CONTENTS CROSS‐REFERENCED

Topic A INTRODUCTION B DEVELOPMENT C EXPLORATION E EXTENSION Topic

1 What is discourse analysis? Three ways of looking at discourse Doing discourse analysis: first The three perspectives revisited 1
steps (Zellig Harris; Henry G.
Widdowson; James Paul Gee)
2 Texts and texture Cohesion and coherence Analyzing texture Two perspectives on texture 2
(Michael A. K. Halliday and
Ruqaiya Hasan; David
Rumelhart)
3 Texts and their social functions All the right moves Analyzing genres Genres, discourse communities 3
and power (John Swales, Vijay
K. Bhatia)
4 Discourse and ideology Constructing reality Other people’s voices Ideologies in discourse (Norman 4
Fairclough; James Paul Gee)
5 Spoken discourse The texture of talk Analyzing speech acts Two perspectives on 5
conversation (John L. Austin;
Emanuel A. Schegloff and
Harvey Sacks)
6 Strategic interaction Negotiating relationships and Analyzing conversational Frames in interaction (Deborah 6
activities strategies Tannen and Cynthia Wallat)
7 Context, culture and The SPEAKING model Analyzing contexts The ethnography of 7
communication communication (Dell Hymes;
TM
Muriel Saville-Troike)
8 Mediated discourse analysis Mediation Doing mediated discourse Discourse and action (Ron 8
analysis Scollon)
9 Multimodal discourse analysis Modes, meaning and action Doing multimodal discourse Two perspectives on 9
analysis multimodality (Gunther Kress,
and Theo van Leeuwen; Sigrid
Norris)
10 Corpus-assisted discourse Procedures for corpus-assisted Analyzing corpora Finding ‘Discourses’ with 10
analysis discourses analysis corpus-assisted analysis (Paul
Baker, and Tony McEnery)

6
A6 STRATEGIC INTERACTION
In the last section we talked about how utterances in conversations are used to
perform certain kinds of actions like greeting, requesting, inviting, warning and
apologizing. But usually these actions taken alone do not constitute
conversations. Conversations happen when multiple actions are put together to
form activities: we chat, we debate, we flirt, we counsel, we gossip, we
commiserate, and we do many other things in our conversations. At the same
time, we use conversations to show that we are certain kinds of people and to
establish and maintain certain kinds of relationships with the people with whom
we are talking. We do not, however, engage in these activities and construct
these identities and relationships all by ourselves. We must always negotiate
‘what we are doing’ and ‘who we are being’ with the people with whom we are
interacting. We call the methods we use to engage in these negotiations
conversational strategies.

In this strand we will focus on two basic kinds of conversational strategies: face
strategies and framing strategies. Face strategies have to do primarily with
showing who we are and what kind of relationship we have with the people with
whom we are talking. Framing strategies have more to do with showing what we
are doing in the conversation, whether we are, for example, arguing, teasing,
flirting or gossiping. As we shall see, however, face strategies also contribute to
the management of conversational activities (especially those involving face
threatening acts), and framing strategies are often central to the discursive
construction of identity.

These two concepts for analyzing how we manage conversations come from an
approach to discourse known as interactional sociolinguistics, which is concerned
with the sometimes very subtle ways people signal and interpret what they think
they are doing and who they think they are being in social interaction. It is
grounded primarily in the work of the anthropologist John Gumperz (1982a,
1982b) who drew on insights from anthropology and linguistics as well as the
fields of pragmatics and conversation analysis that we discussed in the previous
section. One of the most important insights Gumperz had was that people
belonging to different groups have different ways of signaling and interpreting
cues about conversational identity and conversational activities, and this can
sometimes result in misunderstandings and even conflict. Not surprisingly,
interactional sociolinguistics has been used widely in studies of intercultural
communication, including some of the early studies by Gumperz himself of
communication between Anglo‐British and people of South Asian origin in the
U.K.

Another important influence on interactional sociolinguistics comes from the

American sociologist Erving Goffman, who, in his classic book The Presentation of
Self in Everyday Life (1959), compared social interaction to a dramatic
performance. Social actors in everyday life, he argued, like stage actors, use
certain ‘expressive equipment’ like costumes, props, and settings to perform

23
certain ‘roles’ and ‘routines’. Our goal in these performances is to promote our
particular ‘line’ or version of who we are and what is going on. Most of the time,
other people help us to maintain our line, especially if we are willing to help
them to maintain theirs. Sometimes, however, people’s ‘lines’ are not entirely
compatible, which means they need to negotiate an acceptable common ‘line’ or
else risk spoiling the performance for one or more of the participants.
It was Goffman who contributed to discourse analysis the concepts of face and
frames. By ‘face’ he meant ‘the positive social value a person effectively claims
for himself by the line others assume he has taken’ (1967: 41). In other words,
for Goffman a person’s face is tied up with how successful he or she is at ‘pulling
off’ his or her performance and getting others to accept his or her ‘line’. What he
meant by ‘frames’ was ‘definitions of a situation (that) are built up in accordance
with principals of organization which govern events.’ The concept of ‘framing’
relates to how we negotiate these ‘definitions of situations’ with other people
and use them as a basis for communicating and interpreting meaning.

Showing who we are: Face strategies

Social identity is a complex topic and one that we will return to later in this book.
For now we will focus on one fundamental aspect of identity: the fact that our
identities are always constructed in relation to the people with whom we are
interacting. Some people are our friends, and others are complete strangers.
Some people are our superiors and others are our subordinates. When we talk,
along with conveying information about the topic about which we are talking, we
always convey information about how close to or distant from the people with
whom we are talking we think we are, along with information about whether we
are social equals or whether one has more power than the other. The strategies
we use to do this are called face strategies.

The concept of ‘face’ in its more everyday sense will be familiar to many readers.
The term is often used to denote a person’s honor or reputation. Many cultures
have the notions of ‘giving’ people face (helping them to maintain a sense of
dignity or honor) and of ‘losing face’ (when people, for some reason of another
suffer a loss of dignity or honor). The definition of face we will be using here,
while not far from the more everyday meaning, is based primarily on the work of
Goffman and other social scientists. We define face as ‘the negotiated public
image mutually granted to each other by participants in a communicative event’
(Scollon, Scollon and Jones 2011).

There are three important aspects to this definition: The first is that one’s face is
one’s public image rather than one’s ‘true self’. This means that the social image
that constitutes face is not the same in every interaction in which we engage. We
‘wear’ different faces for different people. The second important aspect of this
definition is that this image is ‘negotiated’. That is to say, it is always the result of
a kind of ‘give and take’ with the person or people with whom we are interacting,
and throughout a given interaction the image that we present and the images
others project to us may undergo multiple adjustments. Finally, this image is
‘mutually granted’. In other words, successfully presenting a certain face in

24
interaction depends on the people with whom we are interacting cooperating
with us. This is because face is the aspect of our identity which defines us in
relation to others. If one person’s idea of the relationship is different from the
other person’s idea, chances are one or the other will end up ‘losing face’. And so,
in this regard, the everyday ideas of ‘giving face’ and ‘losing face’ are also quite
important in our definition of face.

There are basically two broad kinds of strategies we use to negotiate our
identities and relationships in interaction. The first we will call involvement
strategies. They are strategies we use to establish or maintain ‘closeness’ with
the people with whom we are interacting – to show them that we consider them
our friends. These include things like calling people by their first names or using
nicknames, using informal language, showing interest in someone by, for
example, asking personal questions, and emphasizing our common experiences
or points of view. While such strategies can be used to show friendliness — as
we will see in the next section — they can also be used to assert power over
people. Teachers, for example, often use such strategies when interacting with
young students, and bosses sometimes use them when interacting with their
employees.

The second class of face strategies is known as independence strategies. These

are strategies we use to establish or maintain distance from the people with
whom we are interacting either because we are not their friends, or, more
commonly, because we wish to show them respect by not imposing on them.
They include using more formal language and terms of address, trying to
minimize the imposition, being indirect, apologizing and trying to depersonalize
the conversation (see Table A6.1).

Table 6.1 Face strategies

Involvement Strategies Independence Strategies

Using first names or nicknames Using tiles (Good afternoon,

(Hey, Rodders!) Professor Jones.)

Expressing interest (What have Apologizing (I’m terribly sorry to

you been up to lately?) bother you.)

Claiming a common point of Admitting differences (Of course,

view (I know exactly what you you know much more about it than
mean.) I do.)

Making assumptions (I know you Not making assumptions (How

love lots of sugar in your coffee.) would you like your coffee today?)

Using informal language (Gotta Using formal language (Pardon me,

minute?) can you spare a few moments?)

Being direct (Will you come?) Being indirect and hedging (I

25
 wonder if you might possibly drop
by.)

Being optimistic (I’m sure you’ll Being pessimistic (I’m afraid you’ll
have a great time.) find it a bit boring.)

Being voluble (talking a lot) Being taciturn (not talking much)

Talking about ‘us’ Talking about things other than

‘us’

These two kinds of face strategies correspond to two fundamentally and, in some
ways, contradictory social needs that all humans experience: we all have the
need to be liked (sometimes referred to as our positive face), and we all have the
need to be respected (in the sense of not being imposed on or interfered with
(sometimes referred to as our negative face). When we interact with others, we
must constantly attend to their need to be liked and respected, and constantly
protect our own need to be liked and respected (Brown and Levinson 1987).
How we balance and negotiate these needs in communication is fundamental to
the way we show who we are in relation to the people around us.
In any given interaction we are likely to use a combination of both of these
strategies as we negotiate our relationships with the people with whom we are
interacting. In Section B6 we will go into more detail about how we decide which
of these strategies to use, when, and with whom.

Showing what we are doing: Framing strategies

In order to understand one another, we have to interpret what other people say
in the context of some kind of overall activity in which we are mutually involved.
One could think of many examples of utterances whose meanings change based
on what the people are doing when they utter them. The meaning of the
utterance by a doctor of the phrase, ‘please take off your clothes’ is different if
uttered in the context of a medical examination or in the context of his or her
apartment. For different kinds of activities we have different sets of expectations
about what kinds of things will be said and how those things ought to be
interpreted. We call these sets of expectations frames.

Goffman took his idea of frames from the work of the anthropologist Gregory
Bateson, who used it to explain the behavior of monkeys he had observed at the
zoo. Sometimes, he noticed, the monkeys displayed hostile signals, seemingly
fighting with or attempting to bite one another. It soon became clear to him,
however, that the monkeys were not actually fighting; they were playing. It then
occurred to him that they must have some way of communicating to one another
how a particular display of aggression should be interpreted, whether as an
invitation to fight or an invitation to play.

We bring to most interactions a set of expectations about the overall activity in

which we will be engaged, which Goffman called the primary framework of the

26
interaction. When we are a patient in a medical examination, for example, we
expect that the doctor will touch us, and we interpret this behavior as a method
for diagnosing our particular medical problem. When we attend a lecture, we do
so with an idea of what the activities of delivering a lecture and of listening to a
lecture involve.

Interaction, however, hardly ever involves just one activity. We often engage in a
variety of different activities within the primary framework. While lecturing, for
example, a lecturer might give explanations, tell jokes, or even rebuke members
of the audience if they are not paying attention. Similarly, medical examinations
might include multiple frames. In the reading in Section D6, Deborah Tannen and
Cynthia Wallat analyze how a doctor uses a ‘playing’ frame while examining a
young child, and then switches back to a ‘consultation’ frame when talking with
the child’s mother. And so, when we are interacting with people, we change the
activities we are involved in as we go along and, like Bateson’s monkeys, we need
ways to signal these ‘frame changes’ and ways to negotiate them with the people
with whom we are interacting.

27
B6 NEGOTIATING RELATIONSHIPS AND ACTIVITIES

Power and politeness

Whenever we interact with someone we always communicate something about

our relationship with them. We do this by using various discursive strategies,
which we have divided into two categories: involvement strategies and
independence strategies. Involvement strategies are strategies people use to
communicate friendliness or solidarity, and independence strategies are
strategies people use to communicate respect or deference.

In many cases, both parties in an interaction share a fairly clear idea about how
close they are and whether one has more power than the other, but in other
cases, participants in interaction need to negotiate their relationship. Such
negotiations are common, for example, as people move from more distant to
closer relationships, or when one person wishes to challenge another person’s
assertion of power or dominance.

Regardless of whether or not a relationship is seen as ‘negotiable’, we always

approach interactions with certain sets of expectations about how independence
and involvement strategies will be used to communicate information about
power and intimacy. We call these expectations face systems. Although
expectations about when independence and involvement strategies are
appropriate and what they mean vary across cultures and groups, most people
enter interaction with three basic ideas: 1) in interactions where the parties are
socially distant but relatively equal, both parties are likely to use independence
strategies (deference face system); 2) in interactions where people are close and
relatively equal, they are likely to use involvement strategies (solidarity face
system); and 3) in interactions in which one person has more power than the
other (regardless of their social distance) the more powerful one is more likely
to use involvement strategies and the less powerful one is more likely to use
independence strategies.

Like the conversational maxims we discussed in the last section, these ‘systems’
should not be treated as ‘rules’, but rather as broad sets of expectations people
draw on to decide how to act towards other people and how to interpret others’
behavior towards them. Since power and distance are relative rather than
absolute, and because interaction often involves the negotiation of power and
distance, people usually employ both independence and involvement strategies,
mixing them tactically depending on the situation and what they are trying to
accomplish in the interaction.

An example of the way participants often strategically mix independence and

involvement strategies can be seen in the following conversation between a
senior engineer (Martin) and his subordinate (Ollie) reported in Ladegaard
(2011):

73
 Martin: Happy birthday or (0.2) whatever it is (laughing)
Ollie: thank you (0.2) it’s actually a while ago
Martin: okay eh: Ollie//
Ollie: //there’s Danish pastry over there if you’re interested (0.2)
Martin: thanks ah: (0.6) (talks about tape recorder)
Martin: okay well to cut a long story short Sam called (0.2) and I’m not
sure how busy you are or what you’re doing right now (0.4)
Ollie: ah: we’re just about to launch the [name] project and ah:
Martin: okay
Ollie: so this is where we are [xxx] quite busy (0.5) but Sam called you
said
Martin: yes (0.2)
Ollie: and he? (0.3)
Martin: he needs some help here and now (0.2) he needs someone to
calculate the price of rubber bands (0.3) for the [name] project in India
Ollie: okay
Martin: they expect the customer to sign today (1.3)
Ollie: okay
(Ladegaard 2011: 14‐15)

In this example, Martin, the more powerful participant, begins using involvement
strategies, wishing Ollie happy birthday (although it is not his birthday) and
laughing. Ollie, on the other hand, though friendly, uses more independence
strategies, accepting the inappropriate, birthday wish and then using words like
‘actually’ and ‘a while’ to soften his revelation that it is not his birthday, and then
offering Martin some pastry in a way which is designed not to impose on him
(‘…if you’re interested’). Were Martin and Ollie equals and friends, the
inappropriate birthday wishes might have been answered in a more direct way
like, ‘What are you talking about? My birthday was ages ago!’, and the offer of
pastry might have been more insistent (Have some Danish!). In other words, the
mixture of involvement and independence strategies in the beginning of the
conversation are what one might expect within a hierarchical face system.

What happens next in the conversation, however, is rather interesting. Martin,

the more powerful person, changes to independence strategies, asking Ollie how
busy he is and making it clear that he does not wish to impose on him. In fact, he
acts so reluctant to make the request that Ollie practically has to drag it out of
him (‘but Sam called you said… and he?’). This, in fact, is the opposite of what one
might expect in a hierarchical relationship. Of course, this shift in politeness
strategies, with the more powerful participant using independence strategies
and the less powerful one showing more involvement does not really reflect a
shift in power. Rather, it is a clever strategy Martin has used to make it more
difficult for Ollie to refuse the request by putting him in the position of soliciting
it.

74
The point of this analysis is that, even though our expectations about face
systems form the background to how we communicate about relationships,
people often strategically confound these expectations to their own advantage.

One further factor, that determines which strategy a person will use to
communicate his or her relationship with another person is the topic of the
conversation he or she is engaged in. In cases in which the topic of the
conversation is serious or potentially embarrassing for either party, or in which
the weight of imposition is seen to be great, independence strategies will be
more common, whereas in situations where the topic is less serious, the outcome
more predictable and the weight of imposition seen to be relatively small,
involvement strategies are more common.

As can be seen in the example above, rather than as simple reflections of a priori
relationships of power and distance or the ‘weightiness’ of a particular topic, face
strategies can be regarded as resources that people use to negotiate social
distance, enact power relationships, and sometimes manipulate others into
doing things which they may not normally be inclined to do. A person might use
involvement strategies with another not because they are close, but because he
or she wants to create or strengthen the impression that there is a power
difference. Similarly, a person might use independence strategies not to create a
sense of distance from the person they are interacting with, but rather to endow
the topic under discussion with a certain ‘weightiness’. In other words, face
strategies are not just reflections of the expectations about relationships that
people bring to interactions but resources they make use of to manage and
sometimes change those relationships on a moment‐by‐moment basis.

 Look deeper into this topic online

Framing and contextualization cues

As we have seen above, conversational strategies like involvement and

independence are not just ways that we communicate and manage our
relationships with other people, but also ways that we communicate something
about what we are doing (the degree, for example, to which we think we are
imposing on other people). We also have other ways of signaling to people what
we think we are doing in an interaction, whether we, for example, are arguing,
joking, commiserating or making small talk. Whenever we speak, in fact, we
communicate not just the message contained in (or implied by) our words, but
also information about what we think we are doing and, therefore, how our
words should be interpreted. We call the signals we use to communicate this
information contextualization cues.

In Section A6 we said that there are basically two kinds of frames: broader
primary frameworks which consist of the relatively stable sets of expectations we
bring to particular situations (like lectures or medical consultations), and
smaller, more dynamic interactive frames, which consist of our negotiated ideas

75
about what we are doing moment by moment in a conversation, ideas which
often change rapidly in the course of an interaction. Although contextualization
cues are often important in signaling primary frameworks, they are particularly
important in the role they play in helping us to manage and negotiate interactive
frames.

Sometimes contextualization cues are verbal, that is, we signal what we are
doing through our choice of topic, vocabulary, grammar, or even the language
that we use. For example, in her analysis of the talk of teachers in bilingual
classrooms, Angel Lin (1996), has pointed out that when English teachers in
Hong Kong are focusing on teaching, they tend to use English, but when they are
engaged in reprimanding their students, they tend to switch to Cantonese.
Sometimes these verbal cues involve adopting a particular social language (see
Sections A4, B4) or certain genres (see Sections A3, B3) associated with
particular kinds of activities. A doctor, for example, might begin a consultation
with a period of small talk in which the language might be extremely informal
and the topic might range from the weather to a local sports team before he or
she ‘shifts gears’ and starts ‘talking like a doctor’.

One of the most obvious ways we signal shifts in frames verbally is through what
are known as discourse markers. These are words or phrases that often rather
explicitly mark the end of one activity and the beginning of another. A lecturer,
for example, might move from the pre‐lecture chatting and milling around frame
to the formal lecture frame with words like ‘Okay, let’s get started…’ Similarly,
the doctor might move from small talk to the more formal medical examination
by saying something like ‘So, how are you feeling?’ Discourse markers typically
consist of words like okay, so, well, and anyway, as well as more formal
connectors like first, next, and however. It is important to remember that
discourse markers do not always signal a shift in frame – sometimes they signal
other things like the relationship between one idea and another (see Section B2).

These verbal strategies are not the only ways, or even the most common ways,
people signal what they are doing when they talk. Contextualization cues also
include non­verbal signals delivered through things like gestures, facial
expressions, gaze, our use of space, and paralinguistic signals delivered through
alterations in the pitch, speed, rhythm or intonation of our voices. For this
reason, people who study frames and contextualization cues often pay a lot of
attention to marking things like stress, intonation and pausing and even facial
expressions, gestures and other movements when they produce transcripts of
the conversations they are studying.

These non‐verbal and paralinguistic contextualization cues are sometimes much

more subtle than verbal strategies and so more easily misunderstood. The way
they are used and interpreted might also vary considerably from group to group
or even person to person. In one of his most famous studies, Gumperz (1982a:
173‐174) found a mismatch between the ways South Asian servers in a staff
canteen in a British airport used intonation as a contextualization cue and the
ways their British customers interpreted them. The South Asian servers, for
example, used falling intonation when asking customers if they wanted gravy on

76
their meat (consistent with the conventions of their variety of English), but the
British customers, expecting the rising intonation they associated with a polite
offer, interpreted the servers’ behavior as rude. What this example tells us is that
contextualization cues do not in themselves contain information about what we
think we are doing – rather, they activate culturally conditioned assumptions
about context, interactional goals and interpersonal relationships that might be
different for different people.

As we said above, interactive frames are not static, but can change rapidly in the
course of an interaction. They are also, as their name implies, interactive – that is,
they are always a matter of negotiation between participants in the conversation,
and the way they are used and interpreted often has a great deal to do with
things that happened previously in the conversation and with the history of the
relationship between those involved. In other words, just as face strategies of
involvement and independence, while primarily providing information about
relationships, also give clues as to what we think we are doing and our attitude
towards it, framing strategies, while primarily signaling what we think we are
doing, also play an important role in managing relationships.

In an article called ‘Talking the Dog’, in which she examines how people use pets
to frame and reframe their utterances in interaction, Deborah Tannen (2004)
gives the following example of a conversation between a woman, Clara, and her
husband, Neil, in the presence of their dog, Rickie.

Clara: You leave the door open for any reason?

((short pause, sound of door shutting))
—> <babytalk> Rickie,
—> he’s helpin burglars come in,
—> and you have to defend us Rick.>
Tannen 2004:413)

In this example, Clara shifts frames from talking to her husband to talking to the
dog by altering her voice quality (adopting the high pitched and playful tone of
‘baby talk’). In a sense, though, she is still talking to her husband, communicating
to him ‘through’ the dog the potential seriousness of leaving the door open. By
addressing her remarks to the dog, however, and by adopting a different tone of
voice, she shifts the frame from scolding to playing, allowing her to get the
message across without threatening her husband’s face.

Sometimes participants in an interaction will experience disagreement regarding

‘what’s going on’. The way one person frames the conversation, for example, may
be at odds with the other person’s wishes, expectations or interpretation of the
situation. In some cases, they may simply accept the framing that has been
imposed by the other person, or they may contest or resist it by either
attempting to reframe the conversation using their own contextualization cues
or by breaking the frame altogether and engaging in a ‘meta‐conversation’ about
‘what’s going on’.

77
The film When Harry Met Sally (1988 Castle Rock Pictures) contains a number of
good examples of characters competitively negotiating frames in interaction. In
the following example, Harry, who is going out with Sally’s best friend, tells Sally
that he thinks she is attractive, and what ensures is a negotiation about what
such a statement means based on what he was ‘doing’ when he said it.

HARRY: You’re a very attractive person.

SALLY: Oh, thank you.
HARRY: Amanda never said you were so attractive.
SALLY: Maybe she doesn’t think I’m attractive.
HARRY: It’s not a matter of opinion. Empirically you are attractive.
SALLY: Harry, Amanda is my friend.
HARRY: So?
SALLY: So you’re going with her.
HARRY: So?
SALLY: So you’re coming on to me.
HARRY: No I wasn’t.
HARRY (continuing): What? Can’t a man say a woman is attractive
without it being a come‐on?
HARRY (continuing): All right. Let’s just say for the sake of argument it
was a come‐on. Okay. What do you want me to do? I take it back. All right,
I take it back.
SALLY: You can’t take it back.
HARRY: Why not?
SALLY: It’s already out there.
An awkward pause
HARRY: Ohm jeez. What are we supposed to do now? Call the cops? It’s
already out there.
SALLY: Just let it lie, okay?
HARRY: Right, right. Let it lie. That’s my policy. Let it lie… So, you want to
spend the night in the motel?
HARRY (continuing): See what I did? I didn’t let it lie.
SALLY: Harry –‐‐
HARRY: I said I would and then I didn’t ‐‐ ‐‐
SALLY: Harry ‐‐‐‐
HARRY: I went the other way ‐‐‐‐
SALLY: Harry ‐‐‐‐
HARRY: Yes?
SALLY: We are just going to be friends, okay?
HARRY: Yeah. Great. Friends. Best thing.

In this example, Harry tries to frame his initial compliment as an ‘objective

observation’ using formal language like ‘empirically’. Sally, however, labels what
he is doing as a ‘come on’, a label which he first resists with the question, ‘Can’t a
man say a woman is attractive without it being a come‐on?’, framing the
accusation as unreasonable and possibly sexist. He then half accepts her framing
and offers to ‘take it back’. This acceptance is only partial because he frames it as
‘hypothetical’ (‘Let’s just say for the sake of argument it was a come‐on…’). Sally,
however, does not accept his retraction, framing a ‘come on’ as an irreparable

78
breech in decorum, which Harry responds to by again shifting frames from
conciliation to mocking (‘What are we supposed to do now? Call the cops?’).
What happens after this, however, is particularity interesting. After agreeing to
‘let it lie’, that is, abandon this particular negotiation about framing, Harry then
issues what is unambiguously a ‘come‐on’, and then deflects her objections by
again engaging in meta‐conversation about his own framing (‘See what I did? I
didn’t let it lie… I said I would and then I didn’t… I went the other way…’).

Part of the humor in this scene lies in the fact that it foregrounds the process of
framing itself, a process which is usually left tacit in conversations. It also shows
how complex and contentious negotiations of framing can be, with parties not
only shifting frames, breaking frames, and attempting to reframe the utterances
of themselves and others, but also superimposing frames on top of other frames
in order to create strategic ambiguity (as when Harry imposes a ‘hypotheictal’
frame onto his admission of guilt).

79
C6 ANALYZING CONVERSATIONAL STRATEGIES
In this section we will further explore the strategies we use to manage
relationships (face) and activities (frames) in interaction. The kinds of
interaction we will use as examples in this section, however, are not from face‐
to‐face conversations, but rather from computer‐mediated interactions, in
particular, interactions using Facebook and MSN Messenger.

As we have explained, mediated interactions are different from face‐to‐face

spoken conversations in a number of ways. For one thing, in much computer‐
mediated communication, people type their ‘utterances’ rather than speaking
them. In addition, these interactions rarely involve the same kind of synchrony
that face‐to‐face conversation does. Whereas face‐to‐face interactions occur in
‘real time’, giving us access to other’s people’s utterances as they are forming
them, most computer mediated interaction is asynchronous, involving a ‘time
lag’ between production and reception, whether it be the momentary lag
between the time when one party types a message and the other person reads it
which we experience in instant messaging or the much longer time lags
associated with email, blogs and social networking sites.

Perhaps the most important difference between face‐to‐face interaction and

computer‐mediated interaction is that many of the non‐verbal and paralinguistic
resources available in face‐to‐face communication are not available in text based
computer mediated communication. This is significant because these are
precisely the resources people often use as contextualization cues to frame their
conversational activities, and they can also play an important role in the face
strategies of involvement and independence. Users of text based communication
tools, then, need to make use of different resources such as graphics, emoticons,
orthography and punctuation to fulfill the functions that non‐verbal and
paralinguistic communication do in face‐to‐face encounters.

Face strategies on Facebook

Perhaps more than any other kind of computer‐mediated communication, social

networking sites are designed to give users tools to communicate about and
manage their social relationships with others. Facebook is about ‘face’ precisely
in the sense in which we defined it in Section A6: ‘the negotiated public image
mutually granted to each other by participants in a communicative event.’ Users
of Facebook are centrally concerned with constructing and maintaining their
‘public images’, with saving face, and with ‘giving face’ to others.

It is not uncommon for people to have many Facebook ‘friends’ (hundreds or

even thousands), and yet they do not enjoy the same kind of relationships with
all of these ‘friends’: with some of them they are socially close, and with others
they are socially distant; some of them are their social equals, while others are in
a hierarchical relationship with them. The problem with Facebook, however, is

126
that it is biased towards a face system of symmetrical solidarity. Nearly all of the
resources it makes available, from the initial mechanism of ‘friending’, to photo
sharing, to the exchange of virtual tokens like ‘pokes’ and ‘vampire bites’ are
designed to express involvement. Some (see for example Kiesler 1986, Landow
1992) have even suggested that it is a fundamental characteristic of all
computer‐mediated communication that it flattens hierarchies and encourages
self‐disclosure, a phenomenon Joseph Walther (1996) calls ‘hyperpersonal
communication’.

For some users this is not a problem — the whole point of a social networking
site for them is to help them get closer to those in their social network — and it
certainly is not a problem for the company that runs Facebook since the more
people share with one another using involvement strategies, the more
information about them is available to sell to advertisers. It does become a
problem, however, when people who are accustomed to hierarchical or
deference face systems in face‐to‐face communication have to negotiate their
relationships in an environment that is biased towards involvement, as when
students and professors or employees and employers become ‘friends’.

These difficulties are especially salient in ‘wall posts,’ since these constitute
‘publically performed conversations’ which people who are not involved in
typically have access to. Therefore, the relationships people enact in these
interactions are not just negotiated between the interactants, but also displayed
to a larger audience.

The example below (figure C6.1) illustrates how one of my students strategically
mixed independence and involvement strategies when ‘tagging’ me in a picture
in her photo album.

Figure C6.1 Excerpt from the author’s Facebook Wall

The first thing that should be noted regarding this example is that ‘tagging’
someone in a photo on Facebook is a clear example of involvement. Not only
does it assume a relationship of solidarity, but also makes the assumption that
the person tagged does not mind advertising this relationship to other users.
Consequently, it is also a threat to the ‘negative face’ of the person who has been
tagged, potentially violating their desire for autonomy and privacy. There are
also other instances of involvement in this example, such as the optimistic and
complimentary message, the informal language and the use of emoticons (like :)

127
and :D) and unconventional spelling and punctuation (like ‘ur’, ‘jokessssss’, and
the repetition of the exclamation point at the end of the message).

At the same time, there are also instances of independence strategies, most
notably the use of the title ‘Prof. Jones’ to address me. What is interesting about
this is that, like many university professors, I am on a ‘first name basis’ with my
students. In other words, this student uses an independence strategy on
Facebook which she probably would not use in face‐to‐face interaction with me.
One reason for this may be to compensate for the involvement strategies that
otherwise dominate the message and to mitigate the potential threat to my
negative face.

 Find additional examples online

Activity

Analyze the postings on Facebook or some other social network service you use.
Does this service encourage the adoption of a particular face system among
users? Do the people in your network (including yourself) use different mixtures
of independence and involvement strategies when interacting with people with
whom they have different kinds of relationships? In particular, how do people
who are socially distant or who are in hierarchal relationships manage face
strategies? Can you find examples of interactions which would have been
managed differently had they taken place face‐to‐face?

Contextualization cues in instant messaging

As we have said above, text‐based computer mediated communication differs

from face‐to‐face conversation in that users do not have access to many of the
resources normally used to issue contextualization cues, such as body language,
facial expressions and paralinguistic signals. As a result, they have, over the
years, developed a multitude of other ways with which to frame and reframe
their utterances, including emoticons, screen names, status updates,
unconventional spelling, creative use of punctuation, and code‐mixing (the
mixing of words from different languages). A number of scholars (see for
example Danet et al. 1997, Herring 2001) have shown how users of chat and
instant messaging systems use such cues to signal ‘what’s going on’ in online
interaction.

Speakers of Chinese like many of the students I teach in Hong Kong also have at
their disposal written ‘final particles’, sounds that often occur at the end of
spoken utterances which signal the speaker’s attitude towards the utterance or
the hearer, which users of chat and instant messaging programs regularly insert
(often in Romanized form) in their written messages (though they hardly ever
appear in more formal written Chinese).

128
Below is an example of how such resources can be used as contextualization cues
in instant messaging exchanges. It is an excerpt from a conversation between
two university students in Hong Kong, one a female named Tina, and the other a
male named Barnett.

Barnett: u're....~?!
Tina: tina ar.......
Tina: a beautiful girl........
Tina: haha...
Tina: ^_^
Barnett: ai~
Barnett: i think i'd better leave right now....^o^!

The conversation starts out with Barnett asking for clarification of Tina’s
identity. The tilde (~) here signifies a lengthening of the previous utterance,
giving it a playful, insistent quality. Tina replies with her name, followed by a
Romanized final particle (‘ar’), which in Cantonese is often used to soften
affirmative statements so they do not sound too abrupt, followed by a number of
ellipsis marks (…) indicating that there is more to come. In her next message she
elaborates on her identity, referring to herself as ‘a beautiful girl’, which might be
interpreted as either a boast or an attempt at seduction. In her next two
messages, however, she puts a ‘joking frame’ around her previous description
with the words ‘haha…’ and a smiling emoticon (^_^). Barnett replies with ‘ai’ a
Romanization of the Cantonese word , often used as an expression of pain,
frustration or indignation, which he lengthens with a tilde (~) in the same way it
might be if spoken in a particularly exaggerated way. He then adds, in the next
message, that he thinks he had better leave the conversation, but reframes this as
a playful threat with the humorous emoticon ^o^ , which represents the face of a
clown.

What is going on in this short exchange, of course, has very little to do with Tina
giving an objective appraisal of her looks or even boasting, or with Barnett
expressing concern and threatening to terminate the conversation. Instead, this
is clearly an episode of playful teasing or flirting. Without the contextualization
cues supplied by such things as punctuation, emoticons, and tokens like ‘haha’,
however, the conversation would take on a very different meaning.

Activity

a. Choose an utterance which you might send to your friend via instant
messaging such as ‘u finish hw?’ (‘have you finished the homework?) and discuss
how the message could be ‘framed’ differently (as, for example, a warning, an
offer, a boast, a complaint, a sympathetic remark, etc.) by attaching to it one of
the emoticons from the range of default choices offered by MSN Messenger
(Figure C6.2).

129
Fig C6.2 MSN Messenger emoticons

b. Save an instant messaging conversation as a ‘history file’ and analyze it in

terms of how things like code choice, spelling, punctuation, capitalization and
emoticons are used to strategically frame and re‐frame messages.

 Do more activities online

130
D6 FRAMES IN INTERACTION
In the following classic article by Deborah Tannen and her collaborator Cynthia
Wallat, the authors give a clear an accessible definition of interactive frames and
the theoretical basis for this concept. They then go on to illustrate how
interactive frames operate in a medical examination between a pediatrician and
a child. As you read, play attention to the strategies the doctor uses to shift
frames from ‘playing’ with the child to explaining the child’s condition to the
mother, and how the doctor and the child’s mother negotiate conflicting frames.

Interactive frames and knowledge schemas in interaction: examples

from a medical examination/interview

Deborah Tannen and Cynthia Wallat (reprinted from. Social Psychology Quarterly
50 (2), 1987: 205‐16)

Interactive frames

The interactive notion of frame refers to a definition of what is going on in

interaction, without which no utterance (or movement or gesture) could be
interpreted. To use Bateson's classic example, a monkey needs to know whether
a bite from another monkey is intended within the frame of play or the frame of
fighting. People are continually confronted with the same interpretative task. In
order to comprehend any utterance, a listener (and a speaker) must know within
which frame it is intended: for example, is this joking? Is it fighting? Something
intended as a joke but interpreted as an insult (it could of course be both) can
trigger a fight.

Goffman (1974) sketched the theoretical foundations of frame analysis in the

work of William James, Alfred Schutz and Harold Garfinkel to investigate the
socially constructed nature of reality. Building on their work, as well as that of
linguistic philosophers John Austin and Ludwig Wittgenstein, Goffman developed
a complex system of terms and concepts to illustrate how people use multiple
frameworks to make sense of events even as they construct those events.
Exploring in more detail the linguistic basis of such frameworks, Goffman (1981)
introduced the term "footing" to describe how, at the same time that participants
frame events, they negotiate the interpersonal relationships, or "alignments,"
that constitute those events.

The interactive notion of frame, then, refers to a sense of what activity is

being engaged in, how speakers mean what they say. As Ortega y Gas'set (1959:
3), a student of Heidegger, puts it, "Before understanding any concrete
statement, it is necessary to perceive clearly 'what it is all about' in this
statement and 'what game is being played.' Since this sense is gleaned from the

194
way participants behave in interaction, frames emerge in and are constituted by
verbal and nonverbal interaction.

One author (Tannen) was talking to a friend on the telephone, when he suddenly
yelled, "YOU STOP THAT!" She knew from the way he uttered this command that
it was addressed to a dog and not her. She remarked on the fact that when he
addressed the dog, he spoke in something approximating a southern accent. The
friend explained that this was because the dog had learned to respond to
commands in that accent, and, to give another example, he illustrated the way
he plays with the dog: "I say, 'GO GIT THAT BALLI'" Hearing this, the dog began
running about the room looking for something to fetch. The dog recognized the
frame "play" in the tone of the command; he could not, however, understand the
words that identified an outer frame, “referring to playing with the dog," and
mistook the reference for a literal invitation to play.

This example illustrates, as well, that people (and dogs) identify frames in
interaction by association with linguistic and paralinguistic cues ‐ the way words
are uttered ‐ in addition to what they say. That is, the way the speaker uttered
"You stop that!" was associated with the frame "disciplining a pet" rather than
"chatting with a friend." Tannen drew on her familiarity with the use of linguistic
cues to signal frames when she identified her friend's interjection "You stop
that!" as addressed to a dog, not her. But she also drew on the knowledge that
her friend was taking care of someone's dog. This was part of her knowledge
schema about her friend. Had her schema included the information that he had a
small child and was allergic to dogs, she might have interpreted the same
linguistic cues as signaling the related frame, "disciplining a misbehaving child."
Furthermore, her expectations about how any speaker might express orders or
emotions, i.e. frame such expressions, were brought to bear in this instance in
conjunction with her expectations about how this particular friend is likely to
speak to her, to a dog and to a child; that is, a schema for this friend's personal
style. Thus frames and schemas interacted in her comprehension of the specific
utterance.

Interactive frames in the pediatric examination

Linguistic registers

A key element in framing is the use of identifiable linguistic registers. Register, as

Ferguson (1985) defines it, is simply "variation conditioned by use,"
conventionalized lexical, syntactic and prosodic choices deemed appropriate for
the setting and audience .... In addressing the child, the pediatrician uses
"motherese": a teasing register characterized by exaggerated shifts in pitch,
marked prosody (long pauses followed by bursts of vocalization), and drawn out
vowel sounds, accompanied by smiling.

For example, while examining Jody's ears with an ophthalmoscope (ear light),

195
the pediatrician pretends to be looking for various creatures, and Jody responds
with delighted laughter:

In stark contrast to this intonationally exaggerated register, the pediatrician uses

a markedly flat intonation to give a running account of the findings of her
examination, addressed to no present party, but designed for the benefit of
pediatric residents who might later view the video‐tape in the teaching facility.
We call this "reporting register." For example, looking in Jody's throat, the doctor
says, with only slight stumbling:

Doctor: Her canals are are fine, they're open, urn her tympanic membrane
was thin, and light.

Finally, in addressing the mother, the pediatrician uses conventional

conversational register, as for example:

Doctor: As you know, the important thing is that she does have difficulty
with the use of her muscles.

Register shifting

Throughout the examination the doctor moves among these registers.

Sometimes she shifts from one to another in very short spaces of time, as in the
following example in which she moves smoothly from teasing the child while
examining her throat, to reporting her findings, to explaining to the mother what
she is looking for and how this relates to the mother's expressed concern with
the child's breathing at night.

196
The pediatrician's shifts from one register to another are sometimes abrupt (for
example, when she turns to the child and begins teasing) and sometimes gradual
(for example, her reporting register in ''high arched palate" begins to fade into
conversational register with ''but there's no cleft," and come to rest firmly in
conversational register with "what we'd want to look for ... "). In the following
example, she shifts from entertaining Jody to reporting findings and back to
managing Jody in a teasing tone:

Frame shifting

Although register shifting is one way of accomplishing frame shifts, it is not the
only way. Frames are more complex than register. Whereas each audience is

197
associated with an identifiable register, the pediatrician shifts footings with each
audience. In other words, she not only talks differently to the mother, the child
and the future video audience, but she also deals with each of these audiences in
different ways, depending upon the frame in which she is operating.
The three most important frames in this interaction are the social encounter,
examination of the child and a related outer frame of its videotaping, and
consultation with the mother. Each of the three frames entails addressing each of
the three audiences in different ways. For example, the social encounter requires
that the doctor entertain the child, establish rapport with the mother and ignore
the video camera and crew. The examination frame requires that she ignore the
mother, make sure the video crew is ready and then ignore them, examine the
child, and explain what she is doing for the future video audience of pediatric
residents.

The consultation frame requires that she talk to the mother and ignore the crew
and the child ‐ or, rather, keep the child "on hold," to use Goffman's term, while
she answers the mother's questions. These frames are balanced nonverbally as
well as verbally. Thus the pediatrician keeps one arm outstretched to rest her
hand on the child while she turns away to talk to the mother, palpably keeping
the child "on hold."

Juggling frames

Often these frames must be served simultaneously, such as when the

pediatrician entertains the child and examines her at the same time, as seen in
the example where she looks in her ear and teases Jody that she is looking for a
monkey. The pediatrician's reporting register reveals what she was actually
looking at (Jody's ear canals and tympanic membrane). But balancing frames is
an extra cognitive burden, as seen when the doctor accidentally mixes the
vocabulary of her diagnostic report into her teasing while examining Jody's
stomach:

198
The pediatrician says the last line, "Is your spleen palpable over there?" in the
same teasing register she was using for peanut butter and jelly, and Jody
responds with the same delighted giggling "No" with which she responded to the
teasing questions about peanut butter and jelly. The power of the paralinguistic
cues with which the doctor signals the frame "teasing" is greater than that of the
words spoken, which in this case leak out of the examination frame into the
teasing register.

In other words, for the pediatrician, each interactive frame, that is, each
identifiable activity that she is engaged in within the interaction, entails her
establishing a distinct footing with respect to the other participants.

The interactive production of frames

Our analysis focuses on the pediatrician's speech because our goal is to show
that the mismatch of schemas triggers the frame switches which make this
interaction burdensome for her. Similar analyses could be performed for any
participant in any interaction. Furthermore, all participants in any interaction
collaborate in the negotiation of all frames operative within that interaction.
Thus, the mother and child collaborate in the negotiation of frames which are
seen in the pediatrician's speech and behavior.

For example, consider the examination frame as evidence in the pediatrician's

running report of her procedures and findings for the benefit of the video
audience.

Although the mother interrupts with questions at many points in the

examination, she does not do so when the pediatrician is reporting her findings
in what we have called reporting register. Her silence contributes to the
maintenance of this frame. Furthermore, on the three of seventeen occasions of
reporting register when the mother does offer a contribution, she does so in
keeping with the physician's style: Her utterances have a comparable clipped
style.

The homonymy of behaviors

199
Activities which appear the same on the surface can have very different
meanings and consequences for the participants if they are understood as
associated with different frames. For example, the pediatrician examines various
parts of the child's body in accordance with what she describes at the start as a
"standard pediatric evaluation." At times she asks the mother for information
relevant to the child's condition, still adhering to the sequence of foci of attention
prescribed by the pediatric evaluation. At one point, the mother asks about a skin
condition behind the child's right ear, causing the doctor to examine that part of
Jody's body. What on the surface appears to be the same activity ‐ examining the
child ‐ is really very different. In the first case the doctor is adhering to a preset
sequence of procedures in the examination, and in the second she is interrupting
that sequence to focus on something else, following which she will have to
recover her place in the standard sequence.

Conflicting frames

Each frame entails ways of behaving that potentially conflict with the demands of
other frames. For example, consulting with the mother entails not only
interrupting the examination sequence but also taking extra time to answer her
questions, and this means that the child will get more restless and more difficult
to manage as the examination proceeds. Reporting findings to the video audience
may upset the mother, necessitating more explanation in the consultation frame.
Perhaps that is the reason the pediatrician frequently explains to the mother
what she is doing and finding and why.

Another example will illustrate that the demands associated with the
consultation frame can conflict with those of the examination frame, and that
these frames and associated demands are seen in linguistic evidence, in this case
by contrasting the pediatrician's discourse to the mother in the examination
setting with her report to the staff of the Child Development Center about the
same problem. Having recently learned that Jody has an arteriovenous
malformation in her brain, the mother asks the doctor during the examination
how dangerous this condition is. The doctor responds in a way that balances the
demands of several frames:

200
The mother's question invoked the consultation frame, requiring the doctor to
give the mother the information based on her medical knowledge, plus take into
account the effect on the mother of the information that the child's life IS in
danger. However, the considerable time that would normally be required for
such a task is limited because of the conflicting demands of the examination
frame: the child is "on hold" for the exam to proceed. (Notice that it is admirable
sensitivity of this doctor that makes her aware of the needs of both frames.
According to this mother, many doctors have informed her in matter‐of‐fact
tones of potentially devastating information about her child's condition, without
showing any sign of awareness that such information will have emotional impact
on the parent. In our terms, such doctors acknowledge only one frame ‐
examination ‐ in order to avoid the demands of conflicting frames ‐ consultation
and social encounter. Observing the burden on this pediatrician, who
successfully balances the demands of multiple frames, makes it easy to
understand why others might avoid this.)

The pediatrician blunts the effect of the information she imparts by using
circumlocutions and repetitions; pausing and hesitating; and minimizing the
significant danger of the arteriovenous malformation by using the word "only"
("only danger"), by using the conditional tense ("that would be the danger"), and
by stressing what sounds positive, that they're not going to get worse. She
further creates a reassuring effect by smiling, nodding and using a soothing tone
of voice.

In reviewing the Video‐tape with us several years after the taping, the
pediatrician was surprised to see that she had expressed the prognosis in this
way‐‐ and furthermore that the mother seemed to be reassured by what was in
fact distressing information. The reason she did so, we suggest, is that she was
responding to the immediate and conflicting demands of the two frames she was
operating in: consulting with the mother in the context of the examination.

Evidence that this doctor indeed felt great concern for the seriousness of the

201
child's condition is seen in her report to the staff regarding the same issue:

Here the pediatrician speaks faster, with fluency and without hesitation or
circumlocution. Her tone of voice conveys a sense of urgency and grave concern.
Whereas the construction used with the mother, "only danger," seemed to
minimize the danger, the listing construction used with the staff ("sudden death,
intracranial hemorrhage"), which actually refers to a single possible event, gives
the impression that even more dangers are present than those listed.

Thus the demands on the pediatrician associated with consultation with the
mother; those associated with examining the child and reporting her findings to
the video audience; and those associated with managing the interaction as a
social encounter are potentially in conflict and result in competing demands on
the doctor's cognitive and social capacities.

Issues to consider

 The different frames the doctor uses with the young patient and with her
mother also involve different face strategies of independence and
involvement. What do you think the relationship between framing
strategies and face strategies is? How do they work together in
interactions? How does this relate to Goffman’s concept of ‘footing’?

 The way we interpret frames depends crucially on our ‘knowledge

schema’ for different situations. How do the knowledge schema the
participants in the interaction described have about a medical
consultations affect how they produce and interpret contextualization
cues? Can you think of a situation in which you or someone you know
approached a situation with an incomplete or faulty knowledge schema?
What were the consequences?

 At one point in this interaction the doctor mixes a formal medical register
(‘Is your spleen palpable over there?’) with a ‘teasing frame’, and the child
reacts as if this is part of the game. The authors use this as an example of
how non‐verbal contextualization cues can sometimes be so powerful as
to override the actual content of an utterance. Why do you think this is?
Can you think of any examples of this from your own experience?

202
 Sometimes we have to manage the demands of two or more frames at one
time. In this interaction, for example, the doctor has to manage
communicating medical information to both the mother and the students
watching the video and, at the same time, manage her young patient. In
what way are the demands of these three different frames incompatible?
What kinds of miscommunication can potentially arise from such
situations?

203