Film audiences as overhearers, Claudia M. Bubel

Film audiences as overhearers

Claudia M. Bubel *

Saarland University, Fachrichtung 4.3 Anglistik, Postfach 15 11 50, 66041 Saarbrucken, Germany

Received 13 July 2004; received in revised form 7 September 2007; accepted 1 October 2007

Abstract

This paper attempts to describe the workings of film discourse by focusing on the cognitive processes

involved in the comprehension and design of verbal exchanges on screen. Several concepts have been

developed to account for various forms of mediated discourse (Burger, 1984; Short, 1981; Clark, 1996);

however, these concepts do not explain how the audience co-constructs meaning (Duranti, 1986) and they

fail to spell out the mental engagement involved in this co-construction. The approach introduced here

accords a central position to the spectator, who acts as an ‘‘overhearer’’ in the sense of Goffman (1976,

1979). The cognitive processes that lead to an understanding of film dialogue are parallel to those happening

in the case of overhearers in everyday situations. This also has consequences for the design of film dialogue.

Both understanding and design of film dialogue are based on underlying patterns of knowledge, including

knowledge about real conversation and film dialogues.

# 2007 Elsevier B.V. All rights reserved.

Keywords: Film discourse; Participation framework; Layered discourse; ‘Overhearer’; ‘Ratified participant’

1. Introduction

Although scholars of both film and human communication have pointed out that linguistic

analyses of film dialogue promise to be rewarding, few people have followed up this suggestion

systematically. Film and media scholars, on the one hand, have occasionally fallen back on

linguistic theories to make up for the lack of scholarly treatment of film dialogue in their own

field (cf. Kozloff, 2000). Language scholars, on the other hand, have sporadically used film

dialogue as data, either when naturally occurring data has not been accessible (cf. Tannen and

Lakoff, 1994), or when they have coincidentally come across a spate of movie or TV dialogue

which served as a suitable example for their line of argument (cf. Hopper and LeBaron, 1999).

www.elsevier.com/locate/pragma

Available online at www.sciencedirect.com

Journal of Pragmatics 40 (2008) 55–71

* Tel.: +49-681-302-3009/938 7236; fax: +49 681 302 4623.

E-mail address: [email protected].

0378-2166/$ – see front matter # 2007 Elsevier B.V. All rights reserved.

doi:10.1016/j.pragma.2007.10.001

mailto:[email protected]

http://dx.doi.org/10.1016/j.pragma.2007.10.001

Finally, film dialogue has been investigated from a stylistic point of view, for example, with

respect to its function in characterization (cf. Culpeper, 2001), or when comparing film versions

of literary works (e.g. Kobus, 1998; Simpson and Montgomery, 1995). One of the reasons for this

rather incidental treatment may be the lack of an approach that tackles the workings of film

discourse and its perception. In this paper, I will present such an approach; I will focus on the

audience’s comprehension of verbal events on screen by considering the spectators as

‘overhearers’, in the sense of Goffman (1976, 1979).

Several concepts have been put forward to account for related modes of mediated discourse,

including dramatic and electronic mass media discourse such as TV and radio. Most of these,

however, are based on the message model or container model of communication, which has been

widely criticized (cf. Arundale, 1999; Akmajian et al., 1980; Krauss, 1987; Reddy, 1979).

In particular, mass media studies rely on this simplified mechanistic model (cf. Klemm, 2000:83;

Scollon, 1998:17). As a consequence, these theories neglect the fact that the audience

co-constructs meaning (Duranti, 1986). The notion of co-construction ties in with the media

studies concept of the ‘active audience’, which implies that the audience is always active and that

media content is always open to interpretation (cf. Klemm, 2000:102–105; Morley, 1994).

In the following, I will present three approaches to various forms of mediated discourse and

use them as a basis for a theoretical description of how an audience arrives at an understanding of

talk on a screen. Doing this, I will refine the existing approaches by integrating Goffman’s

participation framework (1976, 1979) and extensions of this by Clark and his co-workers (Clark,

1996; Clark and Carlson, 1982; Clark and Schaefer, 1992). Taking a cognitive perspective—in

line with a call for a cognitivist approach in film theory (cf. Giora and Ne’eman, 1996; Bordwell,

1989) – I will attempt to explain how the audience comprehends film dialogue and how film

dialogue is designed for the audience.1

2. Approaches to mediated discourse

2.1. Burger’s communication circles

Burger (1984:44ff.) presents a model that describes the structure of dialogues on TVor radio. He

argues that the participants in the dialogue do not talk to each other only, but are always

communicating with the audience that sits in front of the TV or the radio. Structurally, he

distinguishes two communication circles: an ‘inner circle’, in which the dialogue is taking place

(the ‘primary situation’) and an ‘outer circle’ constituted through the relationship between the

participants in the dialogue and the audience in front of the TVor radio (the ‘secondary situation’).2

If a participant of the inner circle dialogue directly addresses the audience in the studio or the

audience in front of the TVor radio, the circle is broken. Burger (1984:44) gives the example of the

skilled politician who, in a TV interview, while answering questions, does not look at the

interviewer, but into the camera and thus at the potential audience in front of the screen. The

distinction between inner and outer circle entails, in Burger’s (1991:7) terms, ‘multiple addressing’,

i.e. the participants to a conversation in the media simultaneously address recipients in the inner and

the outer circle. This affects the nature as well as the course of a conversation (Burger, 1991:7).

C.M. Bubel / Journal of Pragmatics 40 (2008) 55–7156

1 As one of the referees of this paper has rightly suggested, the visual and musical accompaniment are also highly

significant to the perception of film dialogue; however, it would take me too far afield to start discussing this aspect.2 Burger (1984:44) applies his concept mainly to talk shows or interviews on TV and radio, so that there can even be a

third circle in the case of a studio audience being present.

Burger’s concept can be applied to film and TV drama dialogue insofar as we can distinguish

between an inner circle, i.e. the action on the screen and an outer circle, the perception by the

audience of what happens on the screen. The participants on the screen address each other and at

the same time, they indirectly address the audience that is sitting in front of the screen. For film

dialogue, this concept needs to be extended, as there are further participants to be taken into

account, such as the film production crew who design the dialogue of the inner circle for the

participants in the outer circle. Furthermore, Burger simplifies the process of communication to a

simple circle where A addresses B and B decodes the meaning of the message sent by A. The fact

that the participants negotiate meaning in both the inner and outer circles is not taken into

account.

2.2. Short’s embedded discourse

Just as Burger uses the metaphor of one communication circle within another, Short (1981,

1989, 1994) describes dramatic discourse in terms of embeddedness. The playwright addresses

the audience on one level, but there is another level embedded in this first one; this is where a

character A addresses a character B. The embedded level is part of the message that the

playwright communicates to the audience (see Fig. 1).

In Short’s own words,

This ‘doubled’ discourse helps us account for dramatic irony, and in general for the way in

which we know that when we listen to two characters talking on stage we are meant to

deduce, through what they say, what the author is telling us about them in terms of

characterization, etc. (Short, 1994:950).

This model seems to work for film discourse as well. For example, an early scene in the film

Career Girls (Leigh, 1997) presents the first meeting of the two main characters, Annie and

Hannah, in a flashback. Hannah’s upper-crust Received Pronunciation (RP) and her exaggerated

enunciation, followed by the sudden code-switch from RP to a broad London Cockney accent,

works on two levels. By using RP, Hannah is contextualizing the situation as formal; the switch to

Cockney indicates that she is now ready to get personally involved with Annie (cf. Gumperz,

1982). From this, not only her direct addressee, Annie, draws conclusions about Hannah’s

character, but so does the cinema audience, (on what Short calls the overlaying discourse level).

The same holds for Annie’s hesitating way of speaking and her Northern accent: not only Hannah

and Claire, her flatmate, but also the audience watching the interaction on the screen are in this

way able to infer something about Annie’s background and character. The messages exchanged

between Annie, Hannah and Claire are part of the message that Mike Leigh, the screenwriter and

director, addresses to the audience.

C.M. Bubel / Journal of Pragmatics 40 (2008) 55–71 57

Fig. 1. Short’s embedded discourse (Short, 1989:149).

(example 1) Career GirlsAnnie, first year psychology student looking for a flat, rings a doorbell; Hannah opens the door.

1 H oh, hello? ((upper crust RP))

2 A hello.

are you .. eh .. Hannah? ((Northern accent))

3 H it’s Hanna:h actually.

4 A oh .. right .. ((clears her throat))

5 H and this is Cla:

6 H yea:h?

do come in Anja.

7 A thank you.

8 H all right? ((broad Cockney accent))

Even so, the application of Short’s model to film discourse is limited. As Betten (1977:360)

states, film has additional levels of communication: since there is no direct link between the

message of the screenwriter/director and the audience, other factors, such as the cameras and the

editing process, have to be dealt with.

Furthermore, Short’s approach is redolent of the message model of communication, where the

role of the listener in the co-construction of the meaning of the interaction is not taken into

account. Although the cinema audience normally does not take an active part in the interaction –

with the exception of movie premieres, when the audience applauds the filmmaking crew that is

present on stage, and may ask them questions and offer comments – it is still taking an active part

in the construction of meaning. As Duranti expresses it:

. . . interpretation (of texts, sounds etc.) is not a passive activity whereby the audience is just

trying to figure out what the author meant to communicate. Rather, it is a way of making

sense of what someone said (or wrote or drew) by linking it to a world or context that the

audience can make sense of (Duranti, 1986:243–244).

Only with the help of the audience’s knowledge about the accents and dialects of the British

Isles, as well as about the kind of characters that hesitate in speaking, switch codes, and over-

enunciate can the interaction in the Career Girls scene quoted above function as an introduction

of the characters.

2.3. Clark’s layered discourse

Clark’s theory of layering in discourse (1996) is more suitable for explaining the complexities

of dramatic and cinematic discourse. His model resembles Short’s approach in so far as Clark,

too, distinguishes several levels of discourse on which events take place. These levels are called

‘layers (or domains) of action’, and they are characterized by the participants, their roles, the

place, the time, the relevant features of the situation, and the possible actions (Clark, 1996:355).

In the above example from Career Girls, one layer is the first meeting of Annie with her future

flatmates Hannah and Claire. The participants are three first-year students, the place is a drab

terraced house in London, the time is sometime in the 1980s. Annie is looking for a flat and

Hannah and Claire are looking for a flatmate; Annie visits the others to have a look at their flat

and they meet for the first time. On another layer, the movie actors Lynda Steadman, Katrin

Cartlidge, and Kate Byers pretend to be Annie, Hannah, and Claire respectively. The place is the

film set in London, the time is the moment of shooting the film. The participants are enacting


roles that they have developed under the direction of the moviemaker Mike Leigh. A camera

team films the action and the film material is later edited. In yet another layer, a projectionist or a

TV station may present the story of Annie, Hannah, and Claire, as told by the screenwriter,

directed, filmed, and edited by various people in the production crew, to an audience which

pretends that the actors indeed are Annie, Hannah, and Claire. In this layer, the participants are a

projectionist and an audience; the time and place are, for example, the 1997 Edinburgh film

festival.

Schematically, this layering takes the following form:

Layer 3: Characters interact.

Layer 2: The film production team, the actors and the cinema audience jointly pretend that

events in layer 3 take place.

Layer 1: The projectionist and the cinema audience jointly realize layer 2 (alternatively,

the audience in front of a TV and the TV station jointly realize layer 2).

The layers in the model being recursive, one could argue that the filming and the editing

comprise extra layers. This recursiveness makes the model more suited to the complexities of

film discourse than does Short’s embeddedness. Short (1989:149) concedes that there can be

more than two levels, but only within the structure of the play, as, for example, when one

character relates to a second character the words of a third character. His approach is not intended

to integrate factors involved in the production of a play, except for the roles of the playwright and

the recipients; the director, for example, is not taken into account.

Comparable to Short’s notion of embedding one level within the other, Clark argues that

layer 1 is the foundation or base and that, consequently, the participants of this layer are the

primary participants. The higher layers are like theatrical stages created on top of the first layer

(Clark, 1996:355).3 The audience, however, does not pay most of its attention to the basic

layer; rather, they focus on the topmost layer. In this connection, Clark distinguishes between

two principles that hold in layered actions, ‘‘imagination’’ and ‘‘appreciation’’ (Clark,

1996:358ff. and 366ff.):

Principle of imagination: In layered actions, the primary participants are intended to

imagine what is happening in the highest current layer of action.

Principle of appreciation: In layered actions, the primary participants are intended to

appreciate the instigator’s purposes and techniques in creating the highest current layer of

action. (Clark, 1996:359)

Imagining the topmost layer is what engrosses the audience. One effect of film is that it

transports the audience into the realm of the story and, doing that, evokes emotions and

suspense. Clark here does not make explicit the cognitive processes involved, but how

imagination works is documented in the following experiment by Morrow and Clark (1988:282–

285),which shows how people create imaginary representations even when interpreting a single

word, such as approach. The word’s meaning depends on how near to a landmark the recipient

thinks an object has to be in order for the object to be in some kind of proximity relation to the

landmark.


3 Clark himself objects to the notion of embeddedness: ‘‘Although this metaphor is useful for some purposes, it can be

misleading’’. He argues that layer 2 is not a proper part of layer 1 (Clark, 1996:355). Annie and Hannah, for example, are

not part of the 1996 Career Girls film shooting situation.

(1) I am standing on the porch of a farm house looking across the yard at a picket fence.

A tractor is approaching it.

(2) I am standing on the porch of a farm house looking across the yard at a picket fence.

A mouse is approaching it.

To comprehend the sentence the interviewees imagined the situation and described it

accordingly. Whereas they estimated the distance between the fence and the tractor to be around

39 ft, the distance between the mouse and the fence was estimated at only two feet. Depending on

the size of the advancing object, approach is interpreted differently. The interpretation is based on

world experience, on knowledge of the size of tractors and mice, and on experiences with other

approaches. This means, firstly, that ‘‘imagination is needed for even the simplest descriptions’’

(Clark and Van Der Wege, 2001:774), and secondly, that imagination entails linking up descriptions

given, for example, in a narrative with what we have experienced in the world, i.e. with patterns of

knowledge stored in our mind. In this way, Clark’s theory also takes into consideration the role of

the cinema audience in the co-construction of meaning: imagination only works through linking up

what is presented on the screen to the audience’s world knowledge.

The other concept, appreciation, also illustrates how the audience is involved in the co-

construction of meaning. It comprises what the audience does on the lower, more obscure layers,

such as when they recognize how screenwriters and film directors, camera, and editing staff

achieve certain effects (for example, the first, introductory scene of Career Girls has a bluish

tinge to mark the scene as a flashback, representing one of the protagonists’ remembrance).4

The principle of appreciation also accounts for scenes which, in Kozloff’s (2000:56ff.) terms,

contain thematic messages, authorial commentary, or allegories. The inclusion of these elements

. . . breaks the illusion that viewers are merely overhearing characters talking to one another;

it makes plain that the dialogue is addressed at the audience. This both violates the suspension

of disbelief and ‘catches’ the viewer in the act of eavesdropping (Kozloff, 2000:57).

For example, in the film Secrets and Lies (Leigh, 1996), the protagonist Maurice gives a

speech to his assembled family (see example 2). This speech contains the film title, which

engages the viewer in appreciation rather than imagination. The film title is part of layer 1, in

which the audience is watching the movie. They have seen the film poster or the announcement in

the newspaper or TV magazine and are aware of the title, so that when the title is mentioned in the

dialogue in layer 3, the otherwise obscure layers surface into the viewer’s consciousness. The

mentioning of the title is set off from the surrounding discourse not only through forceful

enunciation but also through bracketing it between long pauses (5 and 2 s, on either side). It is as

if screenwriter/director Mike Leigh addresses the audience directly, admonishing them to be

open and truthful, and to share their grievances with their friends and family.

(example 2) Secrets and lies1 M there.

(3.0) I’ve said it.

so where’s the bolt of lightning?

(5.0) secrets and LIES.

(2.0) we’re ALL in pain.

why can’t we SHARe our pain?


4 This of course adds another layer: Annie is remembering the events taking place on the topmost layer.

3. An audience-centred model of film discourse

So far, I have presented three models that have been used to explain the workings of various

forms of mediated discourse. I consider the common feature of their approaches, based on either

embeddedness or layered levels, as useful to explain certain phenomena of cinematic discourse,

for example, how characterization works through verbal interaction between the protagonists, or

how the audience imagines and appreciates movies.

In the following, I will integrate the notion of layeredness into a model of cinematic discourse.

I will focus on the processes occurring on the underlying levels of discourse, namely the acts of

imagination and appreciation by the spectator and the screenwriter and his colleagues’ way of

designing utterances for the spectator. I argue that the cognitive processes in ‘screen-to-face’

discourse are generally parallel to the processes we experience when we overhear conversations

in everyday life and that this is taken into account when a film dialogue is constructed. To

document the processes we experience in everyday life, I will first present the concept of

overhearing as sketched by Goffman (1976, 1979) and further developed by Clark and Carlson

(1982), Clark and Schaefer (1992), and Clark (1996).

3.1. Overhearing and designing utterances for overhearers

3.1.1. The spectator as overhearer

The key concept to an understanding of the workings of cinematic discourse is

‘overhearing’.5 The cognitive processes going on in the spectator while he or she is listening

to film dialogue are generally parallel to those that occur in everyday life, when we take on the

role of an overhearer, whether or not the conversation we are overhearing is meant to be heard by

us, and whether or not the conversationalists are aware of our listening in. Clark and Carlson

(1982), Clark and Schaefer (1992), and Clark (1996) have described the processes that occur

in the overhearer’s mind in everyday situations; in doing this, they build on Goffman’s

participation framework.

Goffman (1976, 1979) distinguishes three main listener roles: firstly, ‘overhearers’, whose

unratified participation can be intentional or not and can be encouraged or not. Secondly, ‘ratified

participants’, who – in the case of there being more than two persons talking – are not specifically

addressed by the speaker. Thirdly, addressees, who are ‘oriented to’ by the speaker in a way that

suggests that his or her words are directed specifically at them. This participation framework, as

expanded by Clark and his co-workers, can be schematically illustrated as in Fig. 2.

The relevant roles for film discourse are overhearers, divided (as seen in Fig. 2) into

bystanders and eavesdroppers. The bystanders are openly present, but not part of the

conversation; for example, a couple is having a conversation on the bus, and people are sitting

opposite them within hearing distance of what is said. Eavesdroppers, by contrast, listen in

without the speakers’ being aware of it, for example, when someone is listening behind a door to

a conversation going on inside a room. According to Clark (1996:14), the roles of bystanders and

eavesdroppers, however, form a continuum, as there are in-between varieties of overhearers; for

example, in the bus scenario, the overhearers who are located behind the conversationalists are


5 In spite of the general claim that the visual takes priority in film, the central role of ‘overhearing’ has been recognized

by film scholars. Weis (1999), for example, draws on psychoanalysis to argue that the adult eavesdropper recapitulates the

primal scene of overhearing one’s parents making love and identifies with either of the parents or the overhearing child.

She thus suggests that ‘‘overhearing is a fundamental experience with profound implications for films’’ (p. 84).

not as openly present as the ones sitting opposite them. As for eavesdroppers, there are cases

when they believe that the conversationalists do not know they are being overheard, while in fact

they are aware of it. I will therefore use the term ‘overhearer’ to denote a listener situated

somewhere on the continuum between bystander and eavesdropper.

Short (1989) has pointed out (as have Holly and Baldauf, 2002) that the spectators in the

theatre or in front of the screen cannot be considered bystanders or eavesdroppers in the sense of

Goffman, because ‘‘the situation is arranged to be overheard on purpose’’ (Short, 1989:149);

consequently, the spectators are ratified participants (Holly and Baldauf, 2002:55). Even so, I

argue, they can still be considered overhearers. Firstly, there are overhearer situations in which

the conversationalists want the bystander or eavesdropper to glean certain information from what

they are saying, without actively interacting with the person in the conversation (see below,

section 3.1.2). Secondly, the characters in the top layer that the spectators are listening to are

not the ones who design the talk destined for the overhearers in front of the screen; this is what

the screenwriter, the director, and the actors in the overlaying layers do. Or, as Goffman

argues:

... the words addressed by one character in a play to another (at least in modern Western

dramaturgy) are eternally sealed off from the audience, belonging entirely to a self-

enclosed, make-believe realm (Goffman, 1979:13).

Consequently, listening to dialogue on the screen is not different from listening to dialogue in

the role of an overhearer; here, too, the talk is ‘sealed off’, as it concerns the ratified participants’

world that as a rule only incidentally overlaps with that of the overhearer.

Scollon, when describing his concept of ‘a watch’ as ‘‘any person or group of people who are

perceived to have attention to some spectacle as the central focus of their (social) activity’’

(Scollon, 1998:92) follows the same line of thinking. No matter whether there is mediation (as

when we see a couple arguing in a movie), or not – (as when we see a couple arguing in the street),

there is a ‘‘barrier’’ between the spectacle and the watchers (Scollon, 1998:93).

Goffman uses the term ‘audience’ to refer to the spectators in the theater, distinguishing this

kind of participation framework from that of a conversation. Audiences are not a feature of

speech events, but of ‘‘stage events’’ (Goffman, 1979:13). He extends the term ‘audience’ to

spectators and listeners in front of a TV screen or the radio, when talk is addressed to ‘‘imagined

recipients’’ (Goffman, 1979:12). Although this favors applying a different participation

framework for film discourse, the points argued above still justify comparing the situation of

watching people and listening to them talking on the screen to the everyday situation of

overhearing conversations.


Fig. 2. The participation framework (Clark, 1996:14).

In order to explain the process of interpreting utterances by unratified participants, Clark and

Schaefer (1992:259ff.) distinguish ‘recognizing’ from ‘conjecturing’. Ratified participants, i.e.

addressees and side-participants (see Fig. 2), recognize what speakers say by inferring from

conclusive evidence, based on the ‘common ground’ between the conversationalists.

Understanding is possible on the basis of knowledge, shared between speaker and listener.

This shared knowledge, or common ground, is distinguished into communal common ground and

personal common ground. Communal common ground is all the knowledge and all the beliefs

held in the communities that the participants share membership of. Personal common ground is

all the mutual knowledge and all the mutual beliefs that the participants share from personal

experience with each other, including the present interaction. Overhearers, on the other hand, can

only conjecture about what speakers mean; that is they draw inferences from inconclusive

evidence, since they generally do not fully share knowledge with the ratified participants. Even if

overhearers share the communal common ground and some personal common ground, they are

unlikely to have taken part in all of the participants’ shared experiences, and thus there is always

some part of the common ground that is sealed off from them. Conjecturing then means

reconstructing the common ground that the speaker presupposes such that the hearer can use the

utterance to expand the common ground, thereby giving the utterance a reasonable interpretation.

For example, if we are overhearing two young people on the bus and one says:

‘‘Psycholinguistics was really interesting yesterday, wasn’t it?’’, we reconstruct the following

common ground: the two young people are students in the same course, namely linguistics, and

they attend a seminar or lecture called ‘Psycholinguistics’, which took place the day before. If

that is part of the common ground between the ratified participants, our reconstruction is

consistent with theirs, and the utterance makes sense to both ratified and unratified participants.

How exactly does the process of making conjectures work? Common ground, as reconstructed

and reproduced in conjectures, is communal or personal knowledge shared by the participants in

a conversation. Knowledge is generally organized into so called ‘idealized cognitive models’

(Lakoff, 1987) or ‘cognitive frames’ (Tannen, 1979; Hedges, 1991). When overhearers make

conjectures, they retrieve stored cognitive models or frames that are prompted by the utterance.

These models are then compared to what is said, and if they do not fit, other models with a better

fit are retrieved; in case there is no matching model, the new information is fitted into the existing

knowledge structures, expanding and combining these or establishing new cognitive models. The

cognitive models that stand up to the comparison constitute the part of the common ground

referred to in the utterance.

Consider the following exchange between Stuart and Maurice from the film Secrets and lies

(example 3; Leigh, 1996). The scene is a meeting between two men who have not seen each other

for a long time. Years ago, Maurice bought a photography business from Stuart, who then

emigrated to Australia. Now Stuart has returned and starts quarreling with Maurice about how to

run the business, especially about how to organize wedding photography.

(example 3) Secrets and lies1 S how many weddings do you do?

2 M oh .. enough.

3 S how many?

4 M oh .. about forty a year.

5 S I used to do a hundred and forty.

6 M what?

personally?


7 S no.

not personally.

no.

nobody does them personally.

8 M I do.

9 S well, then you’re a bloody fool.

you get people in, get them out there.

if the work’s there, take it.

you’ve got to grab it while you can.

10 M it’s not in my interest to get some tosser in.

I mean I’ll have no control,

it could fuck up my reputation.

In order to make sense of this exchange, the overhearer in the cinema audience has to retrieve

cognitive models of weddings and of a photography business. With the help of these models, the

overhearer interprets Maurice’s and Stuart’s dialogue.

The models or frames that spectators retrieve while overhearing film discourse, however, also

encompass knowledge patterns acquired through the experience of watching other movies

(Hedges, 1991:xiv) or by watching TV (Holly and Baldauf, 2002:52). These are part of our

background knowledge just as real world experiences are. Berliner states: ‘‘Movie dialogue

obeys its own customs. We accept it according to the terms of the cinema, not of reality’’

(Berliner, 1999:3). While much of movie dialogue is close to reality, there are certain stock lines

that the audience has heard in countless movies and that have a well-defined additional meaning.

A change about to occur in a character, for example, can be indicated by ‘‘I can’t take it

anymore’’, and the line ‘‘It’s quiet. Too quiet’’ means that something is going to happen soon

(Berliner, 1999:3). With respect to TV discourse in particular, Holly and Baldauf (2002:52) argue

that the TV viewer has expectations of the characters that specific actors impersonate or of

storylines typical of certain broadcasting stations. In their corpus of recorded talk by people

watching TV, they find evidence for these processes. One person, for example, says: ‘‘I like her;

she always plays the part of a real bitch’’ (Holly and Baldauf, 2002:52; my translation). This

finding can be extended to comprise expectations of what specific directors, screenwriters, and

cinematographers are doing, or are going to do. This kind of background knowledge is activated

in the process of appreciation.

To recapitulate, overhearers – whether in the cinema or in everyday situations – are unratified

participants in a conversation. That means they have no conversational rights or responsibilities.

This has consequences for the process of understanding insofar as overhearers draw inferences,

based on the limited amount of common ground they share with the ratified participants; because

of this, overhearers are at a disadvantage, compared to addressees in a conversation. In addition,

they suffer from yet another disadvantage. While addressees are ratified participants in a

conversation and can negotiate meaning with the speaker in a ‘grounding process’, in Clark and

Schober’s terms (1992), the overhearer has no such rights. Troubles in understanding may thus

accumulate. These disadvantages of the spectator in screen-to-face discourse render the design of

successful cinematic dialogue a complex challenge for the film production team.

3.1.2. Overhearer design

In order to explain how film dialogue is designed for the implied spectators located in front of

the screen, it is again helpful to compare screen-to-face conversation to the situation in which


someone overhears face-to-face conversation. Clark and Schaefer (1992:262ff.) distinguish four

attitudes towards overhearers.

The simplest attitude towards overhearers is ‘indifference’ (Clark and Schaefer, 1992:263).

This means that the speaker can design his or her utterance as if the overhearer were not present.

However, the speaker is still bound by politeness obligations; for example, when you are having a

conversation on a bus, while sitting across from an elderly lady, you might refrain from using

strong expletives and shouting.

More difficult is the attitude of ‘disclosure’ (Clark and Schaefer, 1992:264). In this case, the

speaker wants the overhearer to gather certain information from the conversation. The

conversationalists on the bus vaguely know the person sitting opposite and, for example, want

for that person to gain a positive opinion of them. That means that the speaker has to provide the

overhearer with enough evidence so that she or he can make the appropriate conjectures, based on

the conversationalists’ and the overhearer’s common ground. The less common ground there is

between overhearer and conversationalists, the more difficult it is to disclose information to an

overhearer. It could be argued that this attitude blurs the boundaries between overhearer and

addressee, as the overhearers to a certain extent can be considered as being addressed by the

conversationalists. However, the conversationalists do not want the overhearers to actively

participate in the conversation.

In a third attitude, ‘concealment’ (Clark and Schaefer, 1992:265), the conversationalists make

use of the lack of common ground to keep the overhearer away from the evidence needed to infer

correctly what they mean. For each conversational utterance, the addressee is to make sense of

what is said on the basis of the parts of common ground that are open to him or her but closed to

the overhearer. The easiest way to accomplish this is to switch to a language that is not known to

the overhearer, or to some kind of spy code. The less common ground there is between the

conversationalists and the overhearer, the easier it is to practice concealment.

The most complicated attitude is ‘disguisement’ (Clark and Schaefer, 1992:267). In this case,

the conversationalists want the overhearer to reach the wrong conclusions without noticing it.

Clark and Schaefer (1992:273) cite the following example:

Just before Pearl Harbor in World War II, the chief of the American bureau of the Japanese

Foreign Office talked with an associate of the Japanese Ambassador to the United States on

the telephone. They suspected a wiretap and tried to disguise what they were saying with a

pre-arranged code that made their talk sound personal and mundane. In referring to

Japanese-American negotiations, for example, they spoke of a matrimonial question, so

when the associate said The matrimonial question seemed as if it would be settled, he meant

It looked as though we could reach an agreement.

It is important to see that the speaker has to assume some kind of attitude towards potential

overhearers (Clark and Schaefer, 1992:269). Indifference is the default, but that still means

choosing not to disclose, not to conceal, and not to disguise. This in turn means that speakers

always design utterances with overhearers, if present, in mind.

Depending on the speakers’ attitude towards overhearers, more or less of the common ground

is accessible to them. If the speaker, for example, wants to conceal what is said from the

overhearer, she or he will try to base what is said on information that is not accessible to the

overhearer; for example, the overhearer cannot access the personal common ground shared by the

ratified participants. Utterances like ‘You-know-who told me’, for example, allow no correct

conjectures, no reproduction of common ground can be imagined to which this utterance would

be a meaningful addition. If, on the other hand, the attitude is disclosure, where the ratified


participants want the overhearer to glean information from what they are saying, they must

design their utterances so as to allow an interpretation that contradicts those parts of their

common ground that are open to the overhearer (Clark and Schaefer, 1992:262ff.).

The latter is essentially what happens in film discourse: the default attitude here is that of

disclosure.6 Utterances are designed with overhearers in mind, on the basis of an estimate of the

spectators’ world knowledge and on the knowledge the participants have gleaned from interactions

that the spectators have observed. This process is the screen-to-face equivalent of Sacks et al. (1974)

‘recipient design’ in face-to-face conversation. By recipient design, they refer to ‘‘a multitude of

respects in which the talk by a party in a conversation is constructed or designed in ways which

display an orientation and sensitivity to the particular other(s) who are co-participants’’ (Sacks

et al., 1974:727). Correspondingly, the audience in front of the screen is ‘oriented to’, by which I

mean that the spectators’ knowledge structures are anticipated. Every utterance is tailored to them.

The difference between recipient design in face-to-face communication and overhearer design in

screen-to-face conversation is that the anticipation of the recipient’s knowledge is indirect,

established in ‘‘hypothetical intersubjectivity’’ (Ayaß, 2002:149) and consequently more

tentative.7

If relevant information is considered closed to overhearers, it must be included in the

utterance, so that correct conjectures can be made, even though this information might be

redundant for the ratified participants. Imagine, for example, two characters in a science fiction

movie talking about a technological gadget that is standard equipment in their world but does not

exist in the world of the audience; they will have to include information on how it works and what

its purpose is, even though both of them know. They make this information available to the

overhearer who does not share their common ground, but who, with the help of the additional

information, can make sense of what they are talking and make the correct conjectures.

In sum, cinematic dialogue has to take the disadvantages of screen-to-face discourse into

account; generally speaking, its design is based on the condition that the common ground on

which utterances are based is open to the audience. However, this is difficult to realize,

considering the vastly different backgrounds and experiences of the many viewers of a film or TV

drama: different age groups, genders, occupations, to name but a few distinguishing factors. Each

viewer resorts to individual experiences and knowledge patterns, and may consequently come up

with a different interpretation of a line in a movie. Like all mass media texts, cinematic texts are

open to interpretation by an active audience; their fundamental characteristic is their variability

in interpretation (cf. Eco, 1967/1985:152).8

3.2. The workings of film discourse

In the following, I will describe how film discourse works, which processes take place in the

recipients’ minds, and how film producers design the dialogue for an audience. In this, the


6 Sometimes, it is indifference, especially at the beginning of movies when the audience stumbles into the lives of the

characters and at first has to make the transition into the realm of the story. On some occasions, it is concealment;

especially in the creation of suspense it is helpful, if the audience is not able to piece together a common ground, and

fragments are disclosed in a way that delays this process.7 In his analysis of the language of news media, Bell (1991) uses the term ‘audience design’ for this phenomenon.8 Complicating matters even more is the fact that film dialogue is used to provide a source for predictive inferences. For

the audience to generate the predictions that movie makers intend them to, it is first of all necessary for the audience to

make the correct conjectures, so as to arrive at the meaning the movie makers had in mind when assigning an utterance to

a character (cf. Magliano et al., 1996).

spectators are accorded a central position. Consider again the exchange between Maurice and

Stuart cited above (example 3). On the underlying discourse levels, the spectators overhear the

argument about how to organize wedding photography that is taking place between Maurice and

Stuart on the top level. The utterances they overhear are compared to their existing knowledge

structures, and the common ground between the participants is reproduced, thus allowing them to

make sense of what Maurice and Stuart say. For example, when Stuart says ‘I used to do a

hundred and forty’ (line 5), the audience retrieves knowledge structures that are composed of

general world knowledge and of knowledge of the former utterances in the exchange between

Maurice and Stuart; these structures belong to those parts of the participants’ personal common

ground that are open to an audience which has been doing overhearing since the beginning of the

exchange. On the basis of this open personal common ground, the audience knows that the

number 140 refers to weddings, and because of their world knowledge, they know that this is a

high number—photographers being usually present at church weddings, church weddings taking

place on weekends, and a year having 52 weekends. This interpretation then helps them make

sense of Maurice’s reaction in line 6: ‘what? personally?’

The conjecturing process is facilitated because of the utterances being ‘designed’ for the

overhearer (see above, section 3.1). By this, I mean that the characters, for example, generally do

not mumble, as this would close off personal common ground to the audience, who cannot add

mumbled utterances to their knowledge. They also generally do not speak in a variety of language

that is unintelligible to the overhearer; the code they share with the audience is part of the

participants’ common ground that is open to the overhearer.9 Note that in example 3, the two

characters, Maurice and Stuart, are arguing about a part of their business that is open to the audience.

If their attitude towards the audience were to be one of indifference or concealment, they could, for

example, have argued about the advantages of certain techniques of photography, like lenses and

developing solutions. The film production crew has obviously constructed the dialogue in such a

fashion that the audience can comprehend the exchanges with the help of the part of their world

knowledge that overlaps with the world knowledge the production team projects onto the

characters.

Who designs the utterances for the overhearer? In the case at hand, it is not the characters on the

topmost level, Maurice and Stuart, who do this, but the actors, Timothy Spall and Ron Cook, who

enunciate the utterances clearly so that the audience can follow. Neither is it the characters, who

choose the wording but the actors, together with the screenwriter and director, Mike Leigh. Another

important part is played by the camera as it, for example, zooms in on the faces of Maurice and

Stuart uttering the words so that the overhearer also receives non-verbal cues; or, contrariwise, the

camera may show the person speaking from the back, away from the face of the addressee, so that

we only receive verbal cues from the speaker but non-verbal cues from the listener. This again is

determined by the director and expressed in the screenplay, to be finally decided upon in the process

of editing the material filmed on the set. Through the choice of camera angles, too, dialogue can

be structured and emphasized; for example, close-ups are chosen, if a conversation becomes more

emotional or an argument reaches its climax. ‘Knee shots’, on the other hand, in which one sees the

conversing characters from the knee up, are preferred when the dialogue is of little importance to

the film’s plot (Straßner, 2001:1095). Another issue relevant to film dialogue that is decided upon in

the editing process is the length of shots: generally, short shots are preferred for emotional

dialogues, clearly structuring the conversation through cutting from speaker to speaker, whereas


9 This has the effect that, for example, in Hollywood movies, even if foreigners appear, they will speak English. Only

fairly recently, there has been a trend towards linguistic realism and one hears languages other then English.

long shots are favored for less emotional dialogue (Straßner, 2001:1096). Along with the

screenwriter, the director, the camera team, and the cutters, one more person behind the scenes has

an impact on the film dialogue, namely the producer, who usually reads the screenplay, chooses a

director, and may be involved in the editing.

The whole of this process is relevant in the design of utterances for the overhearers in

the cinema, so that the co-construction of meaning in movie discourse becomes a joint effort

of the film recipients, the actors, the director, the screenwriter, the producer, the camera staff, and

the cutters involved in the editing process (see Fig. 3).

Fig. 3 shows that both the members of the production crew and the members of the audience

make use of their world knowledge to design and interpret the film dialogue. Consequently, the

study of film dialogue provides insights into these underlying knowledge patterns, amongst

others insights into the underlying expectations of real conversation. As Tannen and Lakoff argue

(in their analysis of the dialogue in Bergman’s film Scenes from a marriage): ‘‘. . . artificial dialog

may represent an internalized model or schema for the production of conversation – a

competence model that speakers have access to’’ (Tannen and Lakoff, 1994:139).10


Fig. 3. A model of film discourse (cf. Bubel, 2006).

10 This procedure is all the more justified as valid research has also been done on the basis of overhearers’ intuitions

about recorded conversations played to them. McGregor (1990) showed how overhearers’ interpretive, oral responses to

particular discourse fragments may be distinguished into observations and inferences. Fox Tree (2002) uses this research

methodology to show how pauses and filled pauses are interpreted.

4. Conclusion

In this paper, we have seen that theories developed to explain the workings of various forms of

mediated discourse are not sufficient to account for the processes involved in the creation and

comprehension of film discourse. The approaches by Burger and Short neither consider the role of

the film production team nor the role of the audience in the co-construction of meaning. Clark’s

model of layered discourse, on the other hand, while it deals with both these aspects, does not make

explicit the cognitive processes involved in the design and interpretation of film discourse.

The complexity of mediated discourse is a notion that all three models share. This idea was

integrated into a description of film discourse which focuses on the underlying levels of discourse,

with the audience as overhearers of the conversation on the screen. The processes in the spectator’s

mind are considered to correspond with those in everyday overhearer situations, as when we listen

in on a conversation between people sitting in front of us on the bus. Overhearers can only make

conjectures about what they are able to listen in on, as they do not fully share the participants’

common ground. Consequently, in order to be intelligible, film dialogue has to be carefully

designed for overhearers so that they can reconstruct the participants’ common ground, and the film

production crew involved in this design has to construct the dialogue on the basis of the knowledge

patterns they expect the future audience to share with them. This implies that meaning in film

discourse is co-constructed in a joint effort of recipients and the production crew—all drawing on

their respective world knowledges and especially on their knowledge of communicative processes

in everyday situations. This renders film discourse a relevant field of study for linguists who are

seeking to gain insights into language users’ underlying competence. Further research is needed to

strengthen the theoretical assumptions made in this paper with empirical findings, in particular by

interviewing film audiences and determining how they arrived at their understandings.

Transcription conventions

All examples cited in this contribution are transcribed from the films according to the

conventions established by Dressler and Kreuz (2000):

she’s out. A period shows falling tone in the preceding element.

oh yeah? A question mark shows rising tone in the preceding element.

nine, ten. A comma indicates a level, continuing intonation.

DAMN Capitals show heavy stress or indicate that speech is louder than

surrounding discourse.

(2.0) Numbers in parentheses indicate timed pauses.

If the exact duration of the pauses is not crucial and/or not timed:

.. a truncated ellipsis is used to indicate pauses of one-half

second or less.

. . . an ellipsis is used to indicate a pause of more than a half-second.

ha:rd The colon indicates a prolonged the prior sound or syllable.

((desperate whisper)) Special features of the utterance, such as whispers, coughing,

and laughter, are indicated with double parentheses.

References

Akmajian, Adrian, Demers, Richard A., Harnish, Robert M., 1980. Overcoming inadequacies in the ‘message-model’ of

linguistic communication. Communication and Cognition 13, 317–336.

Arundale, Robert B., 1999. An alternative model and ideology of communication for an alternative politeness theory.

Pragmatics 9, 119–153.


Ayaß, Ruth, 2002. Fernsehgattungen in der Aneignung. In: Holly, Werner, Puschel, Ulrich, Bergmann, Jorg (Eds.), Der

sprechende Zuschauer. Westdeutscher Verlag, Wiesbaden, pp. 143–152.

Bell, Allan, 1991. The Language of News Media. Blackwell, Oxford.

Berliner, Todd, 1999. Hollywood movie dialogue and the ‘real realism’ of John Cassavetes. Film Quarterly 52, 2–16.

Betten, Anne, 1977. Moderne deutsche Dramen- und Filmdialoge und ihr Verhaltnis zum spontanen Gesprach.

In: Drachmann, Gaberell (Ed.), Akten der 2. Salzburger Fruhlingstagung fur Linguistik 1975, Narr, Tubingen,

pp. 357–371.

Bordwell, David, 1989. A case for cognitivism. Iris 9, 11–40.

Bubel, Claudia, 2006. The linguistic construction of character relations in TV drama: doing friendship in Sex and the City.

Universitat des Saarlandes Saarbrucken URN: urn:nbn:de:bsz:291-scidok-5988. URL: http://scidok.sulb.uni-saarland.

de/volltexte/2006/598/.

Burger, Harald, 1984. Sprache der Massenmedien. de Gruyter, Berlin.

Burger, Harald, 1991. Das Gesprach in den Massenmedien. de Gruyter, Berlin.

Clark, Herbert H., 1996. Using Language. Cambridge University Press, Cambridge.

Clark, Herbert H., Carlson, Thomas B., 1982. Hearers and speech acts. Language 58, 332–372.

Clark, Herbert H., Schaefer, Edward F., 1992. Dealing with overhearers. In: Clark, Herbert H. (Ed.), Arenas of Language

Use. University of Chicago Press, Chicago, pp. 248–273.

Clark, Herbert H., Schober, Michael F., 1992. Understanding by addressees and overhearers. In: Clark, Herbert H. (Ed.),

Arenas of Language Use. University of Chicago Press, Chicago, pp. 176–197.

Clark, Herbert H., Van Der Wege, Mija, 2001. Imagination in discourse. In: Schiffrin, Deborah (Ed.), The Handbook of

Discourse Analysis. Blackwell, Oxford, pp. 772–786.

Culpeper, Jonathan, 2001. Language and Characterisation: People in Plays and Other Texts. Longman, London.

Dressler, Richard A., Kreuz, Roger J., 2000. Transcribing oral discourse: a survey and model system. Discourse Processes

29, 25–36.

Duranti, Alessandro, 1986. The audience as co-author: an introduction. Text 6, 239–247.

Eco, Umberto, 1985 (1967). Fur eine semiologische Guerilla. In: Eco, Umberto (Ed.), Uber Gott und die Welt, translated

by Burkhart Kroeber, Carl Hanser, Munchen, pp. 146–156.

Fox Tree, Jean E., 2002. Interpreting pauses and ums at turn exchanges. Discourse Processes 34, 37–55.

Giora, Rachel, Ne’eman, Judd, 1996. Categorical organization in the narrative discourse: a semantic analysis of Il

conformista. Journal of Pragmatics 26, 715–735.

Goffman, Erving, 1976. Replies and responses. Language in Society 5, 257–313.

Goffman, Erving, 1979. Footing. Semiotica 25, 1–29.

Gumperz, John J., 1982. Discourse Strategies. Cambridge University Press, Cambridge.

Hedges, Inez, 1991. Breaking the Frame: Film Language and the Experience of Limits. Indiana University Press,

Bloomington.

Holly, Werner, Baldauf, Heike, 2002. Grundlagen des fernsehbegleitenden Sprechens. In: Holly, Werner, Puschel, Ulrich,

Bergmann, Jorg (Eds.), Der sprechende Zuschauer, Westdeutscher Verlag, Wiesbaden, pp. 41–60.

Hopper, Robert, LeBaron, Curtis, 1999. How gender creeps into talk. Research on Language and Social Interaction 31,

59–74.

Klemm, Michael, 2000. Zuschauerkommunikation. Peter Lang, Frankfurt.

Kobus, Isabel, 1998. Dialog in Roman und Film: Untersuchungen zu Joseph Loseys Literaturverfilmungen ‘The Go-

Between’ und ‘Accident’. Peter Lang, Frankfurt.

Kozloff, Sarah, 2000. Overhearing Film Dialogue. University of California Press, Berkeley.

Krauss, Robert M., 1987. The role of the listener: addressee influences on message formulation. Journal of Language and

Social Psychology 6, 81–98.

Lakoff, George, 1987. Women, Fire and Dangerous Things. University of Chicago Press, Chicago.

Leigh, Mike, 1996. Secrets and Lies. Film Four Distributors, London.

Leigh, Mike, 1997. Career girls. Film Four Distributors, London.

Magliano, Joseph P., Dijkstra, Katinka, Zwaan, Rolf A., 1996. Generating predictive inferences while viewing a movie.

Discourse Processes 22, 199–224.

McGregor, Graham, 1990. Eavesdropper response and the analysis of everyday communicative events. In: McGregor,

Graham (Ed.), Reception and Response. Routledge, London, pp. 99–111.

Morley, David, 1994. Active audience theory: pendulums and pitfalls. In: Levy, Mark R., Gurevitch, Michael (Eds.),

Defining Media Studies: Reflections on the Future of the Field. Oxford University Press, Oxford, pp. 255–261.

Morrow, Daniel G., Clark, Herbert H., 1988. Interpreting words in spatial descriptions. Language and Cognitive Processes

3, 275–291.


http://scidok.sulb.uni-saarland.de/volltexte/2006/598/

http://scidok.sulb.uni-saarland.de/volltexte/2006/598/

Reddy, Michael J., 1979. The conduit metaphor: a case of frame conflict in our language about language. In: Ortony,

Anthony (Ed.), Metaphor and Thought. Cambridge University Press, Cambridge, pp. 285–324.

Sacks, Harvey, Schegloff, Emanuel A., Jefferson, Gail, 1974. A simplest systematics for the organization of turn-taking

for conversation. Language 50, 696–735.

Scollon, Ron, 1998. Mediated Discourse as Social Interaction: A Study of News Discourse. Longman, London.

Short, Mick H., 1981. Discourse analysis and the analysis of drama. Applied Linguistics 2, 180–201.

Short, Mick H., 1989. Discourse analysis and the analysis of drama. In: Carter, Ronald, Simpson, Paul (Eds.), Language,

Discourse, and Literature. Unwin Hyman, London, pp. 139–168.

Short, Mick H., 1994. Discourse analysis and drama. In: Asher, R.E. (Ed.), The Encyclopedia of Language and

Linguistics. Pergamon, Oxford, pp. 949–952.

Simpson, Paul, Montgomery, Martin, 1995. Language, literature and film: the stylistics of Bernard MacLaverty’s Cal.

In: Verdoenk, Peter, Weber, Jean Jacques (Eds.), Twentieth Century Fiction: From Text to Context. Routledge,

London, pp. 138–165.

Straßner, Erich, 2001. Kommunikative und asthetische Leistungen der Sprache in der Geschichte des Films. In: Leonhard,

Joachim-Felix, Ludwig, Hans-Werner, Schwarze, Dietrich, Straßner, Erich (Eds.), Handbucher zur Sprach- und

Kommunikationswissenschaft, vol. 15. Medienwissenschaft, de Gruyter, Berlin, pp. 1093–1106.

Tannen, Deborah, 1979. What’s in a frame? Surface evidence for underlying expectations. In: Freedle, Roy O. (Ed.),New

Directions in Discourse Processing. Ablex, Norwood, NJ, pp. 137–181.

Tannen, Deborah, Lakoff, Robin T., 1994. Conversational strategy and metastrategy in a pragmatic theory: the example of

scenes from a marriage. In: Tannen, Deborah (Ed.), Gender and Discourse. Oxford University Press, Oxford, pp. 137–

173.

Weis, Elisabeth, 1999. The narrative functions of the ecouteur. In: Brophy, Philip (Ed.), Cinesonic: The World of Sound in

Film. AFTRS, Sydney, pp. 79–107.


Documents

Film audiences as overhearers, Claudia M. Bubel