half, that I can be teaching for six hours or eight hours, besides I really want to do that otherwise I don't want to study, when I have to continue the way 30 I do now then I'm stuck like this for another two and a half years, then I still don't know whether I can teach, then I have to start with it at that point 35 so I really want it now
//that you can do it] in the number of years that you like //well that's not true all together]
m
yes yes that'~ true by the way do you know a Christian comprehensive school in Buitenvel40 dert? yes the Christian Comprehensive school
A. Scholtens / Planning in ordinary conversation
54
Buitenveldert what type of school is that? well, a little fzctory, not really (h) well somewhat nice
45 since there was a vacancy announced in the teachers' magazine but my mother sent it to me, but it didn't arrive
that's a good reason not to do it ((laughs))
50 (h) yes but she said it was written that it was such an extensive position that also people who had 55 only
yes
less time (.) so you can imagine that there are/,/35 hours] English second degree (h) but do 60 you think you have to be a Christian for that, is it a ponderous school?
//but in August?]
well. 1 thought no: no I know someone who, a girl who attends it, who is not from a Christian family but, that does not mean a good deal of course
65 hm
no: I don't think so, that's almost, but (.) but shouldn't you wait a little or do you really think that there will be such a scarcity of jobs?
70 (h) I don't know so much about it really (.) I keep on thinking I must do some//thing] 75
no according to me they have it 80 in our institute as well but I just haven't looked in it at all my mother gets it delivered so ,//she] looks in it every time and every time you get the most 85 impossible jobs offered (h) by her (h)//like uh] (h) yes but I don't know whether
//but] but don't you regularly go to t~e library on the Keizersgracht or I don't know where, it could also be in the Universlt) Library, where they have the teachers' magazine?
//yesl
//'cause of course] it could be, I mean, there must be nicer schools to think of
A. Schohens / Planning in ordinary conversation
there are so many jobs 90 //given what, what I want] because I really want something peculiar I don't want a full-time job I don't even want half a j o b (.) I really want less than//that] 95 = really, ram, I prefer what you had
//no:]
55
n o / / I don't know that either no idea]
//a couple oq spare hours somewhere = (.) hm-yeah did you mention it to HansErik for instance?//that's also/very clever to spread this about to people who eh:
100 (h) I will do that because that's a nice school as well isn't it? yes, certainly (.) oh there's really something in //what you say] 105
!10
/,/just not not], just that he may thir.k of it some time when there are a few hours !eft, and then it is pretty close or, ah well, rather close when when they are going to move from Buitenveldert to what's it called? the Ignatiusschoo! then it is rather close
where is it?
oh: well that's//altogether] 115 //yes, yes that's true] //I will do that] 120 yes: and then then I know him as well, so I don't feel//quite uh]
125
on the Stadhouderskade on the corner of the Stadhouderskade //and somewhatl (.) well really close that's opposite Frits that's why Frits said//that he was] going to teach there(.) well I would in any case mention it//'cause that's nice], or that's to say, I liked it over there (.) (.) I would do it //yes oh God] and Jaap and and: what's the name of that former friend of Mart's, she is teaching English over there, Marijke (.) Marijke I don't know her name (.) you should ask Mart
mm Mart is her former best friend (.) very nice girl, who teaches English 130 m m not a full-time job either (.) but God you never know yes 135
I mean, when it isn't this year maybe it is next year
A. Scholtens / Planning #2 ordinary conversation
56
yes (.) ah in any case 1 want to do it next year = =ram.=
= and I don't want to (h) I want to 140 have something like that, not to crack up completely if it is going wrong here at the institute, //you know] 'cause it looks like that, because I'm so very busy (h) 145 'cause I can't think (.) I should really regret it if I don't pass those exams, and it looks like that
//yesl
(.) yes, God, that's bad 150 that is very bad, but they are teaching so fast really (3 and, oh well I don't know yet (,) and I want to find out about teaching if it works a n d / / t r y it 155 for a year] (.) (h) yes: well with Dutch class it wcnt wcll, yes Dutch class that's another possibility
yes,//well, yes]
//yes and your teacher-training] that was airight?
you have but are you qualified third degree?
160 yes
that's silly, isn't it? yes: that's crazy but well on the certificate it said that the exam 165 was held on the third degree level so it must be third degree but still it is enough for a comprehensive school, the first two classes is uh//is] OK uh for 170 third degree (h)
//yes] yes
and for Dutch class I think it is alright you know, for English class I would like to tell some 175 more but for Dutch class (h) yes that is very free, and then I'd like to do a puzzle with those little ones ((laughs)) or at least there are so many 180 possibilities with//Dutch class, I like that]
((laughs)) //yes: you can do a lot more (.)] you can read the paper with them and
A. Scholtens / Planning in ordinary conversation
57
SO o n
(.) yes and I am rather inspired 185 to do that, for English lesson I feel that you must have gone pretty far and then it is nice yes since then I can tell something 190 about literature and then you can do something with song texts etc. yes but, I have tried all that but it didn't work because there is 195//simply] //well I tried it on] //well I tried it in pretty high]
//yes but] where didn't it work? //what kind of classes?] //'cause that may differ from case to case]
classes I tried it on Berlage200 comprehensive (h) well then I had a fifth class of the gymnasium yes well I really had great difficulty in talking about a rather simple 205 poem, and then I thought well if that is the case, then I can look forward to doing literature, but yes they look at you like this 210
yes (.) no uh I can imagine that that it does not work in a mavo 3 or 4 mayo, I experienced that myself in the GV-school, but maybe it works - it is extremely elitist - but that it DOES work in 4 or 5 gymnasium
215 mm
that kind of subjects DO exist really (.) but well and song texts I always like very 220 much yes, but explaining the elementary grammar (.) well no thanks I don't like that no
58
A. Scholtens / Planning in ordinary conversation
References Abelson, Robert P., 1973. The structure of belief systems. In: R.C. Schank and K.M. Colby, eds. Computer models of thought and language, 287-339. San Francisco: W.H. Freeman and Company. Atkinson, J. Maxwell and John Heritage, eds., 1984. Structures of social action: Studies in conversation analysis. Cambridge: Cambridge University Press. Brown, Gretchen P., 1984. Linguistic and situational context in a model of task-oriented dialogue. In: L, Vaina and J. Hintikka, eds,, Cognitive constraints on communication, 229-240. Dordrccht: Reidel. Brown, Gitlian and George Yule, 1983. Discourse analysis. Cambridge: Cambridge University Press. Clements, P., 1979. The effects of staging on recall from prose. In: R,O. Freedle, ed,, New directions in discourse processing, 287-329. Norwood, N J: Ablex. Cohen, Robin, 1987. Analyzing the structure of argumentative discourse. Computational Linguistics 13: 11-24. Ensink, Titus, 1986. Over interpretatie. Een studie in cognitieve sociolinguistiek. Dordrecht: Foris. Grimes, Joseph E., 1975. The thread of discourse, The Hague: Mouton, Grosz, Barbara J,, 1978. Discourse knowledge. In: D, Walker, ed., Understanding spoken language, 229-347. New York: North-Holland. Grosz, Barbara J, and Candace L, Sidner, 1986, Attention, intentions and the structure of discourse, Computational Linguistics 12: 175-204. Labor, William and David Fanshel, 1977. Therapeutic discourse. Psychotherapy as conversation. New York: Academic Press. Lakoff, Robin, 1973. The logic of politeness: or, minding your P's and Q's. In: Claudia Corum et al., eds., Papers from the Ninth Regional Meeting of the Chicago Linguistic Society, 292-305. Linde, Charlotte and J.A. Goguen, 1978. Structure of planning discourse. Journal of Social Biological Structure 1: 219-251. Linelt, Per, Lennart Gustavsson and Pfiivi Juvonen, 1988. lnteractional dominance in dyadic communication: A presentation of initiative-response analysis. Linguistics 26: 415--422. Minsky, Marvin, 1975. A framework for representing knowledge. In: P.H. Winston ed., The psychology of computer vision, 195-228. New York: McGraw-Hill. Polanyi, Livia, 1979. So what's the point? Semiotica 25: 207-241. Rich, Elaine, 1983. Artificial intelligence. New York: McGraw-Hill. Riesbeck, Christopher K., 1985. Realistic language comprehension. In: A.M. Aitkenhead and J,M. Slack, eds., Issues in cognitive modeling. London: Lawrence Erlbaum Associates. Ross, Robert N., 1975. Ellipsis and the structure of expectation. San Jos6 State Occasional Papers in Linguistics i: 183-191. Schank, Roger C. and Robert P. Abelson, 1975. Scripts, plans, goals and understanding. New York: Lawrence Erlbaum. Tannen, Deborah, 1979, What's in a frame? Surface evidence for underlying expectations. In: R.O. Freedle, ed., New directions in discourse processing, 137-182. Norwood, N J: Ablex. Van Dijk, Teun A. and Walter Kintsch, 1983. Strategies of discourse comprehension. New York: Academic Press. Wilks, Yorick, 1986. Relevance and beliefs. In: T. Myers, K. Brown and B. McGonigle, eds., Reasoning and discourse processes, 168-185. London: Academic Press.