Working with Discourse

- meaning beyond the clause

J R Martin & David Rose

copyright J R Martin & David Rose

 i

Preface
In this book we have set out to provide readers with a relatively accessible set of tools
for discourse analysis informed by systemic functional linguistics. This book takes
Martin’s 1992 book English Text as point of departure, and extends this work taking into
account developments throughout the 90s, presenting it for a general audience
interested in discourse analysis. We decided to focus on discourse semantics rather
than social context or lexicogrammar because it seemed to us that, while there has
been a lot of genre and grammar analysis going on around the world, there is now a
need and expanding opportunities for work that bridges systematically between these
levels. This book attempts to fill that gap with analyses of meanings beyond the clause,
that makes contact with social context.

We are grateful to our functional linguistic colleagues around the world whose interest
has given value to this work over the years, at meetings and over the internet. Many of
these work in the field of educational linguistics, where a lot of the analysis considered
here has been deployed – especially in the context of Australia’s distinctive genre-based
literacy programs across education sectors. Without this dialectic of theory and
practice, the ideas in this book would never have evolved. In particular we would like to
thank Suzanne Eggins who worked with Jim on an earlier version of this monograph.
Her ideas about how such a book might be organised were a major influence on its
present form.

Our intellectual debt to Ruqaiya Hasan and Michael Halliday will be obvious to readers
on every page. We respectfully dedicate the book to them, in gratitude for all the
meanings they have given us to explore – meanings which have shaped our lives in so
many many ways.

Sydney & Hong Kong, September 2001

copyright J R Martin & David Rose Working with Discourse

 ii

For Ruqaiya & Michael

whose meanings we’re embroidering upon

copyright J R Martin & David Rose Working with Discourse

 iii

Contents
Chapter 1 Interpreting social discourse
1.0 An invitation
1.1 A framework for discussion
1.2 Genre
1.3 Language, power and ideology
1.4 How this book is organised
1.5 How to use this book
Chapter 2 APPRAISAL - negotiating attitudes
2.0 Negotiating attitudes
2.1 Kinds of attitudes
2.2 Amplifying attitudes
2.3 Sources of attitudes
2.4 Prosody and genre
2.5 More detail on kinds of attitude
Chapter 3 IDEATION - representing experience
3.0 Representing experience
3.1 Sequences of meanings
3.2 Doing – focusing on activities
3.3 Being – focusing on entities
3.4 Classifying and describing within elements
3.5 Ways of participating
3.6 Building up a picture - taxonomic relations
3.7 Types of taxonomic relations
3.8 Re-construing experience - ideational metaphor
Chapter 4 CONJUNCTION - connecting events
4.0 The logic of discourse
4.1 Four kinds of logic
4.3 Connecting arguments
4.4 Continuatives
4.5 Countering our expectations
4.6 Conjunction resources in full
4.7 Displaying connections – conjunction analysis
4.8 Logical metaphor
Chapter 5 IDENTIFICATION – tracking participants
5.0 Keeping track
5.1 Who’s who? - identifying people
5.2 What’s what? – identifying things
5.3 Where to look?
5.4 Tracking and genre
5.5 Identification systems in full
Chapter 6 PERIODICITY – information flow
6.0 Waves of information
6.1 Little waves – Themes and News
6.2 Bigger waves – hyperThemes and hyperNews
6.3 Tidal waves – macroThemes, macroNews, and beyond

copyright J R Martin & David Rose Working with Discourse

 iv

6.4 How texts grow – hierarchies and series

6.5 Hard reading
6.6 A note on headings
6.7 Texture - phasing discourse systems
Chapter 7 TACKLING A TEXT
7.1 Getting going
7.2 Outside-in
7.3 Inside-out
7.4 Inauguration day - from past to present
7.5 The Cost of Courage - from domination to freedom
7.6 The Meaning of Freedom - from self to community
7.7 Reprise
Chapter 8 CONNECTIONS
8.1 Context (register and genre)
8.3 Multimodal Discourse Analysis (MDA)
8.4 Critical Discourse Analysis (CDA)
8.5 Data
8.6 Envoi
References

copyright J R Martin & David Rose Working with Discourse

 1

Chapter 1 Interpreting social discourse

1.0 An invitation

In this book we are concerned with interpreting discourse by analysing it. For us this
means treating discourse as more than a sequence of clauses; we want to focus on
meaning beyond the clause – on resources that lead us from one clause to another as a
text unfolds. For us this also means that we treat discourse as more than an incidental
manifestation of social activity; we want to focus on the social as it is constructed
through texts – on the constitutive role of meanings in social life. In a sense then this
book is an invitation to grammarians to reconsider meaning in the clause from the
perspective of meaning in texts; and it is also an invitation to social theorists to
reconsider social activity as meaning we negotiate through texts.

Our starting point then, for interpreting social discourse, is with texts in social contexts.
Social discourse rarely consists of just single clauses, rather social contexts develop as
sequences of meanings comprising texts. Since each text is produced interactively
between speakers, and between writers and (potential) readers, we can use it to
interpret the interaction it manifests. And since each interaction is an instance of the
speakers’ culture, we can also use the text to interpret aspects of the culture it
manifests.

We should emphasise that although we can assign a name to each of these

phenomena, a clause, a text or a culture are not ‘things’, but social processes that
unfold at different time scales. Culture unfolds through uncountable series of situations,
as our lives unfold through such situations as learners, speakers and actors, producing
texts that unfold as sequences of meanings. The relationship between these
phenomena is schematised in Figure 1.1, illustrating the scaling in size and complexity
from clause to text to culture.

copyright J R Martin & David Rose Working with Discourse

 2

Figure 1.1: Clause-text-culture

clause
(Helena in t ime and p lace)
My st o ry b egins
in my lat e t eenage y ears
text
asa f arm g irl (Helena' s sto ry o f injus t ic e)
in th e Beth lehem d istr ict …
As an eighteen -yea r-o ld, I met a young
cult ure
man in his twen ties . He wa s wo rking in
(t he stru ggle f or just ice in South A f rica)
a top secu rity structure. It w as the
beg inning of a beau tiful re lationship.
We even spoke abou t marri age . A
bubbly, vivac ious man who be amed out
wild energy . Sha rply intelligen t. Even
if he was an Eng lish man, he w as popul ar
with all the ‘Boer’ Afrikane rs. And a ll
my girlfriend s env ied me. Then one day
he said he was going on a ‘trip’. ‘We
won ’t see each oth er aga in… maybe
neve r eve r aga in.’ I was torn to pieces .
So was he.

Figure 1.1 shows one clause as an instance of the story of ‘Helena’, whose life was
caught up in the injustices of apartheid South Africa - as Helena’s story is one instance
of the cultural changes that culminated with the release of Nelson Mandela and the
overthrow of apartheid. Helena’s story is one of the texts we use to interpret discourse
throughout this book, and we will return to it often, examining its sequences of meanings
from different perspectives, to understand just how it manifests the changing culture it is
part of.

In order to keep our analyses manageable, we focus intensively in the book on a small
set of texts, all concerned with the processes of Truth and Reconciliation in South
Africa. We chose this context for two reasons: because we expect it will be relatively
familiar to many readers, and because we believe the reconciliation process in post-
apartheid South Africa is an inspiration for engaging with difference in the post-colonial
world. One text we focus on is Helena’s story, which is about the effects on herself and
the men in her life of their violations of other people’s human rights; one is an argument
by Desmond Tutu, about amnesty for such offenders, from his recent book No Future
without Forgiveness; and another is the act of parliament establishing the Truth and
Reconciliation Commission. These are complementary kinds of texts that allow us to
explore a wide range of discourse meanings within a single field of social activity. As
another strategy for keeping things manageable we deal mainly with written texts,
although the tools we introduce can all be used for spoken discourse (see Eggins &
Slade 1997 for excellent scaffolding for spoken discourse analysis). And we’ll be

copyright J R Martin & David Rose Working with Discourse

 3

sticking to English, although we know from our own work on very different languages,
Tagalog and Pitjantjatjara, and from the work of SFL researchers around the world, that
comparable resources are found across languages (Caffarel et al. 2001)

1.1 A framework for discussion

Any description or analysis involving language implies some theory of how language
works, and we want to be explicit about the model we are using, rather than leaving it
unsaid. The framework for our discussion is the model of language in social context
that has been developed within the broad field of systemic functional linguistics (SFL).
SFL researchers have been actively concerned for several generations with the
semantics of discourse. However we will not assume that readers are familiar with this
theory, or with its grammatical descriptions of English and other languages. Rather we
will introduce relevant aspects of theory and description as they are required, illustrating
them with examples from our texts. In the process we will gradually introduce a shared
language for talking about discourse, a metalanguage that includes both the model of
language in context, and the terms we use for talking about it. We will use only those
technical terms that we need to consolidate understandings and make them portable, so
that they can be used easily for as wide a range of analytical tasks as possible. And
portability is one aim of this book. We want to build up a tool kit for discourse analysis
that readers can take away with them when they go. But we’ll build up this
metalinguistic tool kit the way people learn languages, through experiencing meaningful
instances in actual texts.

SFL has been described as an ‘extravagant’ theory; its extravagance has evolved to
manage the complexity of the phenomenon it describes. But despite the complexity of
language in social contexts, the basic principles developed in SFL for managing it are
relatively simple. To begin with we will briefly introduce two general perspectives for
looking at the phenomena of discourse. These two perspectives are:

 three levels of language: as grammar - as discourse – and as social context

(known as the strata of language)

 three general functions of language in social contexts: to enact our relationships,

to represent our experience, and to organise discourse as meaningful text
(known as metafunctions).

Strata: grammar - discourse – social context

The focus of this book is on the analysis of discourse. In SFL, discourse analysis
interfaces with the analysis of grammar and the analysis of social activity – somewhere
between the work of grammarians on the one hand and social theorists on the other.
This has partly to do with the size of what we’re looking at – texts are bigger than a
clause and smaller than a culture. Grammarians are particularly interested in types of
clauses and their elements. But texts are usually bigger than single clauses, so a
discourse analyst has more to worry about than a grammarian (expanded horizons). By
the same token, cultures manifest themselves through a myriad of texts, and social

copyright J R Martin & David Rose Working with Discourse

 4

theorists are more interested in how social contexts are related to one another than in
how they are internally organised as texts (global horizons). Discourse analysis
employs the tools of grammarians to identify the roles of wordings in passages of text,
and employs the tools of social theorists to explain why they make the meanings they
do. These two points of view on discourse are illustrated in Figure 1.2.

Figure 1.2: Points of view on discourse: from social activity and from grammar

social act ivity

My st ory begins
in my lat e t eenage y ears
asa f arm g irl in th e Beth lehem
dist rict E ast ern Free St at e. As an
discourse
eight een-y ear-o ld, I met a y oun g man in
his tw ent ies. H e wa s wor ki ng in a t op
sec urit y stru c tur e. It w as t he beginning
of a beaut if ul r elat io nsh ip. We ev en
spoke about m arriage. A b ubbly ,
v iv ac iou s man who beamed out wild
energ y. Sh arp ly int elligent. Even if he
wasa n Englis hman , h e gramma r with
waspopu lar
all th e ‘Boer’ A f rikaners. And all my
girlf riends env ied me. Then one
day h e said h e wasg oing
on a ‘tr ip’.

In Figure 1.2, grammar, discourse and social activity are symbolised as a series of
circles, in which discourse nestles within social activity and grammar nestles within
discourse, suggesting three complementary perspectives on a single complex
phenomenon. This type of diagram is often used in SFL to symbolise its evolving model
of language in social context.

Realisation: culture-meaning-wording

What is the relation between grammar, discourse and social context? Obviously
cultures aren’t just a combination of texts, and likewise texts aren’t just a combination
clauses. Social activity, discourse and grammar are different kinds of phenomena,
operating at different levels of abstraction: a culture is more abstract than a text, and the
meanings that make up a text are in turn more abstract than the wordings that express
them. The relation between these strata is described in SFL as realisation; social
contexts are realised as texts which are realised as sequence of clauses.

copyright J R Martin & David Rose Working with Discourse

 5

Realisation is a kind of re-coding – like the mapping of hardware through software to the
images and words we see on the screen on our computers. Another way of thinking
about this is symbolisation. The Australian Aboriginal flag for example consists of two
horizontal bands of equal size, red on the bottom and black above, with a yellow circle
in the middle, shown in Figure 1.3.

Figure 1.3: Australian Aboriginal flag

The black band in the flag symbolises the indigenous people of Australia (and the night
sky on which the Dreaming is written in the stars); the red band symbolises the red
Australian earth (and the blood that Aboriginal people have shed struggling to share it
with Europeans); and the yellow circle symbolises the sun (and a new dawn for social
justice for Aboriginal people). Symbolising is an important aspect of realisation, since
grammar both symbolises and encodes discourse, just as discourse both symbolises
and encodes social activity. The concept of realisation embodies the meanings of
‘symbolising’, ‘encoding’, ‘expressing’, ‘manifesting’ and so on.

As the meaning of the Aboriginal flag is more than the sum of its shapes and colours, so
too is discourse more than the sum of its wordings, and culture more than the sum of its
texts. For example, here’s part of the story we’ll be working on later. The narrator,
Helena, is talking about separating from her first love:

Then one day he said he was going on a ‘trip’. ‘We won’t see each other again… maybe never ever
again.’ I was torn to pieces.

The last clause here, I was torn to pieces, tells us how Helena felt; but because of the
way meaning unfolds through the discourse phases of ‘meeting’, ‘description’ and
‘leaving’ it also tells us why she felt upset – there’s an explanation going on which
transcends the meaning of the individual clauses. Taken one by one, each clause
describes what happened; taken together they explain it.

Similarly for text and culture. At the beginning of the argument we’ll be working on
below, Desmond Tutu asks a question and comments on its significance:

copyright J R Martin & David Rose Working with Discourse

 6

So is amnesty being given at the cost of justice being done? This is not a frivolous question, but a very
serious issue, one which challenges the integrity of the entire Truth and Reconciliation process.

He goes on to give reasons why he’d answer no; and earlier he posed a similar
question, which his exposition is just now getting round to answering:

Are the critics right: was the Truth and Reconciliation process immoral?

But there is more to his question that what precedes or follows in his book. The
question strikes to the heart of the whole reconciliation process in South Africa and the
role played in it by the Truth and Reconciliation Commission. A whole range of texts
have posed the question, as a glance at Tutu’s book or a visit to the Truth and
Reconciliation website shows. The social processes round the issue are very complex,
involving all kinds of discourse and a great range of interests. The social meaning
transcends the meaning of the individual texts through which it is negotiated. What Tutu
refers to as the entire Truth and Reconciliation process is a higher order meaning
comprising just an aspect of the cultures we recognise as post-colonial episodes in our
unfolding world.

Metafunctions: interpersonal-ideational-textual

The SFL model of language in social context recognises three general social functions
that we use language for: i) to enact our social relationships; ii) to represent our
experience to each other; and iii) to organise our enactments and representations as
meaningful text. These are known as the metafunctions of language in social activity:

 the interpersonal metafunction to enact relationships

 the ideational metafunction to represent experience
 the textual metafunction to organise text.

As social discourse unfolds, these three functions are interwoven with each other, so
that we can achieve all three social functions simultaneously. In other words we can
look at any piece of discourse from either of these three perspectives, and identify
different functions realised by different patterns of meaning.

In this book, each chapter considers sets of meanings serving one or another of these
metafunctions. These sets of meanings are known as discourse systems. The name of
each chapter is the name of the particular discourse system that it considers. These
are grouped alongside their metafunction in Table 1.1. The table also gives a brief
gloss of the overall function of each system.

Table 1.1: Chapters, discourse systems and metafunctions

Chapters metafunction
 APPRAISAL ‘negotiating attitudes’ interpersonal
 IDEATION ‘representing experience’ ideational
 CONJUNCTION ‘connecting events’ ideational
 IDENTIFICATION ‘tracking people and things’ textual

copyright J R Martin & David Rose Working with Discourse

 7

 PERIODICITY ‘the rhythm of discourse’ textual

1.2 Genre

We use the term genre in this book to refer to different types of texts that enact various
types of social contexts. As children, we learn to recognise and distinguish the typical
genres of our culture, by attending to consistent patterns of meaning as we interact with
others in various situations. Since patterns of meaning are relatively consistent for each
genre, we can learn to predict how each situation is likely to unfold, and learn how to
interact in it.

Such predictable patterns of meaning can vary from the relatively simple range of
language resources we might use to greet our neighbours, or to buy goods in a shop, to
the more complex meanings we might find in scientific reports or political debates. But
even such complex meanings fall into consistent patterns that make it possible for us to
recognise and predict how each genre is likely to unfold, and so manage new
information, and interact appropriately and strategically.

The number of recognisably distinct genres in any culture may be quite large, but not
unmanageably so. In contemporary western culture we could name many spoken
genres whose patterns of meaning are more or less predictable, such as greetings,
service encounters, casual conversations, arguments, telephone enquiries, instructions,
lectures, debates, plays, jokes, games and so on; and within each of these general
types, we could name many more specific genres. In this book we will mainly be
exploring written genres that we will name and describe as we go.

For us a genre is a staged, goal oriented social process. Social because we participate
in genres with other people; goal oriented because we use genres to get things done;
staged because it usually takes us a few steps to reach our goals. In this book we’re
focusing on three families of genres – story, argument and legislation. We’ll look briefly
at their staging here so we can get a feel for the basic organisation of these texts. As
we develop our five strands of discourse analysis we’ll look more and more closely at
how they are organised.

First Helena’s Story. This story is presented in Desmond Tutu’s book No Future without
Forgiveness to support his argument for amnesty for human rights violators, as part of
South Africa’s Truth and Reconciliation process. The narrator is introduced by Tutu,
and Helena orients her story by setting it in time and place:

Orientation My story begins in my late teenage years as a farm girl in the Bethlehem district of
Eastern Free State.

The story then unfolds as a story genre known as an ‘exemplum’ - a kind of moral tale
related to fables, parables and gossip. Its social purpose is to present a problematic
incident and then interpret it for the audience, commenting on the behaviour of the
people involved. This story type contrasts with the ‘narrative’ story type which typically

copyright J R Martin & David Rose Working with Discourse

 8

presents a problem which is then resolved by the lead characters. An exemplum

consists of the basic stages Orientation, Incident and Interpretation.

Helena’s story includes two Incident stages. The first Incident deals with her first love
and the second Incident with her second love. Each Incident has basically the same
organisation, involving three phases. First Helena meets her love, he then starts
working in secret police operations, and he and Helena then have to face the
repercussions. This structure is summarised as follows, with the first clause of each
phase.

Incident 1
‘falling in love’ As an eighteen-year-old, I met a young man…
‘operations’ Then one day he said he was going on a ‘trip’.
‘repercussions’ More than a year ago, I met my first love again…

Incident 2
‘falling in love’ After my unsuccessful marriage, I met another policeman.
‘operations’ Then he says: He and three of our friends have been promoted.
‘repercussions’ After about three years with the special forces, our hell began.

Helena then goes on to interpret the significance of these events, and this Interpretation
stage goes through three phases. The first phase outlines her new knowledge of the
crimes committed by her man under orders from ‘those at the top’; in the second she
understands and identifies with the struggle of black South Africans; in the third she
accuses the cowardly leaders of her own people.

Interpretation
‘knowledge’ Today I know the answer to all my questions and heartache.
‘black struggle’ I finally understand what the struggle was really about.
‘white guilt’ What do we have? Our leaders are too holy and innocent.

Following her Interpretation, Helena ends her story with a Coda, quoting her ‘wasted
vulture’s’ perspective on his punishment.

Coda I end with a few lines that my wasted vulture said to me…

The stages of a genre are relatively stable components of its organisation, that we can
recognise in some form in text after text of the genre - such as the Orientation, Incident
and Interpretation stages of an exemplum. These stages are some of the basic
resources of the culture for organising discourse at the level of the text; we use initial
capitals to label them. But phases within each stage are much more variable; often, as
in Helena’s story, phases may be unique to the particular text, so we label them
notionally, with quotation marks. The story is presented below with its stages and
phases indicated.

Orientation
My story begins in my late teenage years as a farm girl in the Bethlehem district of Eastern Free State.

Incident 1

copyright J R Martin & David Rose Working with Discourse

 9

‘falling in love’
As an eighteen-year-old, I met a young man in his twenties. He was working in a top security structure. It
was the beginning of a beautiful relationship. We even spoke about marriage. A bubbly, vivacious man
who beamed out wild energy. Sharply intelligent. Even if he was an Englishman, he was popular with all
the ‘Boer’ Afrikaners. And all my girlfriends envied me.

‘operations’
Then one day he said he was going on a ‘trip’. ‘We won’t see each other again… maybe never ever
again.’ I was torn to pieces. So was he. An extremely short marriage to someone else failed all because I
married to forget.

‘repercussions’
More than a year ago, I met my first love again through a good friend. I was to learn for the first time that
he had been operating overseas and that he was gong to ask for amnesty. I can’t explain the pain and
bitterness in me when I saw what was left of that beautiful, big, strong person. He had only one desire –
that the truth must come out. Amnesty didn’t matter. It was only a means to the truth.

Incident 2
‘falling in love’
After my unsuccessful marriage, I met another policeman. Not quite my first love, but an exceptional
person. Very special. Once again a bubbly, charming personality. Humorous, grumpy, everything in its
time and place.

‘operations’
Then he says: He and three of our friends have been promoted. ‘We’re moving to a special unit. Now,
now my darling. We are real policemen now.’ We were ecstatic. We even celebrated. He and his friends
would visit regularly. They even stayed over for long periods. Suddenly, at strange times, they would
become restless. Abruptly mutter the feared word ‘trip’ and drive off. I … as a loved one…knew no other
life than that of worry, sleeplessness, anxiety about his safety and where they could be. We simply had to
be satisfied with: ‘What you don’t know, can’t hurt you.’ And all that we as loved ones knew…was what we
saw with our own eyes.

‘repercussions’
After about three years with the special forces, our hell began. He became very quiet. Withdrawn.
Sometimes he would just press his face into his hands and shake uncontrollably. I realised he was
drinking too much. Instead of resting at night , he would wander from window to window. He tried to hide
his wild consuming fear, but I saw it. In the early hours of the morning between two and half-past-two, I jolt
awake from his rushed breathing. Rolls this way, that side of the bed. He’s pale. Ice cold in a sweltering
night – sopping wet with sweat. Eyes bewildered, but dull like the dead. And the shakes. The terrible
convulsions and blood-curdling shrieks of fear and pain from the bottom of his soul. Sometimes he sits
motionless, just staring in front of him. I never understood. I never knew. Never realised what was being
shoved down his throat during the ‘trips’. I just went through hell. Praying, pleading: ‘God, what’s
happening? What’s wrong with him? Could he have changed so much? Is he going mad? I can’t handle
the man anymore! But, I can’t get out. He’s going to haunt me for the rest of my life if I leave him. Why,
God?’

Interpretation
‘knowledge’
Today I know the answer to all my questions and heartache. I know where everything began, the
background. The role of ‘those at the top’, the ‘cliques’ and ‘our men’ who simply had to carry out their
bloody orders… like ‘vultures’. And today they all wash their hands in innocence and resist the realities of
the Truth Commission. Yes, I stand by my murderer who let me and the old White South Africa sleep
peacefully. Warmly, while ‘those at the top’ were again targeting the next ‘permanent removal from
society’ for the vultures.

copyright J R Martin & David Rose Working with Discourse

 10

‘black struggle’
I finally understand what the struggle was really about. I would have done the same had I been denied
everything. If my life, that of my children and my parents was strangled with legislation. If I had to watch
how white people became dissatisfied with the best and still wanted better and got it. I envy and respect
the people of the struggle – at least their leaders have the guts to stand by their vultures, to recognise their
sacrifices.

‘white guilt’
What do we have? Our leaders are too holy and innocent. And faceless. I can understand if Mr F.W. de
Klerk says he didn’t know, but dammit, there must be a clique, there must have been someone out there
who is still alive and who can give a face to ‘the orders from above’ for all the operations. Dammit! What
else can this abnormal life be than a cruel human rights violation? Spiritual murder is more inhumane than
a messy, physical murder. At least a murder victim rests. I wish I had the power to make those poor
wasted people whole again. I wish I could wipe the old South Africa out of everyone’s past.

Coda
I end with a few lines that my wasted vulture said to me one night: ‘They can give me amnesty a thousand
times. Even if God and everyone else forgives me a thousand times – I have to live with this hell. The
problem is in my head, my conscience. There is only one way to be free of it. Blow my brains out.
Because that’s where my hell is.’ [Tutu 1999: 49-51]

Next the argument. Tutu’s text belongs to the argument genre known as ‘exposition’.
An exposition consists of the basic stages Thesis and supporting Arguments. Its social
purpose is to persuade an audience to the writer’s point of view, the ‘thesis’.
Expositions contrast with the argument genre known as ‘discussion’, in which two or
more points of view are presented and one argued for over the others.

In this exposition Tutu is debating whether giving amnesty is just. In the beginning,
instead of stating his Thesis in the usual way, he poses the issue as a question.

Thesis So is amnesty being given at the cost of justice being done?

He then develops three Arguments as to why his answer is ‘No.’ Each of these three
Arguments has two phases. In the first phase Tutu gives the ‘grounds’ on which he is
arguing, and in the second he reaches a ‘conclusion’ on the basis of this evidence.
Tutu uses linkers also and further to guide us from one Argument to the next, and each
conclusion is introduced with the linker thus. These linkers are underlined below.

Argument 1 The Act required that where the offence is a gross violation…
‘grounds’ the application should be dealt with in a public hearing
‘conclusion’ Thus there is the penalty of public exposure and humiliation

Argument 2 It is also not true that … amnesty encourages impunity

‘grounds’ because amnesty is only given to those who plead guilty
‘conclusion’ Thus the process in fact encourages accountability

Argument 3 Further, retributive justice…is not the only form of justice.

‘grounds’ there is another kind of justice, restorative justice,
‘conclusion’ Thus we would claim that… justice, is being served

copyright J R Martin & David Rose Working with Discourse

 11

In Tutu’s original text, Helena’s story follows the first Argument, supporting its
conclusion, and the whole exposition is part of a longer debate. We’ll come back to the
ways genres fit together in discourse in Chapters 7 and 8.

Thesis
So is amnesty being given at the cost of justice being done? This is not a frivolous question, but a very
serious issue, one which challenges the integrity of the entire Truth and Reconciliation process.

Argument 1
‘grounds’
The Act required that where the offence is a gross violation of human rights – defined as an abduction,
killing torture or severe ill-treatment – the application should be dealt with in a public hearing unless such a
hearing was likely to lead to a miscarriage of justice (for instance, where witnesses were too intimidated to
testify in open session). In fact, virtually all the important applications to the Commission have been
considered in public in the full glare of television lights.

‘conclusion’
Thus there is the penalty of public exposure and humiliation for the perpetrator. Many of those in the
security forces who have come forward had previously been regarded as respectable members of their
communities. It was often the very first time that their communities and even sometimes their families
heard that these people were, for instance, actually members of death squads or regular torturers of
detainees in their custody. For some it has been so traumatic that marriages have broken up. That is
quite a price to pay.

[The South Africa Broadcasting Corporation’s radio team covering the Truth and Reconciliation
Commission received a letter from a woman calling herself Helena (she wanted to remain anonymous for
fear of reprisals) who lived in the eastern province of Mpumalanga. They broadcast substantial extracts:
…]

Argument 2
‘grounds’
It is also not true that the granting of amnesty encourages impunity in the sense that perpetrators can
escape completely the consequences of their actions, because amnesty is only given to those who plead
guilty, who accept responsibility for what they have done. Amnesty is not given to innocent people or to
those who claim to be innocent. It was on precisely this point that amnesty was refused to the police
officers who applied for it for their part in the death of Steve Biko. They denied that they had committed a
crime, claiming that they had assaulted him only in retaliation for his inexplicable conduct in attacking
them.

‘conclusion’
Thus the process in fact encourages accountability rather than the opposite. It supports the new culture of
respect for human rights and acknowledgment of responsibility and accountability by which the new
democracy wishes to be characterised. It is important to note too that the amnesty provision is an ad hoc
arrangement meant for this specific purpose. This is not how justice is to be administered in South Africa
for ever. It is for a limited and definite period and purpose.

Argument 3
‘grounds’
Further, retributive justice – in which an impersonal state hands down punishment with little consideration
for victims and hardly any for the perpetrator – is not the only form of justice. I contend that there is
another kind of justice, restorative justice, which is characteristic of traditional African jurisprudence. Here
the central concern is not retribution or punishment but, in the spirit of ubuntu, the healing of breaches, the
redressing of imbalances, the restoration of broken relationships. This kind of justice seeks to rehabilitate
both the victim and the perpetrator, who should be given the opportunity to be reintegrated into the
community he or she has injured by his or her offence. This is a far more personal approach, which sees

copyright J R Martin & David Rose Working with Discourse

 12

the offence as something that has happened to people and whose consequence is a rupture in
relationships.

‘conclusion’
Thus we would claim that justice, restorative justice, is being served when efforts are being made to work
for healing, for forgiveness and for reconciliation. [Tutu 1999: 48-52]

Finally the legislation. Here we have chosen the act of parliament establishing the Truth
and Reconciliation Commission. This is a much longer text including the following
Chapters:

CHAPTER 1 Interpretation and application

CHAPTER 2 Truth and Reconciliation Commission
CHAPTER 3 Investigation of Human Rights Violations
CHAPTER 4 Amnesty mechanisms and procedures
CHAPTER 5 Reparation and rehabilitation of victims
CHAPTER 6 Investigations and hearings by Commission
CHAPTER 7 General provisions

These Chapters are of course divided into smaller sections and sub-sections, which
we’ll pass over here. But before the Chapters get going the act outlines nine ‘purposes’;
and following this the act reviews six constitutional ‘motivations’ for its enactment, each
introduced with the causal linker since, foregrounded with capital letters. Here we will
present just the purposes and motivations phases, leaving the Chapters to an appendix.

PROMOTION OF NATIONAL UNITY AND RECONCILIATION ACT, 1995.

It is hereby notified that the President has assented to the following Act which is hereby published for
general information:-

‘purposes’
To provide for the investigation and the establishment of as complete a picture as possible of the nature,
causes and extent of gross violations of human Rights… ;
the granting of amnesty to persons who make full disclosure of all the relevant facts… ;
affording victims an opportunity to relate the violations they suffered;
the taking of measures aimed at the granting of reparation… ;
reporting to the Nation about such violations and victims;
the making of recommendations aimed at the prevention of the commission of gross violations of human
rights;
and for the said purposes to provide for the establishment of a Truth and Reconciliation Commission, a
Committee on Human Rights Violations, a Committee on Amnesty and a Committee on Reparation and
Rehabilitation;
and to confer certain powers on, assign certain functions to and impose certain duties upon that
Commission and those Committees;
and to provide for matters connected therewith.

‘motivations’
SINCE the Constitution of the Republic of South Africa, 1993 (Act No. 200 of 1993), provides a historic
bridge between the past of a deeply divided society characterized by strife, conflict, untold suffering and
injustice, and a future founded on the recognition of human rights, democracy and peaceful co-existence
for all South Africans, irrespective of colour, race, class, belief or sex;
AND SINCE it is deemed necessary to establish the truth in relation to past events as well as the motives
for and circumstances in which gross violations of human fights have occurred, and to make the findings
known in order to prevent a repetition of such acts in future;

copyright J R Martin & David Rose Working with Discourse

 13

AND SINCE the Constitution states that the pursuit of national unity, the well-being of all South African
citizens and peace require reconciliation between the people of South Africa and the reconstruction of
society;
AND SINCE the Constitution states that there is a need for understanding but not for vengeance, a need
for reparation but not for retaliation, a need for ubuntu but not for victimization;
AND SINCE the Constitution states that in order to advance such reconciliation and reconstruction
amnesty shall be granted in respect of acts, omissions and offences associated with political objectives
committed in the course of the conflicts of the past;
AND SINCE the Constitution provides that Parliament shall under the Constitution adopt a law which
determines a firm cut-off date, which shall be a date after 8 October 1990 and before the cut-off date
envisaged in the Constitution, and providing for the mechanisms, criteria and procedures, including
tribunals, if any, through which such amnesty shall be dealt with;

BE IT THEREFORE ENACTED by the Parliament of the Republic of South Africa, as follows:… [Office of
the President of South Africa. No. 1111. 26 July 1995]

The Chapters that follow spell out the ‘provisions’ of the act, which can themselves be
divided into ‘definitions’ and ‘actual provisions’. The overall structure of the act is thus
purpose, followed by motivation, followed by provisions, as follows. Since these may
not be generalisable stages in this genre, we have labelled them with quotes.

‘purpose’
‘motivation’
‘provisions’
definitions
actual provisions

1.3 Language, power and ideology

In our view ideology and power run through the whole ensemble of language and
culture, positioning people within each social context as having more or less power, and
opening or narrowing their access to resources for meaning. Of course, up to a point all
speakers of a language share an equal range of meaning making resources, but there
are also certain varieties of meanings that are not equally distributed. These include
resources for engaging in the written discourses of contemporary social institutions,
such as sciences, government and education. One important strand of work in SFL has
been to provide access to these discourses through literacy pedagogies grounded in
discourse analysis. Another strand has been to investigate the principles by which
access to meaning is unequally distributed, along the lines of generation, gender, class,
incapacity and ethnicity.

Until very recently, perhaps the most stark example of ideological divisions based on
ethnicity was apartheid South Africa. After a long struggle, black South Africans finally
overturned the apartheid regime, not with bullets but with words. They succeeded in
persuading governments, multinational corporations, and eventually South Africa’s white
rulers themselves that apartheid was unacceptable and could not viably continue.
Today they have embarked on a long term program to heal the deep rifts created by
generations of state-sponsored racial hatred, large-scale violations of human rights, and
impoverishment of the majority of the nation’s people. This process has been
institutionalised as the Truth and Reconciliation Commission, from which our texts are
drawn.

copyright J R Martin & David Rose Working with Discourse

 14

Stated in these terms, the victory over apartheid seems like a simple one of right over
wrong, good over evil. But of course social conflicts are rarely so simple, as all South
Africans may attest. Rather there are usually multiple facets to any contested issue,
and multiple positions by different groups. The voices of many groups are to be heard
in the texts we analyse in this book, arguing, or being argued for or against, in many
subtle ways. The discourse analyses we outline in each chapter enable these voices to
emerge clearly, explicitly, from the patterns of meaning in which they are encoded.
We’ll return to the question of language and ideology in Chapter 8.

1.4 How this book is organised

The chapters of this book are organised around five key sets of resources for making
meaning as text. These sets of resources are set out here in the order of the chapters
in which they are discussed.

Appraisal is concerned with evaluation – the kinds of attitudes that are negotiated in a
text, the strength of the feelings involved and the ways in which values are sourced and
readers aligned. Appraisals are interpersonal kinds of meanings, that realise variations
in the tenor of a text. We begin with appraisal in order to foreground the interactive
nature of discourse, including written discourse.

Ideation focuses on the content of a discourse – what kinds of activities are

undertaken, and how participants undertaking these activities are described and
classified. These are ideational kinds of meaning, that realise the field of a text.

Conjunction looks at inter-connections between activities – reformulating them, adding

to them, sequencing them, explaining them and so on. These are also ideational types
of meanings, but of the subtype ‘logical’. Logical meanings are used to form temporal,
causal and other kinds of connectivity.

Identification is concerned with tracking participants – with introducing people, places

and things into a discourse and keeping track of them once there. These are textual
resources, concerned with how discourse makes sense to the reader by keeping track
of identities.

Periodicity considers the rhythm of discourse – the layers of prediction that flag for
readers what’s to come, and the layers of consolidation that accumulate the meanings
made. These are also textual kinds of meanings, concerned with organising discourse
as pulses of information.

Following the discussion of these discourse systems, we then apply them in Chapter 7
to the analysis of one significant text - the final chapter of Nelson Mandela’s 1995
autobiography, Long Road to Freedom. And in Chapter 8 we then contextualise the
discourse systems in models of the social contexts of discourse, including register and
genre theory, and we make connections to multi-modal discourse analysis and critical
discourse analysis.

copyright J R Martin & David Rose Working with Discourse

 15

We’ll now illustrate each of the discourse systems very briefly, by way of flagging what’s
to come.

First APPRAISAL (evaluation). The focus here is on attitude – the feelings and values
that are negotiated with readers. The key resources here have to do with evaluating
things, people’s character and their feelings. Helena for example records how she and
her partner responded emotionally to his promotion.

Then he says: He and three of our friends have been promoted. ‘We’re moving to a special unit. Now,
now my darling. We are real policemen now.’ We were ecstatic. We even celebrated.

Later, she judges its horrific consequences in moral terms.

Dammit! What else can this abnormal life be than a cruel human rights violation? Spiritual murder is
more inhumane than a messy, physical murder. At least a murder victim rests. I wish I had the power to
make those poor wasted people whole again.

One important aspect of evaluation is the source of the opinions, which we would
naturally attribute to Helena in the examples given above. But we need to be careful,
since her story comes from a letter she sent to the South African Broadcasting
Corporation’s radio team who broadcast extracts. Tutu introduces her story as follows:
The South Africa Broadcasting Corporation’s radio team covering the Truth and reconciliation Commission
received a letter from a woman calling herself Helena (she wanted to remain anonymous for fear of
reprisals) who lived in the eastern province of Mpumalanga. They broadcast substantial extracts:

So in fact what we are looking at here is Tutu writing -> that the SABC broadcast -> that
Helena wrote -> that (for example) spiritual murder is more inhumane than physical
murder. Each step in the reporting nuances the evaluation, and we need to be
systematic about keeping track of the effects this has.

Next IDEATION (the content of a discourse). Here we’re concerned with people and
things, and the activities they’re involved in. Since Helena’s telling her story, there’s lots
of activity involved and it unfolds in sequences. There’s a courtship sequence for
example (woman meets man, they begin a relationship, they plan to marry):

I met a young man in his twenties… It was the beginning of a beautiful relationship. We even spoke about
marriage.

And a later sequence about consumption (people become dissatisfied, they want better,
they get it):

If I had to watch how white people became dissatisfied with the best and still wanted better and got it.

As well as sequences of activities, ideation is concerned with describing and classifying

people and things. Helena’s second love for example is classified (policeman, man,
murderer, vulture), partitioned (face, hands, eyes, throat, head, brains; personality, soul,
conscience) and variously described (bubbly, charming, bewildered, dull like the dead,
wasted etc.).

copyright J R Martin & David Rose Working with Discourse

 16

Next CONJUNCTION (inter-connections between processes). Later in the narrative

Helena comments on her understanding of the struggle against apartheid, outlining the
conditions under which she herself would have joined the struggle:

I finally understand what the struggle was really about. I would have done the same had I been denied
everything. If my life, that of my children and my parents was strangled with legislation. If I had to watch
how white people became dissatisfied with the best and still wanted better and got it.

To demonstrate her understanding she places herself in victims’ shoes, outlining the
conditions under which she would have done the same. The key resources here for
establishing conditions are conditional conjunctions If..., If…, and the Subject-verb
inversion had I… . These realisations serve to link Helena’s intended action I would
have done the same, with the conditions under which she would have done so, had I
been…If my life…If I had to watch

I would have done the same

discourse function wording (grammar)
had I been denied everything condition Subject verb inversion
If my life...was strangled with “ conjunction
legislation.
If I had to watch how white “ conjunction
people became dissatisfied...

Next IDENTIFICATION (concerned with tracking people and things). Helena’s narrative
focuses on the two loves of her life and the way their violation of human rights destroyed
their humanity. Her first love is introduced as a young man, and his identity is then kept
track of using the pronouns his and he.

As an eighteen-year-old, I met a young man in his twenties. He was working in a top security structure…

Years later Helena meets him once again, and he is reintroduced as my first love, to
distinguish him from the other men in her life.
…More than a year ago, I met my first love again through a good friend…

The key English resources here are indefinite reference (a) to introduce the young man,
pronouns to maintain his identity (his, he, my) and comparison (first) to distinguish him
from Helena’s second love.

discourse functions wording

A young man presenting a participant indefinite reference
his twenties tracking a participant pronoun
he tracking a participant pronoun
my first love comparing participants pronoun, ordinal number

Next PERIODICITY (the rhythm of discourse). Here we’re concerned with information
flow – the way in which meanings are organised so that readers can process phases of
meaning. Helena for example doesn’t launch straight into her story by telling us she

copyright J R Martin & David Rose Working with Discourse

 17

met a young man. To begin, she lets us know that she’s going to tell a story about a
teenage farm girl in Eastern Free State:

My story begins in my late teenage years as a farm girl in the Bethlehem district of Eastern Free State

And Tutu himself provided us with some more background to this story as he introduces
it:

The South Africa Broadcasting Corporation’s radio team covering the Truth and Reconciliation
Commission received a letter from a woman calling herself Helena (she wanted to remain anonymous for
fear of reprisals) who lived in the eastern province of Mpumalanga. They broadcast substantial extracts

This means that by the time Helena begins we know what to expect – which genre (a
story), and something about where and when it took place and who was involved. This
kind of predictability is absolutely critical for digesting information, and we need to look
carefully at the ways in which texts tell us what’s coming – alongside reminding us
where we’ve been. Helena for example is just as clear about where her story ends:

I end with a few lines that my wasted vulture said to me one night…

Here she lets us know that the predictions that helped us through the story are closing
down, and that a transition to something different is coming – in this case a big hop
back to Tutu’s argument. We use the term periodicity for these resources because they
organise texts as waves of information; we surf the waves, taking a look back and
forward on crests of informational prominence, so that we can glide smoothly through
the troughs on the flow of meanings we expect.

1.5 How to use this book

The aim of the book is to enable discourse analysts to use the tools we offer for text
analysis. Some applications will call for the full set of analytical tools presented in all
chapters. Others may require analyses from just one or another of the chapters.

Each chapter starts with examples from our texts to give a simple outline of the relevant
discourse system. The sections in each chapter then discuss resources in each part of
the discourse system in turn. Again in each section we start with examples of discourse
to illustrate the set of resources for meaning. The resources in each system are then
summarised in tables, for easy reference, and also where appropriate in system
diagrams to show how the whole system is organised.

Most chapters present more than one set of resources, within the overall system under
focus. This means that texts may be analysed from more than one perspective. But we
have endeavoured to set out each chapter and section as a clear set of steps that build
on those that have gone before.

The text analyses and interpretations in each chapter are intended as models for the
reader to apply to their own texts. The tables of resources are intended as references
to assist the analyst to identify other examples in their texts, and the system diagrams

copyright J R Martin & David Rose Working with Discourse

 18

help to distinguish between each type of meaning. If there is uncertainty about which
category of meaning in a table applies to a particular instance in a text, the analyst can
refer to the relevant discussion in the chapter.

In Chapter 7 we apply the tools we have built up to analyse the final chapter of
Mandela’s Long Road to Freedom. This analysis is intended to illustrate how the
discourse analysis tools can be used for a variety of purposes. We are particularly
concerned in this chapter to show the discourse strategies that Mandela uses to
evaluate his own and his country’s transition from oppression to freedom, and to share
his evaluations with readers.

Finally in Chapter 8 we outline connections between the discourse analysis tools we

have discussed, and other modes of analysis. These connections include firstly the
model of social context we introduced briefly above, and assume throughout the
following chapters. This model of register and genre is a crucial for interpreting the
roles of interpersonal, ideational and textual meanings in social discourse. Secondly
we introduce a set of tools for analysing discourse in modes other than language,
illustrated by re-interpreting Mandela’s Freedom text in terms of its visual images.
Thirdly we connect the tools for analysing discourse, register and genre to work in
critical discourse analysis, extending our toolkit to build in a model of ideology and
change in the development of texts, of speakers and of their cultures. And finally we
briefly outline connections between our own work on discourse analysis and that of
other functional linguists and SFL in general.

It needs to be emphasised that what we offer the reader here are only tools for
discourse analysis. The tasks they can be applied to are many and varied, and each
analyst will do so in their own way, depending on their needs. As with tools of any kind,
skill is needed to use them with fluency and confidence, and this can only come with
practice. What we have tried to do in this book is provide models that can guide the
reader to develop such skills. What we hope is that readers will apply them as they
choose, adapt them as they need, and develop them further.

copyright J R Martin & David Rose Working with Discourse

 19

Chapter 2 APPRAISAL - negotiating attitudes

Appraisal is concerned with evaluation – the kinds of attitudes that are negotiated in a
text, the strength of the feelings involved and the ways in which values are sourced and
readers aligned.

2.0 Negotiating attitudes

APPRAISAL is a system of interpersonal meanings. We use the resources of

APPRAISAL for negotiating our social relationships, by telling our listeners or readers
how we feel about things and people (in a word, what our attitudes are). We have
chosen to begin with this chapter to foreground the interactive nature of discourse,
whether spoken or written – as negotiation.

Attitudes have to do with evaluating things, people’s character and their feelings. Such
evaluations can be more or less intense, that is they may be more or less amplified.
And the attitude may be the writer’s own or it may be attributed to some other source.
These are the three aspects of appraisal - attitudes, how they are amplified and their
sources – that we will explore in this chapter. We will begin with a brief synopsis of
each one, and then explore them in more detail.

Let’s begin with kinds of attitudes. Helena’s story is highly evaluative as she describes
intense feelings and strong reactions to people and things. In the following passage she
outlines her attitude to her first love’s work and their relationship (things), his character
(people), and the emotions of those involved (feelings).

He was working in a top security structure. It was the beginning of a beautiful relationship. We even
spoke about marriage. A bubbly, vivacious man who beamed out wild energy. Sharply intelligent.
Even if he was an Englishman, he was popular with all the ‘Boer’ Afrikaners. And all my girlfriends envied
me. Then one day he said he was going on a ‘trip’. ‘We won’t see each other again… maybe never ever
again.’ I was torn to pieces. So was he.

She begins with the value of her love’s work and their forthcoming relationship:

a top security firm

a beautiful relationship

She then turns to her love’s character, which she holds in such high esteem:

a bubbly vivacious man

wild energy
sharply intelligent
popular

Next she describes her girlfriends’ emotional response to their relationship, and her own
and her lover’s feelings on their separation:

copyright J R Martin & David Rose Working with Discourse

 20

envied
torn to pieces

So these evaluations can be divided into three basic kinds according to what is being
appraised – (i) the value of things, (ii) people’s character and (iii) people’s feelings.

Next let’s look at how attitudes are amplified. One thing that involves us in Helena’s
story is that her evaluations are highly charged. She judges her love for example not
just as intelligent but as sharply intelligent, not just as energetic but as wildly so:

sharply intelligent
wild energy

As she wasn’t just upset at their separation, but absolutely shattered:

torn to pieces

So attitudes are gradable – their volume can be turned up and down depending on how
intensely we feel. We will refer to the resources we use to show how strong our
reactions are as amplification.

Finally let’s look at the sources of attitudes. One thing we need to consider about
attitudes is who they’re coming from. Since it’s a story, we assume that it’s the narrator,
Helena, who is evaluating, unless we’re told otherwise. Helena does in fact suggest that
the ‘Boer’ Afrikaners enjoyed her love’s company, and that it was her girlfriends who
envied her – although in each case we have to keep in mind that it’s Helena who’s
telling us how they felt:

he was popular with all the ‘Boer’ Afrikaners.

And all my girlfriends envied me.

One very common way to attribute feelings is of course to create another voice in the
story, by using direct or indirect speech. Later in the story for example she quotes from
her second love, who evaluates his life as a living hell:

I end with a few lines that my wasted vulture said to me one night: ‘… I have to live with this hell…’

And as we noted in Chapter 1, Helena’s story was itself quoted by Tutu, who is himself
quoting from a broadcast by the SABC. So the immediate, intermediate and ultimate
source of opinions in discourse is an important variable in discourse that we need to
keep track of when analysing evaluations.

In simple terms then, what we see here are a range of resources for expressing
attitudes, amplifying them and attributing them to sources. And there are three main
types of attitude – expressing emotion, judging character and valuing the worth of
things. Technically we’ll refer to resources for expressing feelings as affect, resources
for judging character as judgement and resources for valuing the worth of things as
appreciation. These basic resources are set out as Table 2.1 and then as a system
network in Figure 2.1.

copyright J R Martin & David Rose Working with Discourse

 21

Table 2.1: Basic options for appraisal

attitude affect envied
torn to pieces
judgement a bubbly vivacious man
wild energy, sharply intelligent
appreciation a top security firm
a beautiful relationship
amplification sharply intelligent
wild energy
source he was popular with all the ‘Boer’ Afrikaners.
And all my girlfriends envied me.

Figure 2.1: Basic system for appraisal

affect (feelings)…

ATTITUDE judgement (character)…

appreciation (value)…
APPRAISAL

AMPLIFICATION…

SOURCE…

In Figure 2.1, the enclosing brackets on the left mean that the enclosed options for
APPRAISAL are all selected at the same time - since when we express an attitude, we
also choose how amplified it is, and what its source is. Enclosing brackets in a system
network like this means we select an attitude and its amplification and its source. By
contrast the system shown on the right for ATTITUDE gives three alternative options.
This means that, at this stage in our outline, we can express either affect or judgement
or appreciation.

Next we will explore each of these options for APPRAISAL in more detail, beginning
with kinds of attitudes.

2.1 Kinds of attitudes

In this section we look more closely at the three kinds of attitude we have identified -
affect (people’s feelings), judgement (people’s character) and appreciation (the value of
things).

Expressing our feelings - affect

As we explore how people express their feelings in discourse, we find that they vary in
two general ways. Firstly, we can have good feelings or bad feelings – so affect can be

copyright J R Martin & David Rose Working with Discourse

 22

positive or negative. Secondly people can express their feelings directly, or we can
infer how people are feeling indirectly from their behaviour - so affect can be expressed
directly or implied.

First let’s look at positive and negative affect. More perhaps than any other family of
genres, stories involve us in people’s feelings. We empathise and sympathise with
characters as they take part in extraordinary events. In her exemplum, Helena
describes her own emotions as follows:

I was torn to pieces.

I can’t explain the pain and bitterness in me…
We were ecstatic.
We even celebrated.
Abruptly mutter the feared word ‘trip’
…knew no other life than that of worry, sleeplessness, anxiety about his safety…
We simply had to be satisfied with:
Today I know the answer to all my questions and heartache.
I envy and respect the people of the struggle
I wish I had the power
those poor wasted people
I wish I could wipe…

And her second love’s emotional states are even more fully elaborated:
Humorous, grumpy, everything in its time and place.
We were ecstatic.
We even celebrated.
they would become restless.
Abruptly mutter the feared word ‘trip’
as a loved one…
And all that we as loved ones knew
Withdrawn.
Sometimes he would just press his face into his hands and shake uncontrollably
He tried to hide his wild consuming fear…
I jolt awake from his rushed breathing.
Eyes bewildered, but dull like the dead.
And the shakes.
The terrible convulsions
and blood-curdling shrieks of fear and pain from the bottom of his soul.

It’s not a happy story, as we can see. Most of the feelings are negative ones – things
we’d rather not feel:

torn to pieces, pain, bitterness, feared, worry, anxiety, heartache, envy;

grumpy, restless, feared, withdrawn, press his face into his hands, shake uncontrollably, wild consuming
fear, rushed breathing, eyes bewildered, shakes, terrible convulsions, blood-curdling shrieks of fear and
pain

Positive vibrations are few and far between:

ecstatic, celebrated, satisfied, respect, wish;
humorous, loved

copyright J R Martin & David Rose Working with Discourse

 23

This contrast between good and bad vibes is a basic one as far as emotions, and
attitudes in general are concerned.

Next we’ll look at direct and implicit expressions of feelings. From the lists above we
can see Helena describes emotions in different ways. Often she refers directly to a
mental state, using words that name specific emotions:

torn to pieces, pain, bitterness, ecstatic, feared, worry, anxiety, satisfied, heartache, envy, respect, wish,
humorous, grumpy, ecstatic, loved, wild consuming fear, bewildered, blood-curdling, fear, pain

Alongside this, she refers at times to emotion physically, describing behaviour that also
directly expresses emotion – uncontrollable shakes to express anxiety for example, or
shrieks to express fear:

celebrated, restless, withdrawn, press his face into his hands, shake uncontrollably, rushed breathing, the
shakes, terrible convulsions, shrieks

Related to this, and sometimes hard to distinguish from it, is her description of unusual
behaviour which we read as an indirect sign of emotion:

very quiet; drinking too much; wander from window to window; rolls this way, that side of the bed; pale; ice
cold in a sweltering night, sopping wet with sweat; sits motionless, just staring in front of him

Taken out of context, from this unusual behaviour we know something is wrong but we
can’t be quite so sure about the exact emotion being expressed; we need to use a bit of
psychology perhaps. Read in context however, we do know what Helena’s on about -
because these symptoms are surrounded by explicit references to emotions which tell
us what the strange behaviour means. We can see the interplay in the text below,
where physical symptoms are underlined and direct affect is in bold.

He became very quiet. Withdrawn. Sometimes he would just press his face into his hands and shake
uncontrollably. I realised he was drinking too much. Instead of resting at night, he would wander from
window to window. He tried to hide his wild consuming fear, but I saw it. In the early hours of the
morning between two and half-past-two, I jolt awake from his rushed breathing. Rolls this way, that side
of the bed. He’s pale. Ice cold in a sweltering night – sopping wet with sweat. Eyes bewildered, but dull
like the dead. And the shakes. The terrible convulsions and blood-curdling shrieks of fear and pain
from the bottom of his soul. Sometimes he sits motionless, just staring in front of him.

We can also note here the role that metaphor plays in constructing emotion. Helena’s
love’s eyes are dull like the dead, he’s cold as ice, his fear and pain come from the
bottom of his soul:

ice cold in a sweltering night;

eyes bewildered, but dull like the dead;
terrible convulsions and blood-curdling shrieks from the bottom of his soul

As we can see, Helena uses a range of resources to build up a picture of her second
love’s living hell – including direct expressions of emotional states and physical
behaviour, and implicit expressions of emotion through extraordinary behaviour and
metaphor.

copyright J R Martin & David Rose Working with Discourse

 24

In Helena’s story these resources work together, reinforcing for example the
desperation of her second love’s emotional devastation – his spiritual murder as she
describes it. This accumulative effect over a phase of text reflects the ‘prosodic’ nature
of attitude, and of interpersonal meaning in general. Interpersonal meanings are often
realised not just locally, but tend to sprawl out and colour a passage of discourse,
forming a ‘prosody’ of attitude. By looking at phases of attitude, we can explore how
readers are being aligned rhetorically as a text unfolds; we’ll return to this issue of
aligning the reader below.

Summing up then, we’ve seen that affect can be positive or negative, and that they can
be realised directly or implicitly in text. And we’ve seen that direct and implicit
realisations often work together to establish the mood of phases of discourse. These
options for affect are set out in Table 2.2.

Table 2.2: Options for affect

positive We were ecstatic.
We even celebrated.
negative I was torn to pieces.
I can’t explain the pain and bitterness in me…
direct emotional state ecstatic
wild consuming fear
physical Withdrawn
expression shake uncontrollably
implicit extraordinary wander from window to window
behaviour rolls this way, that side of the bed
metaphor ice cold in a sweltering night
eyes…dull like the dead

Judging people’s character

As with affect, judgements of people’s character can be positive or negative, and they
may be judged explicitly or implicitly. But unlike affect, we find that judgements differ
between personal judgements of admiration or criticism and moral judgements of
praise or condemnation.

We’ll start with personal judgements - positive (admiring) and negative (criticising). As
we showed in Chapter 1, Helena’s story is an exemplum. Exemplums relate an incident
in order to comment on the behaviour of the people involved. This means that
alongside telling how people feel emotionally, Helena judges them – she evaluates their
character.

Helena’s first love is at first characterised admiringly as bubbly, vivacious, energetic,

intelligent, popular and later, retrospectively, as beautiful, big and strong. And he is also
admired implicitly as working in a top security structure, i.e. an admirable role. Helena’s
second love is not quite so special, but described initially as exceptional, special, bubbly
and charming. In both cases her lovers change, as a result of their security operations.
Helena doesn’t explicitly re-assess her first love, rather she implies criticism by telling us
how she felt when she saw what was left of him:

copyright J R Martin & David Rose Working with Discourse

 25

I can’t explain the pain and bitterness in me when I saw what was left of that beautiful, big, strong
person.

But she does directly criticise her second love as having something wrong with him, as
maybe having gone mad, and as wasted. Their transformations from admiring
judgements to critical ones are central to the impact of the two Incident stages of the
story.

Next moral judgements - positive (praising) or negative (condemning). In the

Interpretation stage of her story Helena judges South African leaders. She condemns
her own leaders for their dishonesty:

Our leaders are too holy and innocent. And faceless. I can understand if Mr (F.W.) de Klerk says he
didn’t know, but dammit, there must be a clique, there must have been someone out there who is still alive
and who can give a face to ‘the orders from above’ for all the operations.

And for their inhumanity:

Dammit! What else can this abnormal life be than a cruel human rights violation? Spiritual murder is
more inhumane than a messy, physical murder. At least a murder victim rests.

But she praises the leaders of ‘the people of the struggle’ for having the courage to
stand by their resistance forces and honour their activities:

… at least their leaders have the guts to stand by their vultures, to recognise their sacrifices.

This shift in gears from Incident to Interpretation stages is significant. In the Incidents,
Helena is not blaming anyone. At first she’s full of admiration for her lovers, then
worried sick about their problems. In the Interpretations however she both condemns
and praises on moral grounds, as she deals with honesty and dishonesty, and with guilt
and innocence in the face of the Truth Commission. It’s this shift to moral values in the
later stages of the exemplum that drives home the point of the story and attracts Tutu to
it by way of exemplifying one of his arguments about the cost of justice.

As with admiration and criticism, moral judgements can also be made directly or implied.
When Helena says I envy and respect the people of the struggle for example, she’s
telling us how she’s feeling emotionally; but both emotions imply something
praiseworthy about the character of the people. Similarly, she morally condemns those
at the top for bloody murder, without explicitly judging their character:

…while ‘those at the top’ were again targeting the next ‘permanent removal from society’ for the vultures.

As with emotion, explicit judgements in the relevant phase of discourse tell us exactly
how she wants us to judge the people targeted by accusations of this kind (i.e. as
murderers):
And today they all wash their hands in innocence and resist the realities of the Truth Commission. Yes, I
stand by my murderer who let me and the old White South Africa sleep peacefully. Warmly, while ‘those
at the top’ were again targeting the next ‘permanent removal from society’ for the vultures.

copyright J R Martin & David Rose Working with Discourse

 26

Metaphor also plays a role in judging character, as leaders wash the blood off their
hands, operatives gnaw at carcasses and African families have the life choked out of
them with legislation:

And today they all wash their hands in innocence;

… ‘our men’ who simply had to carry out their bloody orders… like ‘vultures’;
If my life, that of my children and my parents was strangled with legislation.

Perhaps the most powerful image in Helena’s exemplum is that of ‘spiritual murder’ – by
way of capturing the immorality of ‘those at the top’ in relation to their bloody orders.

We can sum up the options for judgement that we’ve seen so far in Helena’s story as
Table 2.3.

Table 2.3: Examples of judgement of character (from Helena’s story)

direct implied
personal admire bubbly, vivacious, energetic, He was working in a top security
intelligent, popular structure.
criticise What’s wrong with him? …I can’t I can’t explain the pain and
handle the man anymore! bitterness in me when I saw…
moral praise their leaders have the guts to I envy and respect the people of the
stand by their vultures… struggle...
condemn Our leaders are too holy and …’those at the top’ were again
innocent. And faceless. targeting the next ‘permanent
removal from society’ …

Let’s now turn to Desmond Tutu’s exposition. Tutu is involved in a moral argument – an
exposition dealing with the integrity of the Truth and reconciliation process.

the cost of justice

the integrity of the entire Truth and Reconciliation process

So we would naturally expect judgements of character to be foregrounded there. In fact

we do find some emotion (intimidated, humiliation, traumatic, fear, respect, wishes), but
this is indeed overwhelmed by the concern with moral issues.

Some of his judgements are like Helena’s – everyday evaluations of character involving
respectability, responsibility, accountability and veracity:

respectable members of their communities

who accept responsibility
encourages accountability
It is also not true

But many more of his judgements are judicial – they work as a kind of ‘technicalised
morality’ that we associate with legal institutions. Note for example that he offers a
definition of a gross violation of human rights, taken from the Promotion of National

copyright J R Martin & David Rose Working with Discourse

 27

Unity and Reconciliation Act. Definitions are a sure sign that we are moving from
common sense into uncommon sense knowledge:
a gross violation of human rights – defined as an abduction, killing, torture or severe ill-treatment

Here are some more examples of Tutu’s judgmental legalese:

had committed a crime,
sees the offence

the perpetrator
regular torturers of detainees
the victim

who plead guilty

innocent people
those who claim to be innocent

of reprisals
in retaliation
not retribution or punishment

encourages impunity
the granting of amnesty

a miscarriage of justice

For certain analytical purposes we might argue that these technical judgements should
be left out of an appraisal analysis, since each in a sense refers to an ideational
meaning that is precisely situated within legal institutions, rather than an interpersonal
meaning like appraisal. But we’re not sure their technicality totally robs them of their
evaluative role – most seem to us to carry with them some of their everyday attitudinal
power, certainly for lay readers. By way of another example, when Robert Manne
wrote that Australia’s policy of removing Aboriginal children from their families by force
was “technically an act of genocide”, we doubt that for most Australians its
technicalisation completely softened the moral blow:

A national inquiry last year found that the government policy of forced removal was a gross violation of
human rights and technically an act of genocide because it has the intention of destroying Australia’s
indigenous culture by forced assimilation. [Manne, R 1998 The stolen generations. Quadrant No. 343.
Volume XLII. Number 1-2. 53-63]

Finally we can examine the Act of parliament for its judgements. Like Tutu’s exposition,
the Act foregrounds judgement over affect - and its judgements are mainly technical
ones, as we’d expect from a legislative document. Some examples are highlighted
below.
To provide for the investigation and the establishment of as complete a
picture as possible of the nature, causes and extent of gross violations of human
Rights committed during the period from 1 March 1960 to the cut-off date
contemplated in the Constitution, within or outside the Republic, emanating from the
conflicts of the past, and the fate or whereabouts of the victims of such
violations;

copyright J R Martin & David Rose Working with Discourse

 28

the granting of amnesty to persons who make full disclosure of

all the relevant facts relating to acts associated with a political objective
committed in the course of the conflicts of the past during the said period;

affording victims an opportunity to relate the violations they suffered;

the taking of measures aimed at the granting of reparation to, and the
rehabilitation and the restoration of the human and civil dignity of, victims
of violations of human rights;

reporting to the Nation about such violations and victims;

the making of recommendations aimed at the prevention of the commission of gross violations of human
rights;

and for the said purposes to provide for the establishment of a Truth and Reconciliation Commission, a
Committee on Human Rights Violations, a Committee on Amnesty and a Committee on Reparation and
Rehabilitation;

and to confer certain powers on, assign certain functions to and impose certain duties upon that
Commission and those Committees;

and to provide for matters connected therewith.

Appreciating things

To this point we’ve looked at how people feel about people and the way the behave.
What about things? Appreciation of things includes our attitudes about TV shows,
films, books, CDs; about paintings, sculptures, homes, public buildings, parks; about
plays, recitals, parades or spectacles and performances of any kind; feelings about
nature for that matter – panoramas and glens, sunrises and sunsets, constellations,
shooting stars and satellites on a starry night. As with affect and judgement, things can
be appreciated positively or negatively.

Helena’s narrative is more about people than things, and so foregrounds affect and
judgement. But it does include evaluations of relationships:

a beautiful relationship
an extremely short marriage… failed
my unsuccessful marriage

And of quality of life as well:

hell
hell
this abnormal life
hell
hell

Relationships and qualities of life are abstract sorts of things, but can be evaluated as
things nevertheless. Tutu’s exposition, as we have seen, foregrounds judgement. But

copyright J R Martin & David Rose Working with Discourse

 29

early on it does evaluate semiotic things, including a question, an issue and

applications:

a frivolous question
a very serious issue
virtually all the important applications to the Commission

And towards the end, Tutu does focus on relationships, by way of exploring the meaning
of restorative justice:

I contend that there is another kind of justice, restorative justice, which is characteristic of traditional
African jurisprudence. Here the central concern is not retribution or punishment but, in the spirit of
ubuntu, the healing of breaches, the redressing of imbalances, the restoration of broken
relationships. This kind of justice seeks to rehabilitate both the victim and the perpetrator, who should
be given the opportunity to be reintegrated into the community he or she has injured by his or her
offence. This is a far more personal approach, which sees the offence as something that has happened to
people and whose consequence is a rupture in relationships. Thus we would claim that justice,
restorative justice, is being served when efforts are being made to work for healing, for forgiveness and
for reconciliation.

Tutu’s use of the term justice in the context of ubuntu seems at first blush to indicate
that he’s judging behaviour here. But in fact he is more concerned with restoring the
fabric of social relations than with western notions of retribution and punishment. On
the positive evaluation side we have terms concerned with communal healing:

the healing of breaches

the redressing of imbalances
the restoration of broken relationships
rehabilitate both the victim and the perpetrator
the opportunity to be reintegrated into the community
restorative justice
healing
reconciliation.

On the negative we have terms concerned with damage done:

broken relationships
the community he or she has injured by his or her offence
a rupture in relationships

We can summarise the positive and negative appreciations we’ve examined so far in
Table 2.4.

Table 2.4: Examples of appreciation

positive a beautiful relationship
a very serious issue
healing of breaches
redressing of imbalances
restoration of broken relationships
negative my unsuccessful marriage
a frivolous question
broken relationships

copyright J R Martin & David Rose Working with Discourse

 30

the community he or she has injured

In order to illustrate a prosody of positive appreciations, we’ll switch fields for a moment
and consider a review of the current CD edition of Stevie Ray Vaughan’s album Texas
Flood (courtesy of Amazon.com):
This legendary 1983 debut by the fallen torchbearer of the ‘80s-’90s blues revival sounds even more
dramatic in its remixed and expanded edition. Stevie Ray Vaughan’s guitar and vocals are a bit brighter
and more present on this 14-track CD. And the newly included bonus numbers (an incendiary studio
version of the slow blues “Tin Pan Alley” that was left off the original release, and live takes of “Testify,”
“Mary Had a Little Lamb,” and the instrumental “Wham!” from a 1983 Hollywood concert) illuminate the
raw soul and passion that propelled his artistry even when he was under the spell of drug addiction.
Texas Flood captures Vaughan as rockin’ blues purist, paying tribute in his inspired six-string diction to
his influences Larry Davis (who wrote the title track), Buddy Guy, Albert King, and Jimi Hendrix. His own
contemplative “Lenny,” a tribute to his wife at the time, also suggests a jazz-fueled complexity that
would infuse his later work. --Ted Drozdowski

This is a rave review from an in-house editor, designed to persuade Amazon’s

customers to purchase Vaughan’s debut album. The album in general and certain
tracks in particular are described in very positive terms:
legendary, even more dramatic, bit brighter, more present, incendiary, contemplative, jazz-fueled
complexity

To these appraisals we might add some arguably experiential meanings with a positive
value in the context of this new edition of Texas Flood:
remixed, expanded, bonus

The borderline of character and value

There are several instances of attitude in our texts that could perhaps be analysed as
either judgement of character or appreciation of things. For example, closely related to
the positive appreciation of Vaughan’s album and its tracks are the evaluations of his
performance:
raw soul and passion, artistry, inspired six-string diction

These bring us to the border of character and value (of judgement and appreciation).
Because they directly value Vaughan’s guitar playing rather than the man, we’ll take
them here as concerned with value rather than character. But they can also additionally
be coded as tokens of Vaughan’s enormous guitar playing abilities – as betokening one
positive dimension of his character (as opposed to the negative dimension of drug
addiction, also noted in the review). Even more borderline perhaps are the
generalisations of these positive capacities when Vaughan is referred to as a
torchbearer of the ‘80s-’90s blues revival and a rockin’ blues purist.
torchbearer, rockin’ blues purist

copyright J R Martin & David Rose Working with Discourse

 31

In the prosodic domain of this positive appreciation of the CD, these can arguably be
included as positive appreciations; but just as strong a case might be made for reading
items such as these as positive judgements of Vaughan’s capacity as an artist,
especially in contexts where character rather than performance is being evaluated. The
context sensitivity of these borderline items underlines the importance of analysing
appraisal in prosodic terms. So it is important to take co-text into account, rather than
analysing simply item by item.

The key term for Tutu, judging from the title of his book, is forgiveness – which seems in
this context to comprise aspects of both judgement and appreciation. Judgement in the
sense that someone is generous enough to stop feeling angry and wanting to punish
someone who has done something wrong to them; appreciation in the sense that peace
is restored. It also seems that for Tutu, forgiveness involves a spiritual dimension,
underpinned by his Christianity; the concept transcends ethical considerations towards
a plane of peace and spiritual harmony. In appraisal terms what this means is that the
politicised aesthetics of appreciation has recontextualised the moral passion plays of
judgement.

If we take communal healing as one dimension of value analysis, then the Act can also
be seen to be concerned with repairing social relations.

SINCE the Constitution of the Republic of South Africa, 1993 (Act No. 200 of 1993), provides a historic
bridge between the past of a deeply divided society characterised by strife, conflict, untold suffering
and injustice, and a future founded on the recognition of human rights, democracy and peaceful co-
existence for all South Africans, irrespective of colour, race, class, belief or sex;
AND SINCE it is deemed necessary to establish the truth in relation to past events as well as the
motives for and circumstances in which gross violations of human rights have occurred, and to make the
findings known in order to prevent a repetition of such acts in future;
AND SINCE the Constitution states that the pursuit of national unity, the well-being of all South African
citizens and peace require reconciliation between the people of South Africa and the reconstruction of
society;
AND SINCE the Constitution states that there is a need for understanding but not for vengeance, a
need for reparation but not for retaliation, a need for ubuntu but not for victimisation;
AND SINCE the Constitution states that in order to advance such reconciliation and reconstruction
amnesty shall be granted in respect of acts, omissions and offences associated with political objectives
committed in the course of the conflicts of the past;
AND SINCE the Constitution provides that Parliament shall under the Constitution adopt a law which
determines a firm cut-off date, which shall be a date after 8 October 1990 and before the cut-off date
envisaged in the Constitution, and providing for the mechanisms, criteria and procedures, including
tribunals, if any, through which such amnesty shall be dealt with;

For this analysis we’ve concentrated on items that don’t directly involve judgement. But
the following paragraph gives us pause:

AND SINCE the Constitution states that there is a need for understanding but not for vengeance, a
need for reparation but not for retaliation, a need for ubuntu but not for victimisation;

Here the Act systematically opposes what we treated as appreciation above to terms
which more explicitly involve ethical considerations, i.e. judgements about impropriety of
people’s behaviour:

copyright J R Martin & David Rose Working with Discourse

 32

appreciation (healing) judgement (impropriety)

understanding vengeance
reparation retaliation
ubuntu victimisation

Afro-Christian values are constructed as transcending western justice. Perhaps a better

reading of the drift of feeling in the Act would be one that follows Tutu’s comments on
the meaning of ubuntu:

the spirit of ubuntu, the healing of breaches, the redressing of imbalances, the restoration of broken
relationships

Here order subsumes disorder; peace breaks out. These are the values the Act wants
people to align with in the new rainbow republic. Accordingly it might be wise to group
judgement and appreciation together here, under the headings of order and disorder –
by way of displaying the attitude to reconciliation the Act is designed to enact.

order
democracy, peaceful co-existence, national unity, peace, reconciliation, reconstruction of society,
understanding, reparation, ubuntu, reconciliation, reconstruction;
recognition of human rights, truth, well-being, amnesty, amnesty

disorder
deeply divided society, strife, conflict, conflicts;
injustice, violations of human rights, vengeance, retaliation, victimisation, omissions, offences

It might be even wiser to pause for a moment and consider the extent to which our
affect, judgement and appreciation framework represents a western construction of
feeling. Tutu’s Afro-Christian heritage might not factor attitude along these lines. We’re
not wise enough to gaze beyond our categories here. But we are confident that other
cultures will take pause, and look at what we’ve done through different eyes.

2.2 Amplifying attitudes

One distinctive feature of attitudes is that they are gradable – this means that we can
say how strongly we feel about someone or something. For example, Helena describes
her first love as sharply intelligent. By doing so she places his intelligence on a scale
and ranks it highly in relation to other choices she could have made:

extremely intelligent high grading

sharply intelligent
really intelligent
quite intelligent
fairly intelligent
somewhat intelligent low grading

copyright J R Martin & David Rose Working with Discourse

 33

As we can see, some choices turn the volume up (e.g. extremely, sharply) and others
tone it down (e.g. fairly, somewhat). In English we seem to have more resources for
turning the volume up than down, and use them more often.

In this section we will look at two kinds of resources for amplification. The first is for
‘turning the volume up or down’. These include words that intensify meanings, such as
very/really/extremely, and vocabulary items that include degrees of intensity, such as
happy/delighted/ecstatic. We refer to this kind of amplifying as force. The second kind
involves ‘sharpening’ or ‘softening’ categories of people and things, using words such as
about/exactly or real/sort of/kind of. We refer to this kind of amplifying as focus.

Amplifying the force of attitudes

We can begin with words that amplify the force of attitudes, such as
very/really/extremely. These kinds of words are known as intensifiers. Helena for
example intensifies how special her second love was, and how quiet he became, as well
as how long her unsuccessful marriage lasted.

very special.
very quiet.
An extremely short marriage to someone else

Tutu also uses intensifying of this kind.

a very serious issue

quite a price to pay.

Intensifiers make it possible for us to compare things – to say how strongly we feel
about someone or something, by comparison to something else. Helena for example
describes how white people had the best of everything and still wanted more.

If I had to watch how white people became dissatisfied with the best and still wanted better and got it.

The best is implicitly compared with the worst, which is all the ‘people of the struggle’
had. And best is also compared with better, which is what white people wanted.
These comparisons are possible because the worth of things is gradable.

best/better/good/bad/worse/worst

Comparison is also found in Tutu’s exposition and the Act.

a far more personal approach

as complete a picture as possible

These intensifiers belong to scales such as the following.

slightly more/a little more/a lot more/far more
less than/as much as/more than

copyright J R Martin & David Rose Working with Discourse

 34

Some comparison refers to an excess of feeling, as when Helena criticises the lack of
responsibility taken by the leaders of white South Africa and Tutu notes the problem of
intimidation for some witnesses.

too holy and innocent

too intimidated to testify in open session

Too contrasts with enough in this region of meaning.

not enough/enough/too much

We won’t go into more detail about resources for intensifying feelings here. There is a
useful discussion of ‘amplifiers’, ‘downtoners’ and ‘emphasizers’ in Quirk et al. 1985,
and see also Hyland 1998 on ‘hedging’. And there is a very useful outline of grading
adverbs in Collins Cobuild’s Grammar Patterns 2: nouns and adjectives. Cobuild shows
in particular that many intensifiers themselves involve attitude.

amazingly beautiful
unusually beautiful
dangerously beautiful
breathtakingly beautiful

There are also several other areas of meaning that involve grading – for example
quantity, manner and modality.

quantity all/several/some of my questions

manner degree shake frantically/uncontrollably/excitedly
modality there must/would/might have been someone out there…

A complete analysis of amplification would usefully include these kinds of meanings.

We’ll discuss modality further below under Sources, but here we’ll restrict ourselves to
graded feelings.

Next let’s examine vocabulary items that include degrees of intensity, such as
happy/delighted/ecstatic. These kinds of words are known as attitudinal lexis – ie.
‘lexis with attitude’. The intensifiers we have already looked at, like better/best,
all/several/some, must/would/might, are grammatical items. That is their meaning
depends on being combined with ‘content words’. By contrast, ‘content words’ are
referred to technically as lexical items, or simply lexis.

Attitudinal lexis plays a very important role in Helena’s narrative, as it does in general
across story genres. Helena for example says that she and her second love were
ecstatic abut his promotion, as opposed to say happy, chuffed, delighted or elated.
These are all lexical items that refer to degrees of happiness. It’s not always easy to
arrange groups of words like these confidently along a scale, but there are obviously
various degrees of feeling involved. With these items, amplification is fused into the
words themselves, so that in the dictionary chuffed is defined as ‘very pleased’ - with the
amplification factored out as very.

copyright J R Martin & David Rose Working with Discourse

 35

Here are some more examples of attitudinal lexis from Helena’s Incidents, with some
suggested scales of intensity.

vivacious man dull/placid/lively/vivacious

torn to pieces saddened/grief stricken/torn to pieces
ecstatic happy/chuffed/delighted/elated/ecstatic
bewildered bemused/puzzled/confused/bewildered
blood-curdling shrieks of fear whimper/groan/cry/screech/shriek
pleading ask/request/pray/beseech/plead

With lexical resources like these, the line between categories can be hard to draw – and
it is not always clear just how many items to include as pushing up the volume in
analysis. As a rule of thumb, the words recognised will be ‘non-core vocabulary’ (Carter
1987), i.e. lexical items other than those most commonly used in English, and they will
tend to be defined in dictionaries with intensifiers like very.

Beyond this, we can also be guided by the prosody of feeling that colours a whole
phase of discourse. In Helena’s narrative for example, attitudinal lexis is more a feature
of her Incidents than her Orientation or Interpretations. And genre is also a factor. Tutu
uses less of this resource in his exposition, but there are some examples.

a frivolous question
the full glare of television lights
humiliation for the perpetrator
impunity

On the other hand, the Act arguably uses no attitudinal lexis at all, just as it avoids
intensifiers like very. So we can score various genres on how much amplification they
are likely to display: narratives tend to amplify most, expositions less so, and
administrative genres like the Act amplify very little.

Another feature of certain genres is that grading is erased when we technicalise

attitude. For example, in commonsense terms gross is at the extreme of scales such as
minor/unacceptable/gross or unpleasant/disturbing/gross. But once we define a gross
violation of human rights then gross doesn’t scale how unacceptable or unpleasant the
violation is any more – gross simply becomes part of the name of the offence -
classifying the type of offence, rather than intensifying it.

a gross violation of human rights – defined as an abduction, killing, torture or severe ill-treatment

As well as the lexical items we’ve seen above, attitudinal lexis also includes metaphors
and swearing. We’ve already considered Helena’s metaphors in relation to affect, but
we can note here that they also have an amplifying effect.

ice cold in a sweltering night

dull like the dead
blood-curdling shrieks

copyright J R Martin & David Rose Working with Discourse

 36

These metaphors tell us how cold her second love was, how dull his eyes were, and
how frightening his screams were.

As well as metaphors, Helena also uses swearing in her Interpretation to express her
frustration with white South African leaders.

Our leaders are too holy and innocent. And faceless. I can understand if Mr (F.W.) de Klerk says he didn’t
know, but dammit, there must be a clique, there must have been someone out there who is still alive and
who can give a face to ‘the orders from above’ for all the operations. Dammit! What else can this
abnormal life be than a cruel human rights violation?

Perhaps what we are looking at here is feeling which becomes so amplified it explodes
– a kind of short-circuit which disengages amplification from what is being appraised
(the leaders’ character) and ‘cuts loose’ as a swear word. The role of swearing needs
further exploration, including its relationship with ‘interjections’ (Quirk et al. 1985), such
as noises like ugh, phew, gr-r-r-r, ow, whew, tut-tut etc. Eggins & Slade 1997 also have
some relevant discussion.

Sharpening and softening focus

Now let’s look briefly at focus - the sharpening and softening of experiential categories.
What we’ve looked at so far are resources for adjusting the volume of gradable items.
By contrast, focus is about resources for making something that is inherently non-
gradable gradable. For example, Helena introduces her second love as a policeman.

After my unsuccessful marriage, I met another policeman.

Experientially, this sets him up as having one kind of job rather than another (tinker,
tailor, soldier, spy etc.). Classifications of this kind are categorical distinctions – he was
a policeman as opposed to something else. After his promotion, however, her second
love describes himself as a real policeman, as if he hadn’t quite been one before.

We are real policemen now.

This in effect turns a categorical boundary between types of professions into a graded
one, allowing for various degrees of ‘policeman-hood’. It implies that when Helena met
him he was less of a policeman than after his promotion.

I met a kind of policeman

I met a policeman sort of

Grading resources of this kind don’t so much turn the volume up and down as sharpen
and soften the boundaries between things. Real policeman sharpens the focus, a sort
of policeman softens it. As well as things, we can also sharpen or soften types of
qualities, such as deep blue or bluish. Even categorical concepts like numbers can be
pushed around in this way.

copyright J R Martin & David Rose Working with Discourse

 37

After about three years with the special forces

vs
After exactly three years with the special forces

Here’s another example of sharpened focus from Helena’s story.

was what we saw with our own eyes

Here own sharpens the category ‘our eyes’, ie. ‘ours and no-one else’s’ – it’s definitely
not hearsay. And here’s an example of softened focus.

not quite my first love

Tutu also sharpens focus a couple of times in his exposition, in order to be precise.

the very first time

precisely this point

However, as with force, the Act appears to avoid focus entirely, preferring categorical
distinctions as a matter of legislation.

We won’t pursue the full range of these resources for grading experiential boundaries in
detail. For a very useful discussion of vague language of this kind see Channel 1994.
Here, however, are a few more examples from Stevie Ray Vaughan’s admiring fans.

to what real blues sounds like

then you aren’t a fan of true, god-blessed American music
two of the songs sound exactly alike
Whether you’re a hardcore Stevie fan
Among Stevie was his brother, Dr. John, Angela Strehler.(or something)
Here is an authentic blues artist
This video was the epitome of Stevie
It’s pure perfection
Absolute intensity
If you’re even a part-time blues fan

To sum up then, amplifying attitude involves a set of resources for adjusting how
strongly we feel about people and things. Technically we refer to these resources as
force. We use them to turn the volume up or down. Grading experiential boundaries
involves resources that sharpen or blur apparently categorical distinctions. Technically
these resources are referred to as focus. They make cut and dried distinctions
negotiable. These options for amplification are set out in Table 2.5 and as a system in
Figure 2.3. Technically these resources are referred to as graduation.

Table 2.5: Options for graduation

force intensifiers he still plays great
attitudinal lexis the second part is fantastic…
metaphors ice cold in a sweltering night
swearing dammit, there must be a clique
focus sharpen a true guitar legend
soften a part-time blues fan

copyright J R Martin & David Rose Working with Discourse

 38

Figure 2.3: Options for graduation

intensifiers

attitudinal lexis

metaphor
FORCE
swearing

raise
GRADUATION
lower

sharpen
FOCUS
soften

2.3 Sources of attitudes

The final region of appraisal we need to consider has to do with the source of attitudes –
who are the evaluations coming from?

We’ll begin once again with Helena’s narrative. On the face of it, we might argue that
the evaluation in Helena’s story comes from Helena. She’s the narrator after all. So
when she describes her relationship with her first love as beautiful, that’s her opinion:
It was the beginning of a beautiful relationship.

And in a sense Helena is responsible for all of the evaluation – since all of it is filtered
through her narration. Helena does however explicitly give voice to other players by
quoting or reporting what her first and second loves said.

Then he says: He and three of our friends have been promoted. ‘We’re moving to a special unit. Now,
now my darling. We are real policemen now.’

This potential for sourcing what is said was one of the factors that got the Russian
linguist Bakhtin thinking about the dialogic nature of discourse – even in texts we
traditionally think of as monologues. The French discourse analyst Kristeva introduced
the term heteroglossia (‘different voices’) for this notion of multiple voicing in all kinds
of discourse. Here we will use the term heterogloss where the source of an attitude is
other than the writer, and monogloss (‘single voice’) where the source is simply the
author.

Projecting sources

One thing we are able to do in discourse is quote or report what people say or think.
Halliday (1994) calls this type of linguistic resource ‘projection’. Projection is the

copyright J R Martin & David Rose Working with Discourse

 39

relation between he says in example above, and what he said He and three of our
friends have been promoted. ‘We’re moving to a special unit. Now, now my darling. We
are real policemen now.’ We can illustrate projection with a ‘speech bubble’ in Figure
2.4.

Figure 2.4: Projection

He and three of our friends have been

promoted. ‘We’re moving to a special unit.
Now, now my darling. We are real policemen
now.’
he says

Projections may quote the exact words that someone said, in which case ‘speech
marks’ are usually used in writing .

‘We’re moving to a special unit. Now, now my darling. We are real policemen now.’

Or they may report the general meaning that was said, which normally doesn’t require
speech marks.

He and three of our friends have been promoted.

As well as ‘saying’, it is also possible to quote or report what we think or feel.

I realised he was drinking too much.

I know where everything began, the background.
I wish I could wipe the old South Africa out of everyone’s past.

Through projection then we can introduce additional sources of evaluation. And

because we can choose from projection over and over again – ‘recursively’, we can use
it to explore the source of sources, and the source of the source of sources – as Helena
does when talking about what white South African leaders did and didn’t know:
I can understand if Mr (F.W.) de Klerk says he didn’t know, but dammit, there must be a clique, there must
have been someone out there who is still alive and who can give a face to ‘the orders from above’ for all
the operations.

In this sentence, Helena chooses projection three times. Two projections are ‘thinking’
– understand and know, and one is ‘saying’ – says. We can present these recursive
sources as a diagram, in Figure 2.5, in which thought bubbles and speech bubbles
represent what is thought or said.

copyright J R Martin & David Rose Working with Discourse

 40

Figure 2.5: Recursive sources

(who gave) ‘the orders
from above’ for all the
operations

he didn’t know

if Mr F.W. de Klerk
says

I can understand

In Helena’s narrative, projection doesn’t just happen within sentences, from ‘saying’ to
‘what is said’. It can also happen across whole texts and text phases. For example
Helena begins by presenting herself as narrator (my story begins):

My story begins in my late teenage years as a farm girl in the Bethlehem district of Eastern Free State.

The rest of her story then is what she tells. And she closes her story by handing over to
her second love (a few lines…):

I end with a few lines that my wasted vulture said to me one night:

In both cases Helena’s sentence ‘projects’ the sentences that follow, just as the SABC
‘projected’ Helena’s story:

they broadcast substantial extracts:

And Tutu in turn projects the SABC broadcast:

The South Africa Broadcasting Corporation’s radio team covering the Truth and Reconciliation
Commission received a letter from a woman calling herself Helena…

So ultimately we have Tutu saying that the SABC said that Helena said that her second
love said what he said. This is managed between sentences by naming ‘speech acts’ ,
such as my story, a few lines , a letter, substantial extracts. This kind of projection
between sentences is often associated with the beginning and end of texts.

Projections can also be found within clauses, where they explicitly assign responsibility
for opinions to sources. Tutu uses this resource four times in relation to claims of
innocence, the meaning of ubuntu, reputations and the values of the new South African
democracy:

Amnesty is not given to innocent people or to those who claim to be innocent.

This is a far more personal approach, which sees the offence as something that has happened to people
and whose consequence is a rupture in relationships.

Many of those who have come forward had previously been regarded as respectable members of their
communities

copyright J R Martin & David Rose Working with Discourse

 41

the new culture of respect for human rights and acknowledgment of responsibility and accountability by
which the new democracy wishes to be characterised

These projections within clauses include ‘saying’ claim to be, ‘seeing’ sees, been
regarded as and ‘feeling’ wishes to be. The Act also uses projections within clauses in
relation to claims of victimhood, and in relation to the powers of the Truth and
Reconciliation Commission:

… the gathering of information and the receiving of evidence from any person, including persons claiming
to be victims of such violations or the representatives of such victims…

…establish such offices as it may deem necessary for the performance of its functions

…conduct any investigation or hold any hearing it may deem necessary and establish the investigating unit
referred to in section 28

These are all examples of ‘saying’ claiming to be, may deem.

Finally we need to consider cases where punctuation is used to signal that someone
else’s words are being used. Helena does this several times in her story:

Even if he was an Englishman, he was popular with all the ‘Boer’ Afrikaners. And all my girlfriends envied
me. Then one day he said he was going on a ‘trip’.

Abruptly mutter the feared word ‘trip’ and drive off.

The role of ‘those at the top’, the ‘cliques’ and ‘our men’ who simply had to carry out their bloody orders…
like ‘vultures’. And today they all wash their hands in innocence and resist the realities of the Truth
Commission. Yes, I stand by my murderer who let me and the old White South Africa sleep peacefully.
Warmly, while ‘those at the top’ were again targeting the next ‘permanent removal from society’ for the
vultures.

…there must have been someone out there who is still alive and who can give a face to ‘the orders from
above’ for all the operations.

This device is sometimes referred to as ‘scare quotes’, and warns readers that these
are not Helena’s words but someone else’s – for example the wording of her second
love or white South African leaders. In spoken discourse speakers might use special
intonation or voice quality to signal projection of this kind, and sometimes people use
gesture to mimic quotation marks – acting out the special punctuation. The effect of this
is to disown the evaluation embodied in the highlighted terms, attributing it to an
alternative, unspecified, but usually recoverable source.

In sum we have seen four ways in which projection is used to attribute sources – as
projecting clauses, as names for speech acts, as projecting within clauses, and as scare
quotes. Examples of these are given in Table 2.6.

copyright J R Martin & David Rose Working with Discourse

 42

Table 2.6: Projecting sources

projecting Then he says: He and three of our friends have been promoted.
clauses I know where everything began, the background.
names for I end with a few lines that my wasted vulture said to me
‘speech acts’ they broadcast substantial extracts:
projecting within Many of those who have come forward had previously been regarded
clauses as respectable
such offices as it may deem necessary
scare quotes ‘those at the top’, the ‘cliques’ and ‘our men’

Modality

Alongside projection, another way of introducing additional voices into a text is via
modality, which we introduced above in relation to amplification. Halliday (1994)
describes modality as a resource which sets up a semantic space between yes and no
– a cline running between positive and negative poles. There are two general kinds of
modality, one for negotiating services, and the other for negotiating information.
Demands for a service can be negotiated as follows.

do it positive
you must do it
you should do it
you could do it
don’t do it negative

On this scale we can say ‘how obliged’ you are to act. Statements that give information
can be negotiated as follows.

it is positive
it must be
it should be
it might be
it isn’t negative

On this scale we can say ‘how probable’ a statement is. At each pole of these scales of
modality is the choice of positive or negative polarity. As we’ve said, modality can be
used as a resource for introducing additional voices into a text, and this includes
polarity. To see how this works, let’s start with polarity and the role of negation. Tutu
begins his exposition with a question, which he follows up immediately with a negative
clause:

So is amnesty being given at the cost of justice being done? This is not a frivolous question, but a very
serious issue, one which challenges the integrity of the entire Truth and Reconciliation process.

What Tutu is doing here is countering anyone who thinks that the cost of justice issue is
a frivolous question (or perhaps anyone who says Tutu thinks it’s frivolous). He uses a
negative clause to pre-empt this position before it can cloud the discussion. Negation
places his voice in relation to a potential opposing one; two voices are implicated. In
this respect negative polarity is different from positive polarity; all things being equal,

copyright J R Martin & David Rose Working with Discourse

 43

positive polarity invokes one voice - whereas negative polarity invokes two. Here are
some more examples from Tutu’s exposition:

It is also not true that the granting of amnesty encourages impunity in the sense that perpetrators can
escape completely the consequences of their actions, because amnesty is only given to those who plead
guilty, who accept responsibility for what they have done. Amnesty is not given to innocent people or to
those who claim to be innocent.

It is important to note too that the amnesty provision is an ad hoc arrangement meant for this specific
purpose. This is not how justice is to be administered in South Africa for ever. It is for a limited and
definite period and purpose.

Further, retributive justice – in which an impersonal state hands down punishment with little consideration
for victims and hardly any for the perpetrator – is not the only form of justice.

Negation of this kind is a feature of persuasive writing where contesting positions need
to be addressed and set aside. Tutu does not in fact allow his opposition to speak, as
he would have to do if using projection to give them a voice; the position he gives them
is that of a voice acknowledged but denied.

As we have seen, modality can be interpreted as a resource for grading polarity – for
setting up degrees of positivity and negativity (an intermediate space between yes and
no). Here are more two examples of these scales, beginning with positive at the top
and sliding through to negative at the bottom:

there is a clique positive

there must certainly be a clique
there would probably be a clique
there might possibly be a clique
there isn’t a clique negative

the application is dealt with in a public hearing positive

the application must be dealt with in a public hearing
the application should be dealt with in a public hearing
the application could be dealt with in a public hearing
the application is not dealt with in a public hearing negative

Modality functions very much like negation when we look at it in terms of these scales
(cf. Fuller 1998). Arguing that something must be the case, for example, sounds
assertive but in fact allows an element of doubt; it’s stronger than saying something
would be true, but not as strong as avoiding modality completely and arguing it is the
case. So modality, like polarity, acknowledges alternative voices around a suggestion
or claim. Unlike polarity, it doesn’t take these voices on and deny them; rather it opens
up a space for negotiation, in which different points of view can circulate around an
issue – a space perhaps for mediation and possible reconciliation.

Tutu uses a range of modal resources in his exposition to acknowledge alternative

positions, including usuality when he is generalising about the effects of a public
hearing:

copyright J R Martin & David Rose Working with Discourse

 44

It was often the very first time that their communities and even sometimes their families heard that these
people were, for instance, actually members of death squads or regular torturers of detainees in their
custody.

Here the grading has to do with ‘how often’ something happened, along a scale like the
following:

it was the first time their families heard…

it was always the first time their families heard
it was often the first time their families heard
it was sometimes the first time their families heard
it wasn’t the first time their families heard

Helena uses modality even more often, across a range of modal meanings:

Negotiating information
how usual He and his friends would visit regularly
how probable there must have been someone out there who is still alive

Negotiating services
how obliged I had to watch how white people became dissatisfied with the best
how inclined I would have done the same had I been denied everything
how able who can give a face to ‘the orders from above’ for all the operations

These examples show the five types of modality discussed by Halliday (1994) – usuality,
probability, obligation, inclination and ability.

The Act, because it is concerned with what should happen is mainly concerned with
obligation (how obliged people are to act).

AND SINCE it is deemed necessary to establish the truth in relation to past events as well as the
motives for and circumstances in which gross violations of human rights have occurred, and to make the
findings known in order to prevent a repetition of such acts in future;
AND SINCE the Constitution states that the pursuit of national unity, the well-being of all South African
citizens and peace require reconciliation between the people of South Africa and the reconstruction of
society;
AND SINCE the Constitution states that there is a need for understanding but not for vengeance, a
need for reparation but not for retaliation, a need for ubuntu but not for victimisation;
AND SINCE the Constitution states that in order to advance such reconciliation and reconstruction
amnesty shall be granted in respect of acts, omissions and offences associated with political objectives
committed in the course of the conflicts of the past…

The last example here makes use of what we might call ‘legislative’ shall – to signal
incontestable obligation. By Chapter 2 of the Act, this use of shall becomes dominant
as the various processes around the establishment of the Truth and Reconciliation
Commission are prescribed:

copyright J R Martin & David Rose Working with Discourse

 45

(3) In order to achieve the objectives of the Commission-

(a) the Committee on Human Rights Violations, as contemplated in Chapter 3, shall deal, among
other things, with matters pertaining to investigations of gross violations of human rights;
(b) the Committee on Amnesty, as contemplated in Chapter 4, shall deal with matters relating to
amnesty;
(c) the Committee on Reparation and Rehabilitation, as contemplated in Chapter 5, shall deal with
matters referred to it relating to reparations;
(d) the investigating unit referred to in section 5(d) shall perform the investigations contemplated in
section 28(4)(a); and
(e) the subcommittees shall exercise, perform and carry out the powers, functions and duties
conferred upon, assigned to or imposed upon them by the Commission.

Some projections also include modality or polarity in their meaning, and so can be
interpreted as heteroglossic with respect to both projection and modalisation (Hyland
1998). Tutu uses three of these:

They denied that they had committed a crime, claiming that they had assaulted him only in retaliation for
his inexplicable conduct in attacking them.

I contend that there is another kind of justice, restorative justice.

Denied includes the meaning of ‘not true’; claiming allows for doubt; contend is less
strong than claim (more ‘should be’ than ‘must be true’).

Concession

The third resource we need to consider, as far as heteroglossia in discourse is

concerned, is known as ‘counterexpectancy’. This is more a feature of Helena’s
narrative than the exposition or act, and has to do with the way she tracks readers’
expectations, adjusting them as her story unfolds. In her prayer for example, she tells
God she can’t handle her second love anymore, creating an expectation as she does so
that she will try to leave. Then she counters this by saying that she can’t leave.

I can’t handle the man anymore! But, I can’t get out.

In this example Helena uses the conjunction but to signal that she is countering an
expectation that she’s created for the reader. At any point in a text, readers have an
expectation about what is likely to follow, and Helena takes this into account as she
counters it. In other words she is acknowledging voices in addition to her own, in this
case her readers’. Here some more examples of this kind of monitoring from Helena’s
story:

Not quite my first love, but an exceptional person.

He tried to hide his wild consuming fear, but I saw it.
Eyes bewildered, but dull like the dead.
I can understand if … de Klerk says he didn’t know, but dammit, there must be a clique,

Conjunctions like but, that counter expectations, are termed concessive. Concessive
conjunctions are discussed further in Chapter 4. Here we will review how they are used
to monitor the reader’s expectations.

copyright J R Martin & David Rose Working with Discourse

 46

But is the most common conjunction used to signal concession. But there are other
possibilities, including however and although, and variations on the theme including
even if and even by; in fact, at least, indeed; and nevertheless, needless to say, of
course, admittedly, in any case etc.

Even if God and everyone else forgives me a thousand times – I have to live with this hell.

Even if here means ‘more than expected’ – given the condition of forgiveness his
continued hell is unexpected.
I envy and respect the people of the struggle – at least their leaders have the guts to stand by their
vultures, to recognise their sacrifices.

Spiritual murder is more inhumane than a messy, physical murder. At least a murder victim rests.

Monitoring expectancy is in fact a pervasive feature of conjunctions, realised as time

and contrast as well as causes. In the next examples suddenly means ‘sooner than
expected’ and instead of resting at night implies that ‘resting at night’ is what we’d
normally expect.

They even stayed over for long periods. Suddenly, at strange times, they would become restless.

Instead of resting at night, he would wander from window to window.

Tutu also makes some use of concession in his exposition…

Here the central concern is not retribution or punishment but, in the spirit of ubuntu, the healing of
breaches, the redressing of imbalances, the restoration of broken relationships.

…including the ‘internal’ rhetorical sense of ‘in spite of what I’ve led you to expect me to
say’ (as opposed to the ‘external’ meaning ‘in spite of what you expect to happen’).
Here Tutu means that although he’s granted that public hearings weren’t an absolute
requirement, in fact virtually all important cases were heard that way:

The Act required that where the offence is a gross violation of human rights – defined as an abduction,
killing, torture or severe ill-treatment – the application should be dealt with in a public hearing unless such
a hearing was likely to lead to a miscarriage of justice (for instance, where witnesses were too intimidated
to testify in open session). In fact, virtually all the important applications to the Commission have been
considered in public in the full glare of television lights.

Alongside conjunctions another important set of resources for adjusting expectations

are continuatives. These are like conjunctions but they occur inside the clause, rather
than at the beginning. They include words like already, finally still and only, just, even.
Continuatives that express time indicate that something happens sooner or later, or
persists longer than one might expect. In the following example Helena’s comments on
white peoples’ greed as persisting longer than one might reasonably expect:

If I had to watch how white people became dissatisfied with the best and still wanted better and got it.

copyright J R Martin & David Rose Working with Discourse

 47

Other continuatives indicate that there is more or less to a situation than has been
implied:

It was the beginning of a beautiful relationship. We even spoke about marriage.

Amnesty didn’t matter. It was only a means to the truth.

Tutu uses much less of this resource to adjust expectancy:

They denied that they had committed a crime, claiming that they had assaulted him only in retaliation for
his inexplicable conduct in attacking them.

Summing up then, what we have are three main appraisal systems – attitude,
amplification and source. Attitude comprises affect, judgement and appreciation – our
three major regions of feeling. Amplification covers grading, including force and focus;
force involves the choice to raise or lower the intensity of gradable items, focus the
option of sharpening or softening an experiential boundary. Source covers resources
that introduce additional voices into a discourse, via projection, modalisation or
concession; the key choice here is one voice (monogloss) or more than one voice
(heterogloss). Technically sourcing resources are referred to as engagement. These
key appraisal systems are outlined in Figure 2.6. In order to more accurately reflect the
ways in which people combine different kinds of engagement, attitude and graduation in
discourse, the network now contains three simultaneous systems for these regions of
appraisal. That is we can choose from all of them at the same time.

Figure 2.6: Appraisal systems – an overview

monogloss
PROJECTION…
ENGAGEMENT
heterogloss MODALITY…

CONCESSION…
AFFECT…

APPRAISAL ATTITUDE JUDGEMENT…

APPRECIATION…

FORCE…

GRADUATION

FOCUS…

copyright J R Martin & David Rose Working with Discourse

 48

2.4 Prosody and genre

Appraisal resources are used to establish the tone or mood of a passage of discourse,
as choices resonate with one another from one moment to another as a text unfolds.
The pattern of choices is thus ‘prosodic’ – they form a prosody of attitude running
through the text that swells and diminishes, in the manner of a musical prosody. The
prosodic pattern of appraisal choices constructs the ‘stance’ or ‘voice’ of the appraiser,
and this stance or voice defines the kind of community that is being set up around
shared values. In everyday language, these stances are often discussed as ranging
along a scale - from more objective to more subjective.

As we have seen, among our genres the Act is well down towards the objective end of
this scale, especially where it is defining terms:

CHAPTER 1
Interpretation and application
Definitions
1. (1) In this Act, unless the context otherwise indicates-
(i) “act associated with a political objective” has the meaning ascribed thereto in section 20(2) and (3);
(ii)
(ii) “article” includes any evidence, book, document, file, object, writing, recording or transcribed
computer printout produced by any mechanical or electronic device or any device by means of which
information is recorded, stored or transcribed; (xix)
(iii) “Commission” means the Truth and Reconciliation Commission established by section 2; (ix)
(iv) “commissioner” means a member of the Commission appointed in terms of section 7(2)(a); (viii)
(v) “committee” means the Committee on Human Rights Violations, the Committee on Amnesty or the
Committee on Reparation and Rehabilitation, as the case may be; (vii)
(vi) “Constitution” means the Constitution of the Republic of South Africa, 1993 (Act No. 200 of 1993); (iv)
(vii) “cut-off date” means the latest date allowed as the cut-off date in terms of the Constitution as set out
under the heading “National Unity and Reconciliation”; (i)
(viii) “former state” means any state or territory which was established by an Act of Parliament or by
proclamation in terms of such an Act prior to the commencement of the Constitution and the territory
of which now forms part of the Republic; (xvii)…

As far as appraisal resources are concerned, this kind of ‘objectivity’ appears to involve
a range of factors – basically as little attitude, graduation and heteroglossia as possible.
We might think of this as a kind of faceless stance. But the absence of feelings,
intensification and alternative voices is itself a face – a cool excluding one perhaps, but
it is a face. In the phase reproduced above, the Act is in fact taking pains to construct a
very precise interpretation of the Truth and Reconciliation Commission. We could argue
that the resources used to define terms (highlighted above) are in fact monoglossing
resources designed to ensure that the Act speaks with one clear voice as far as its
administrative goals are concerned. The absence of amplification helps to keep the
definitions tight; the near absence of projection, modalisation and concession works to
reinforce the monovocal stance. This is not in other words a text to argue with; it’s the
law.

At the other end of the spectrum we have Helena’s narrative, which uses a full range of
attitudinal, graduating and engaging resources. Perhaps this is why we find narrative so
enjoyable across a range of modalities (books, film, chat, TV, comics, radio and so on).

copyright J R Martin & David Rose Working with Discourse

 49

There’s all kinds of feelings to share; the volume pumps up and down to keep us
interested and boundaries are squished to loosen or tighten things up; lots of voices are
invoked – we get to be a part of things. It all hangs out.

Tutu’s exposition is somewhere in between. It uses attitude, but not a full range. There
is some, but not a lot of intensification; focus is used sparingly, to be precise
(sharpening boundaries). Alternative voices are acknowledged, but mainly for rebuttal.
Tutu is talking us round, persuading us – he makes the effort to argue for his position,
unlike the Act which just declares.

These kinds of differences between genres are reflected in differences between stages
within genres. As noted above, Helena has more affect in her Incidents than her
Interpretations, which tend in their turn to foreground judgement. Tutu’s first and
second arguments focus on judgement, but appreciation (around the concept of ubuntu)
plays a major role in his third. The Act avoids appraisal almost completely in its
definitions, but earlier projects the voice of the South African Constitution, including its
modalisation and reconciliation oriented appreciations:

AND SINCE the Constitution states that the pursuit of national unity, the well-being of all South African
citizens and peace require reconciliation between the people of South Africa and the reconstruction of
society;
AND SINCE the Constitution states that there is a need for understanding but not for vengeance, a
need for reparation but not for retaliation, a need for ubuntu but not for victimisation;

As texts unfold in other words they try to move us in different ways – to form different
kinds of relationship with us – to commune with us strategically. Appraisal is to rhetoric
as conjunction is to logic we might say; it unfolds dynamically to engage us, to get us on
side – not with one appeal, but through a spectrum of manoeuvres that work themselves
out phase by phase.

We’ll close this section with an examples of stance shifting from one of Jim’s papers
where he tried to figure out what he found so moving about the final couple of pages of
Nelson Mandela’s autobiography Long Walk to Freedom (a text we’ll return to in 7). He
worked on this as a functional linguist and semiotician and organised his paper in 6
sections:

1. The text
2. Abstracting freedom
3. Enlightenment
4. Engagement
5. Recontextualisation
6. Grace

In order to get it published in a scholarly journal (Discourse Studies) he adopted a

relatively objective stance (of the scholarly kind studied by Hunston 1994 and Myers
1989) – not completely faceless, but tending towards the monovocal and cool. By the
end of this study however he was still feeling a little mystified about why the text he was
analysing was so moving – his communion with Mandela, at such a distance in so many
respects, seemed to transcend the sum of the analyses he had undertaken, however

copyright J R Martin & David Rose Working with Discourse

 50

focussed he tried to make them on what was going on. In exasperation, he decided to
shift stance and wrote:
6. Grace

In this paper I have tried, from my own specific reading position, to analyse this instance of discourse in
relation to the meanings I’ve been trained to decode. If allowed a reaction, the term that comes to mind is
grace, in every meaning of the word. The gracefulness with which the recount unfolds, the charm of its
rhetoric, the goodwill to all peoples... I can’t help admiring the texture, and the Mandela it construes for
me. And in this kind of reaction, I am not alone. Consider for example the evaluative terms used
promotionally on the covers of Mandela 1995: anger, sorrow, love, joy, grace, elegance, riveting, brilliantly,
emotive, compelling, uplifting, exhilarating, epic, hardship, resilience, triumph, clarity, eloquence, burns
with the luminosity of faith, invincible, hope, dignity, enthralling, great, indispensable, unique, truly stunning,
extraordinary, vivid, unusual, courage, persistence, tolerance, forgiveness, extraordinary, well worth,
greatness, epic, struggle, idealism, inspired, cynicism, compulsory. What is the appeal?

I suspect what we are examining here is Mandela’s ability to naturalise radical values in terms that disarm
rather than confront. Both the recount and its multimodal recontextualisation promote a politics of freedom
that involves respecting and enhancing the freedom of others. Put into practice, this involves more than an
end to apartheid and reconciliation with its perpetrators. Ultimately it involves the reconfiguration of a
global economic order which distributes resources so unevenly that it has to be propped up by all manner
of unbearable regimes. In a sense then, Mandela is promoting socialism in the name of freedom; he
naturalises a comfortable reading position for those who might oppose his aims, and at the same time
gives his sympathisers an inspirational shot in the arm. If discourse analysts are serious about wanting to
use their work to enact social change, then they will have to broaden their coverage to include discourse of
this kind - discourse that inspires, encourages, heartens; discourse we like, that cheers us along. We
need, in other words, more positive discourse analysis (PDA?) alongside our critique; and this means
dealing with texts we admire, alongside those we dislike and try to expose (Wodak 1996)… [Martin 1999:
51-52]

As the first paragraph indicates, Jim decided to change gears (invoking professorial
privilege perhaps), to stop analysing the text for a moment and just react… he evaluates
the text as graceful, he’s charmed by it, he admires it and of course the man who wrote
it. And he shows that in these reactions he’s not alone, canvassing the range of
attitudes used for promotional purposes on the covers of Mandela’s book. Then quickly,
before getting trounced by his editor, he pulls back in the following paragraph, gets cool,
goes scholarly again – and tries to figure out again, in academic terms, how Mandela
manages to pull it off. He’s wrestling with the sublime perhaps, but felt he got a little
closer by changing gears than by plodding along. Allowing himself a reaction got him
thinking about how Mandela finesses his radical politics, disarming people round the
world into communities of admiring fans. And it allowed Jim to make a further point
about the importance of focussing on heartening discourses of this kind instead of being
so depressingly critical all the time by focussing solely on hegemony and all that’s wrong
with the world. Ultimately what the change of voices achieved perhaps was to bring a
community of critical discourse analysts into contact with a community of political
activists and supporters – by way of saying, c’mon guys, let’s do something about this.
These guys are heroes; let’s see how they move the world along. Two voices, in
tandem worked, where a single voice, the scholarly one, seemed to be kind of missing
the point.

Just as we’ve been changing voices here; we’ll pull back now, just in case our scholarly
credentials are wearing a little thin.

copyright J R Martin & David Rose Working with Discourse

 51

2.5 More detail on kinds of attitude

Appraisal is a huge resource for constructing communities of feeling, and a great deal of
it is realised through lexis as well as grammar, which makes it even more unwieldy to
deploy than we’ve allowed for here. Lexis is the part of language after all that changes
most quickly – it’s flexi-language, designed to adjust quickly to changing times, new
needs. There’s a lot going on and it’s hard to keep up. And a lot of attitude is
specialised, used in some specific registers but not all – when my son, currently 11
years old and a speaker of Marrickville English in Sydney Australia, says something
‘rocks’ you may know where he’s coming from or you may not. Does it mean something
is good or bad? Is rocks an evaluative item at all? So as a final step, we’ll try and give
you some more scaffolding for analysing attitude here. It won’t be enough for all
purposes, but it will get you going.

Affect

Affect can be realised in various grammatical niches. In Halliday’s 1994 terms, these
include ‘qualities’, ‘processes’ and ‘comments’. (Grammatical functions for each type
are given to the right.)

affect as ‘quality’
- describing participants a happy boy Epithet
- attributed to participants the boy was happy Attribute
- manner of processes the boy played happily Circumstance
1
affect as ‘process’
- affective sensing the present pleased the boy Process(effective)
- affective behaving the boy smiled Process (middle)

affect as ‘comment’
- desiderative comment happily, he had a long nap Modal Adjunct

To classify affect, we can ask the following questions:

 Are the feelings positive or negative?
 Are the feelings a surge of emotion or an ongoing mental state?
 Are the feelings reacting to some specific external agency or an ongoing mood?
 Are the feelings as more or less intense?
 Do the feelings involve intention rather than reaction?
 Are the feelings to do with un/happiness, in/security or dis/satisfaction?

Let’s now examine these six question in more detail.

1 Are the feelings popularly construed by the culture as positive (good vibes that
are enjoyable to experience) or negative ones (bad vibes that are better avoided)? We
are not concerned here with the value that a particular psychological framework might
1
Affect as ‘process’ also includes relations such as I'm pleased that..., It's pleasing that... .

copyright J R Martin & David Rose Working with Discourse

 52

place on one or another emotion (cf. “It’s probably productive that you’re feeling sad
because it’s a sign that...”).

- positive affect the boy was happy

- negative affect the boy was sad

2 Are the feelings realised as a surge of emotion involving some kind of embodied
paralinguistic or extralinguistic manifestation, or more prosodically experienced as a
kind of predisposition or ongoing mental state? Grammatically this distinction is
constructed as the opposition between behavioural (e.g. She smiled at him) vs mental
(e.g. She liked him) or relational (e.g. She felt happy with him) processes.

- behavioural surge the boy laughed

- mental disposition the boy liked the present/the boy felt happy

3 Are the feelings construed as directed at or reacting to some specific external

agency (typically conscious) or as a general ongoing mood for which one might pose the
question “Why are you feeling that way?” and get the answer “I’m not sure.”

- reaction to other the boy liked the teacher/the teacher pleased the boy
- undirected mood the boy was happy

4 How are the feelings graded - towards the lower valued end of a scale of intensity
or towards the higher valued end; or somewhere in between? We don’t wish at this
stage to imply that low, median and high are discrete values (as with MODALITY - cf.
Halliday 1994:358-9), but expect that most emotions offer lexicalisations that grade
along a evenly clined scale.

- low the boy liked the present

- ‘median’ the boy loved the present
- high the boy adored the present

5 Do the feelings involve intention (rather than reaction), with respect to a stimulus
that is irrealis (rather than realis).

- realis the boy liked the present

- irrealis the boy wanted the present

Irrealis affect seems always to be directed at some external agency, and so can be
outlined as in Table 2.7 (setting aside parameter iii).

Table 2.7: Irrealis affect

DIS/INCLINATION SURGE (of behaviour) DISPOSITION
fear tremble wary
shudder fearful
cower terrorised
desire suggest incomplete (miss)
request lonely (long for)
implore bereft (yearn for)

copyright J R Martin & David Rose Working with Discourse

 53

6 Finally we can group emotions into three major sets having to do with
un/happiness, in/security and dis/satisfaction.2

- in/security the boy was anxious/confident

- dis/satisfaction the boy was fed up/absorbed
- un/happiness the boy sad/happy

2
The framework for un/happiness, in/security and dis/satisfaction emerged from Jim’s observations of his
young sons, when they were in their first stages of socialisation (up to about 2 years of age), and in
particular on a cycle of demands structuring my elder son’s temper tantrums over a period of several
months. During these tantrums he would insist on having baggy (his blanket), and then when it was
proffered and rejected his bopple (bottle), and then when this was proffered and rejected Mummy or
Daddy (whichever was not present), and then baggy again, then bopple... for up to an hour. If we take
these primal screams as primitives, than a framework involving in/security (blanket), dis/satisfaction
(bottle) and un/happiness (Mummy/Daddy) can be entertained. The in/security variable covers emotions
concerned with ecosocial well-being - anxiety, fear, confidence and trust; the dis/satisfaction variable
covers emotions concerned with telos (the pursuit of goals) - ennui, displeasure, curiosity, respect; the
un/happiness variable covers emotions concerned with ‘affairs of the heart’ - sadness, anger, happiness
and love. Unfortunately we have not been able to develop a more principled basis for classifying emotions
in recent years and take little comfort from the array of divergent frameworks available elsewhere in the
literature (including the evolving variations in Martin 1992a, 1996 and 1997a2).

copyright J R Martin & David Rose Working with Discourse

 54

Table 2.8: Realis affect

UN/HAPPINESS SURGE (of behaviour) DISPOSITION
unhappiness: misery whimper down [low]
[mood: ‘in me’] cry sad [median]
wail miserable [high]
unhappiness: antipathy rubbish dislike
[directed feeling: ‘at you’] abuse hate
revile abhor
happiness: cheer chuckle cheerful
laugh buoyant
rejoice jubilant
happiness: affection shake hands fond
hug loving
cuddle adoring
IN/SECURITY
insecurity: disquiet restless uneasy
twitching anxious
shaking freaked out
insecurity: surprise start taken aback
cry out surprised
faint astonished
security: confidence declare confident
assert assured
proclaim boastful
security: trust delegate comfortable with
commit confident in/about
entrust trusting
DIS/SATISFACTION
dissatisfaction: ennui fidget bored
yawn fed up
tune out exasperated
dissatisfaction: displeasure caution cross
scold angry
castigate furious
satisfaction: interest attentive curious
busy absorbed
flat out engrossed
satisfaction: admiration pat on the back satisfied
compliment impressed
reward proud

Judgement

Judgement can be thought of as the institutionalisation of feeling, in the context of

proposals (norms about how people should and shouldn’t behave). Like affect, it has a
positive and negative dimension - corresponding to positive and negative judgements
about behaviour. Media research reported in Iedema et al. 1994 has suggested dividing
judgements into two major groups, social esteem and social sanction. Social esteem
involves admiration and criticism, typically without legal implications; if you have
difficulties in this area you may need a therapist. Social sanction on the other hand
involves praise, and condemnation, often with legal implications; if you have problems in

copyright J R Martin & David Rose Working with Discourse

 55

this area you may need a lawyer. Judgements of esteem have to do with normality
(how unusual someone is), capacity (how capable they are) and tenacity (how resolute
they are); judgements of sanction have to do with veracity (how truthful someone is) and
propriety (how ethical someone is).

Table 2.9: Types of judgement

SOCIAL ESTEEM ‘venial’ positive [admire] negative [criticise]
normality [fate] lucky, fortunate, charmed... ; unfortunate, pitiful, tragic...
‘is s/he special?’ normal, average, everyday... ; odd, peculiar, eccentric... ;
in, fashionable, avant garde... dated, daggy, retrograde...
capacity powerful, vigorous, robust... ; mild, weak, whimpy... ;
‘is s/he capable?’ insightful, clever, gifted... ; slow, stupid, thick... ;
balanced, together, sane... flaky, neurotic, insane...
tenacity [resolve] plucky, brave, heroic... ; rash, cowardly, despondent... ;
‘is s/he dependable?’ reliable, dependable... ; unreliable, undependable...
tireless, persevering, resolute weak, distracted, dissolute...

SOCIAL SANCTION ‘mortal’ positive [praise] negative [condemn]

veracity [truth] truthful, honest, credible... ; dishonest, deceitful... ;
‘is s/he honest?’ real, authentic, genuine... ; glitzy, bogus, fake... ;
frank, direct... deceptive, manipulative...
propriety [ethics] good, moral, ethical... ; bad, immoral, evil... ;
‘is s/he beyond reproach?’ law abiding, fair, just... ; corrupt, unfair, unjust... ;
sensitive, kind, caring... insensitive, mean, cruel...

The kinds of judgement speakers take up is very sensitive to their institutional position.
For example, only journalists with responsibility for writing editorials and other comment
have a full range of judgmental resources at their disposal; reporters writing hard news
that is meant to sound objective have to avoid explicit judgements completely (Iedema
et al. 1994). The distinction between social esteem and social sanction in other words
has important implications for the subjective or objective flavour of an appraiser’s
stance.

Appreciation

Appreciation can be thought of as the institutionalisation of feeling, in the context of

propositions (norms about how products and performances are valued). Like affect
and judgement it has a positive and negative dimension - corresponding to positive and
negative evaluations of texts and processes (and natural phenomena). The system is
organised around three variables - reaction, composition and valuation. Reaction has to
do with the degree to which the text/process in question captures our attention

copyright J R Martin & David Rose Working with Discourse

 56

(reaction:impact) and the emotional impact it has on us (reaction:quality). Composition

has to do with our perceptions of proportionality (composition:balance) and detail
(composition:complexity) in a text/process. Valuation has to do with our assessment of
the social significance of the text/process.

Table 2.10: Types of appreciation

positive negative
reaction: impact arresting, captivating, involving, dull, boring, tedious, staid...
‘did it grab me?’ engaging, absorbing, imposing,
stunning, striking, compelling,
interesting...
fascinating, exciting, moving... dry, ascetic, uninviting...
remarkable, notable, sensational... unremarkable, pedestrian, ...
lively, dramatic, intense... flat, predictable, monotonous...
reaction: quality lovely, beautiful, splendid... plain, ugly...
‘did I like it?’ appealing, enchanting, pleasing, repulsive, off-putting, revolting,
delightful, attractive, welcome... irritating, weird...
composition: balance balanced, harmonious, unified, unbalanced, discordant,
‘did it hang together?” symmetrical, proportional... unfinished, incomplete..

composition: complexity simple, elegant... ornamental, over-complicated,

‘was it hard to follow?’ intricate, rich, detailed, precise... extravagant, puzzling...
monolithic, simplistic...
valuation challenging, significant, deep, profound, shallow, insignificant, unsatisfying,
‘was it worthwhile?’ provocative, daring... sentimental...
experimental, innovative, original, conservative, reactionary,
unique, fruitful, illuminating... generic...
enduring, lasting... unmemorable, forgettable...

Of these dimensions, valuation is especially tied up with field, since the criteria for
valuing a text/process are for the most part institutionally specific. But beyond this,
since both judgement and appreciation are in a sense institutionalisations of feeling, all
of the dimensions involved will prove sensitive to field. An example of this coupling of
ideational and interpersonal meaning is presented for appreciations of research in the
field of linguistics below.

copyright J R Martin & David Rose Working with Discourse

 57

Table 2.11: Appreciation in a specialised field (linguistics)

linguistics positive negative
reaction: impact timely, long awaited, engaging, untimely, unexpected, overdue, surprising,
[notability] landmark... dated...

reaction: quality fascinating, exciting, interesting, dull, tedious, boring, pedantic, didactic,
[likeability] stimulating, impressive, uninspired...
admirable...
composition: consistent, balanced, thorough, fragmented, loose ended, disorganised,
[balance] considered, unified, logical, well contradictory, sloppy...
argued, well presented...
composition: simple, lucid, elegant, rich, simplistic, extravagant, complicated,
[complexity] detailed, exhaustive, clear, precise Byzantine, labyrinthine, overly elaborate,
... narrow, vague, unclear, indulgent, esoteric,
eclectic...
valuation useful, penetrating, illuminating, shallow, ad hoc, reductive, unconvincing,
[field genesis] challenging, significant, deep, unsupported, fanciful, tendentious, bizarre,
profound, satisfying, fruitful... counterintuitive, perplexing, arcane...

Further complicating this issue is the implicit coupling of field with appreciation (the
evocation variable noted above). As with affect and judgement, ideational meanings
can be used to appraise, even though explicitly evaluative lexis is avoided. It perhaps
should be stressed again here that appraisal analysts do need to declare their reading
position - in particular since the evaluation one makes of evocations depends on the
institutional position one is reading from. For example, according to reading position,
formal and functional linguists will evaluate terms in the following sets of oppositions in
complementary ways - with firm convictions about what the good guys and the bad guys
should celebrate:
rule/resource:: cognitive/social:: acquisition/development:: syntagmatic/paradigmatic:: form/function::
language/parole:: system/process:: psychology&philosophy/sociology&anthropology:: cognitive/social::
theory/description:: intuition/corpus:: knowledge/meaning:: syntax/discourse:: pragmatics/context::
parsimony/extravagance:: cognitive/critical:: technicist/humanist:: truth/social action::
performance/instantiation:: categorical/probabilistic:: contradictory/complementary:: proof/exemplification::
reductive/comprehensive:: arbitrary/natural:: modular/fractal:: syntax&lexicon/lexicogrammar...

copyright J R Martin & David Rose Working with Discourse

 143

copyright J R Martin & David Rose Working with Discourse