CHAPTER ONE

THE WORK OF REPRESENTATION

Stuart Hall

1 REPRESENTATION, MEANING AND LANGUAGE

In this chapter we will be concentrating on one of the key processes in the ‘cultural circuit’ (see
Du Gay et al., 1997, and the Introduction to this volume) – the practices of representation. The aim
of this chapter is to introduce you to this topic, and to explain what it is about and why we give it
such importance in cultural studies.
The concept of representation has come to occupy a new and important place in the study of
culture. Representation connects meaning and language to culture. But what exactly do people
mean by it? What does representation have to do with culture and meaning? One common-sense
usage of the term is as follows: ‘Representation means using language to say something meaning-
ful about, or to represent, the world meaningfully, to other people.’ You may well ask, ‘Is that all?’
Well, yes and no. Representation is an essential part of the process by which meaning is produced
and exchanged between members of a culture. It does involve the use of language, of signs and
images which stand for or represent things. But this is a far from simple or straightforward process,
as you will soon discover.
How does the concept of representation connect meaning and language to culture? In order to
explore this connection further, we will look at a number of different theories about how language is
used to represent the world. Here we will be drawing a distinction between three different accounts
or theories: the reflective, the intentional and the constructionist approaches to representation. Does
language simply reflect a meaning which already exists out there in the world of objects, people
and events (reflective)? Does language express only what the speaker or writer or painter wants to
say, his or her personally intended meaning (intentional)? Or is meaning constructed in and through
language (constructionist)? You will learn more in a moment about these three approaches.
Most of the chapter will be spent exploring the constructionist approach, because it is this
perspective which has had the most significant impact on cultural studies in recent years. This

01-Hall_Ch-01.indd 1 18/04/2013 12:23:49 PM

 Representation

chapter chooses to examine two major variants or models of the constructionist approach – the semi-
otic approach, greatly influenced by the Swiss linguist, Ferdinand de Saussure, and the discursive
approach, associated with the French philosopher and historian, Michel Foucault. Later chapters
in this book will take up these two theories again, among others, so you will have an opportunity
to consolidate your understanding of them, and to apply them to different areas of analysis. Other
chapters will introduce theoretical paradigms which apply constructionist approaches in different
ways to that of semiotics and Foucault. All, however, put in question the very nature of representation.
We turn to this question first.

1.1 Making meaning, representing things

What does the word representation really mean, in this context? What does the process of repre-
sentation involve? How does representation work?
To put it briefly, representation is the production of meaning through language. The Shorter
Oxford English Dictionary suggests two relevant meanings for the word:

1 To represent something is to describe or depict it, to call it up in the mind by description or por-
trayal or imagination; to place a likeness of it before us in our mind or in the senses; as, for exam-
ple, in the sentence, ‘This picture represents the murder of Abel by Cain.’
2 To represent also means to symbolize, stand for, to be a specimen of, or to substitute for; as in the
sentence, ‘In Christianity, the cross represents the suffering and crucifixion of Christ.’

The figures in the painting stand in the place of, and at the same time, stand for the story of Cain and
Abel. Likewise, the cross simply consists of two wooden planks nailed together; but in the context of
Christian belief and teaching, it takes on, symbolizes or comes to stand for a wider set of meanings
about the crucifixion of the Son of God, and this is a concept we can put into words and pictures.

ACTIVITY 1
Here is a simple exercise about representation. Look at any familiar object in the room. You will immediately
recognize what it is. But how do you know what the object is? What does ‘recognize’ mean?
Now try to make yourself conscious of what you are doing – observe what is going on as you do it. You
recognize what it is because your thought processes decode your visual perception of the object in terms
of a concept of it which you have in your head. This must be so because, if you look away from the object,
you can still think about it by conjuring it up, as we say, ‘in your mind’s eye’. Go on – try to follow the
process as it happens: there is the object ... and there is the concept in your head which tells you what it
is, what your visual image of it means.
Now, tell me what it is. Say it aloud: ‘It’s a lamp’ – or a table or a book or the phone or whatever. The
concept of the object has passed through your mental representation of it to me via the word for it which

01-Hall_Ch-01.indd 2 18/04/2013 12:23:49 PM

 The Work of Representation

you have just used. The word stands for or represents the concept, and can be used to reference or
designate either a ‘real’ object in the world or indeed even some imaginary object, like angels dancing on
the head of a pin, which no one has ever actually seen.

This is how you give meaning to things through language. This is how you ‘make sense of’ the
world of people, objects and events, and how you are able to express a complex thought about
those things to other people, or communicate about them through language in ways which other
people are able to understand.
Why do we have to go through this complex process to represent our thoughts? If you put down
a glass you are holding and walk out of the room, you can still think about the glass, even though
it is no longer physically there. Actually, you can’t think with a glass. You can only think with the
concept of the glass. As the linguists are fond of saying, ‘Dogs bark. But the concept of “dog” can-
not bark or bite.’ You can’t speak with the actual glass, either. You can only speak with the word
for glass – GLASS – which is the linguistic sign which we use in English to refer to objects out
of which you drink water. This is where representation comes in. Representation is the production
of the meaning of the concepts in our minds through language. It is the link between concepts and
language which enables us to refer to either the ‘real’ world of objects, people or events, or indeed
to imaginary worlds of fictional objects, people and events.
So there are two processes, two systems of representation, involved. First, there is the ‘system’
by which all sorts of objects, people and events are correlated with a set of concepts or mental rep-
resentations which we carry around in our heads. Without them, we could not interpret the world
meaningfully at all. In the first place, then, meaning depends on the system of concepts and images
formed in our thoughts which can stand for or ‘represent’ the world, enabling us to refer to things
both inside and outside our heads.
Before we move on to look at the second ‘system of representation’, we should observe that what
we have just said is a very simple version of a rather complex process. It is simple enough to see how
we might form concepts for things we can perceive – people or material objects, like chairs, tables
and desks. But we also form concepts of rather obscure and abstract things, which we can’t in any
simple way see, feel or touch. Think, for example, of our concepts of war, or death, or friendship or
love. And, as we have remarked, we also form concepts about things we have never seen, and possibly
can’t or won’t ever see, and about people and places we have plainly made up. We may have a clear
concept of, say, angels, mermaids, God, the Devil, or of Heaven and Hell, or of Middlemarch (the
fictional provincial town in George Eliot’s novel), or Elizabeth (the heroine of Jane Austen’s Pride
and Prejudice).
We have called this a ‘system of representation’. That is because it consists not of individual
concepts, but of different ways of organizing, clustering, arranging and classifying concepts, and
of establishing complex relations between them. For example, we use the principles of similarity
and difference to establish relationships between concepts or to distinguish them from one another.
Thus, I have an idea that in some respects birds are like planes in the sky, based on the fact that they
are similar because they both fly – but I also have an idea that in other respects they are different,

01-Hall_Ch-01.indd 3 18/04/2013 12:23:49 PM

 Representation

because one is part of nature while the other is man-made. This mixing and matching of relations
between concepts to form complex ideas and thoughts is possible because our concepts are arranged
into different classifying systems. In this example, the first is based on a distinction between flying/not
flying and the second is based on the distinction between natural/man-made. There are other princi-
ples of organization like this at work in all conceptual systems: for example, classifying according
to sequence – which concept follows which – or causality – what causes what – and so on. The point
here is that we are talking about not just a random collection of concepts, but concepts organized,
arranged and classified into complex relations with one another. That is what our conceptual system
actually is like. However, this does not undermine the basic point. Meaning depends on the relation-
ship between things in the world – people, objects and events, real or fictional – and the conceptual
system, which can operate as mental representations of them.
Now it could be the case that the conceptual map which I carry around in my head is totally differ-
ent from yours, in which case you and I would interpret or make sense of the world in totally different
ways. We would be incapable of sharing our thoughts or expressing ideas about the world to each
other. In fact, each of us probably does understand and interpret the world in a unique and individual
way. However, we are able to communicate because we share broadly the same conceptual maps and
thus make sense of or interpret the world in roughly similar ways. That is indeed what it means when
we say we ‘belong to the same culture’. Because we interpret the world in roughly similar ways, we
are able to build up a shared culture of meanings and thus construct a social world which we inhabit
together. That is why ‘culture’ is sometimes defined in terms of ‘shared meanings or shared conceptual
maps’ (see Du Gay et al., 1997).
However, a shared conceptual map is not enough. We must also be able to represent or
exchange meanings and concepts, and we can only do that when we also have access to a shared
language. Language is therefore the second system of representation involved in the overall
process of constructing meaning. Our shared conceptual map must be translated into a common
language, so that we can correlate our concepts and ideas with certain written words, spoken
sounds or visual images. The general term we use for words, sounds or images which carry
meaning is signs. These signs stand for or represent the concepts and the conceptual relations
between them which we carry around in our heads and together they make up the meaning-
systems of our culture.
Signs are organized into languages and it is the existence of common languages which enable
us to translate our thoughts (concepts) into words, sounds or images, and then to use these, oper-
ating as a language, to express meanings and communicate thoughts to other people. Remember
that the term ‘language’ is being used here in a very broad and inclusive way. The writing system
or the spoken system of a particular language are both obviously ‘languages’. But so are visual
images, whether produced by hand, mechanically, electronically, digitally or some other means,
when they are used to express meaning. And so are other things which aren’t ‘linguistic’ in any
ordinary sense: the ‘language’ of facial expressions or of gesture, for example, or the ‘language’
of fashion, of clothes, or of traffic lights. Even music is a ‘language’, with complex relations
between different sounds and chords, though it is a very special case since it can’t easily be used

01-Hall_Ch-01.indd 4 18/04/2013 12:23:49 PM

 The Work of Representation

to reference actual things or objects in the world (a point further elaborated in Du Gay, ed., 1997,
and Mackay, ed., 1997). Any sound, word, image or object which functions as a sign, and is
organized with other signs into a system which is capable of carrying and expressing meaning is,
from this point of view, ‘a language’. It is in this sense that the model of meaning which I have
been analysing here is often described as a ‘linguistic’ one; and that all the theories of meaning
which follow this basic model are described as belonging to ‘the linguistic turn’ in the social sciences
and cultural studies.
At the heart of the meaning process in culture, then, are two related ‘systems of representa-
tion’. The first enables us to give meaning to the world by constructing a set of correspondences
or a chain of equivalences between things – people, objects, events, abstract ideas, etc. – and our
system of concepts, our conceptual maps. The second depends on constructing a set of correspond-
ences between our conceptual map and a set of signs, arranged or organized into various languages
which stand for or represent those concepts. The relation between ‘things’, concepts and signs lies
at the heart of the production of meaning in language. The process which links these three elements
together is what we call ‘representation’.

1.2 Language and representation

Just as people who belong to the same culture must share a broadly similar conceptual map, so
they must also share the same way of interpreting the signs of a language, for only in this way can
meanings be effectively exchanged between people. But how do we know which concept stands for
which thing? Or which word effectively represents which concept? How do I know which sounds or
images will carry, through language, the meaning of my concepts and what I want to say with them
to you? This may seem relatively simple in the case of visual signs, because the drawing, painting,
camera or TV image of a sheep bears a resemblance to the animal with a woolly coat grazing in a
field to which I want to refer. Even so, we need to remind ourselves that a drawn or painted or digital
version of a sheep is not exactly like a ‘real’ sheep. For one thing, most images are in two dimensions
whereas the ‘real’ sheep exists in three dimensions.
Visual signs and images, even when they bear a close resemblance to the things to which they
refer, are still signs: they carry meaning and thus have to be interpreted. In order to interpret them,
we must have access to the two systems of representation discussed earlier: to a conceptual map
which correlates the sheep in the field with the concept of a ‘sheep’; and a language system which in
visual language, bears some resemblance to the real thing or ‘looks like it’ in some way. This argu-
ment is clearest if we think of a cartoon drawing or an abstract painting of a ‘sheep’, where we need
a very sophisticated conceptual and shared linguistic system to be certain that we are all ‘reading’
the sign in the same way. Even then we may find ourselves wondering whether it really is a picture
of a sheep at all. As the relationship between the sign and its referent becomes less clear-cut, the
meaning begins to slip and slide away from us into uncertainty. Meaning is no longer transparently
passing from one person to another ...

01-Hall_Ch-01.indd 5 18/04/2013 12:23:49 PM

 Representation

FIGURE 1.1 William Holman Hunt, Our English Coasts (‘Strayed Sheep’), 1852

FIGURE 1.2
Q: When is a sheep not a sheep?
A: When it’s a work of art. (Damien Hirst, Away from the Flock, 1994)

So, even in the case of visual language, where the relationship between the concept and the sign
seems fairly straightforward, the matter is far from simple. It is even more difficult with written or
spoken language, where words don’t look or sound anything like the things to which they refer. In

01-Hall_Ch-01.indd 6 18/04/2013 12:23:50 PM

 The Work of Representation

part, this is because there are different kinds of signs. Visual signs are what are called iconic signs.
That is, they bear, in their form, a certain resemblance to the object, person or event to which they
refer. A photograph of a tree reproduces some of the actual conditions of our visual perception in
the visual sign. Written or spoken signs, on the other hand, are what is called indexical.
They bear no obvious relationship at all to the things to which they refer. The letters T, R, E, E do
not look anything like trees in nature, nor does the word ‘tree’ in English sound like ‘real’ trees (if
indeed they make any sound at all!). The relationship in these systems of representation between
the sign, the concept and the object to which they might be used to refer is entirely arbitrary. By
‘arbitrary’ we mean that in principle any collection of letters or any sound in any order would do
the trick equally well. Trees would not mind if we used the word SEERT – ‘trees’ written back-
wards – to represent the concept of them. This is clear from the fact that, in French, quite different
letters and a quite different sound is used to refer to what, to all appearances, is the same thing –
a ‘real’ tree – and, as far as we can tell, to the same concept a large plant that grows in nature.
The French and English seem to be using the same concept. But the concept which in English is
represented by the word TREE is represented in French by the word ARBRE.

1.3 Sharing the codes

The question, then, is: how do people who belong to the same culture, who share the same
conceptual map and who speak or write the same language (English) know that the arbitrary
combination of letters and sounds that makes up the word TREE will stand for or represent the
concept ‘a large plant that grows in nature’? One possibility would be that the objects in the world
themselves embody and fix in some way their ‘true’ meaning. But it is not at all clear that real
trees know that they are trees, and even less clear that they know that the word in English which
represents the concept of themselves is written TREE whereas in French it is written ARBRE! As
far as they are concerned, it could just as well be written COW or VACHE or indeed XYZ. The
meaning is not in the object or person or thing, nor is it in the word. It is we who fix the meaning
so firmly that, after a while, it comes to seem natural and inevitable. The meaning is constructed
by the system of representation. It is constructed and fixed by the code, which sets up the cor-
relation between our conceptual system and our language system in such a way that, every time
we think of a tree, the code tells us to use the English word TREE, or the French word ARBRE.
The code tells us that, in our culture – that is, in our conceptual and language codes – the concept
‘tree’ is represented by the letters T, R, E, E, arranged in a certain sequence, just as in Morse code,
the sign for V (which in the Second World War Churchill made ‘stand for’ or represent ‘Victory’)
is Dot, Dot, Dot, Dash, and in the ‘language of traffic lights’, Green = Go! and Red = Stop!
One way of thinking about ‘culture’, then, is in terms of these shared conceptual maps, shared lan-
guage systems and the codes which govern the relationships of translation between them. Codes fix
the relationships between concepts and signs. They stabilize meaning within different languages and
cultures. They tell us which language to use to convey which idea. The reverse is also true. Codes tell
us which concepts are being referred to when we hear or read which signs. By arbitrarily fixing the

01-Hall_Ch-01.indd 7 18/04/2013 12:23:50 PM

 Representation

relationships between our conceptual system and our linguistic systems (remember, ‘linguistic’ in a
broad sense), codes make it possible for us to speak and to hear intelligibly, and establish the translat-
ability between our concepts and our languages which enables meaning to pass from speaker to hearer
and be effectively communicated within a culture. This translatability is not given by nature or fixed
by the gods. It is the result of a set of social conventions. It is fixed socially, fixed in culture. English
or French or Hindi speakers have, over time, and without conscious decision or choice, come to an
unwritten agreement, a sort of unwritten cultural covenant that, in their various languages, certain
signs will stand for or represent certain concepts. This is what children learn, and how they become
not simply biological individuals but cultural subjects. They learn the system and conventions of
representation, the codes of their language and culture, which equip them with cultural ‘know-how’,
enabling them to function as culturally competent subjects. Not because such knowledge is imprinted
in their genes, but because they learn its conventions and so gradually become ‘cultured persons’ – i.e.
members of their culture. They unconsciously internalize the codes which allow them to express
certain concepts and ideas through their systems of representation – writing, speech, gesture, visuali-
zation, and so on – and to interpret ideas which are communicated to them using the same systems.
You may find it easier to understand, now, why meaning, language and representation are such
critical elements in the study of culture. To belong to a culture is to belong to roughly the same con-
ceptual and linguistic universe, to know how concepts and ideas translate into different languages, and
how language can be interpreted to refer to or reference the world. To share these things is to see the
world from within the same conceptual map and to make sense of it through the same language sys-
tems. Early anthropologists of language, like Sapir and Whorf, took this insight to its logical extreme
when they argued that we are all, as it were, locked into our cultural perspectives or ‘mind-sets’, and
that language is the best clue we have to that conceptual universe. This observation, when applied to
all human cultures, lies at the root of what, today, we may think of as cultural or linguistic relativism.

ACTIVITY 2
You might like to think further about this question of how different cultures conceptually classify the world
and what implications this has for meaning and representation.
The English make a rather simple distinction between sleet and snow. The Inuit (Eskimos) who have to
survive in a very different, more extreme and hostile climate, apparently have many more words for snow
and snowy weather. Consider the list of Inuit terms for snow from the Scott Polar Research Institute in
Table 1.1. There are many more than in English, making much finer and more complex distinctions. The
Inuit have a complex classificatory conceptual system for the weather compared with the English. The
novelist Peter Hoeg, for example, writing about Greenland in his novel, Miss Smilla’s Feeling For Snow
(1994, pp. 5–6), graphically describes ‘frazzil ice’ which is ‘kneaded together into a soapy mash called
porridge ice, which gradually forms free-floating plates, pancake ice, which one, cold, noonday hour, on a
Sunday, freezes into a single solid sheet’. Such distinctions are too fine and elaborate even for the English
who are always talking about the weather! The question, however, is – do the Inuit actually experience
snow differently from the English? Their language system suggests they conceptualize the weather
differently. But how far is our experience actually bounded by our linguistic and conceptual universe?

01-Hall_Ch-01.indd 8 18/04/2013 12:23:50 PM

 The Work of Representation

TABLE 1.1 Inuit terms for snow and ice

snow ice siku

blowing — piqtuluk — pan, broken — siqumniq

is snowstorming piqtuluktuq — water immiugaq
falling — qanik melts — to make water immiuqtuaq
— is falling; — is snowing qaniktuq candle — illauyiniq
light falling — qaniaraq flat — qaimiq
light — is falling qaniaraqtuq glare — quasaq
first layer of — in fall apilraun piled — ivunrit
deep soft — mauya rough — iwuit
packed — to make water aniu shore — tugiu
light soft — aquluraq shorefast — tuvaq
sugar— pukak slush — quna
waterlogged, mushy — masak young — sikuliaq
— is turning into masak masaguqtuaq
watery — maqayak
wet — misak
wet falling — qanikkuk
wet — is falling qanikkuktuq
— drifting along a surface natiruvik
— is drifting along a surface natiruviktuaq
— lying on a surface apun
snowflake qanik
is being drifted over with — apiyuaq

One implication of this argument about cultural codes is that, if meaning is the result, not of
something fixed out there, in nature, but of our social, cultural and linguistic conventions, then
meaning can never be finally fixed. We can all ‘agree’ to allow words to carry somewhat different
meanings – as we have, for example, with the word ‘gay’, or the use, by young people, of the word
‘wicked’ as a term of approval. Of course, there must be some fixing of meaning in language or we
would never be able to understand one another. We can’t get up one morning and suddenly decide
to represent the concept of a ‘tree’ with the letters or the word VYXZ, and expect people to follow
what we are saying. On the other hand, there is no absolute or final fixing of meaning. Social and
linguistic conventions do change over time. In the language of modern managerialism, what we
used to call ‘students’, ‘clients’, ‘patients’ and ‘passengers’ have all become ‘customers’. Linguistic
codes vary significantly between one language and another. Many cultures do not have words for
concepts which are normal and widely acceptable to us. Words constantly go out of common usage,
and new phrases are coined: think, for example, of the use of ‘down-sizing’ to represent the process
of firms laying people off work. Even when the actual words remain stable, their connotations shift
or they acquire a different nuance. The problem is especially acute in translation. For example, does the

01-Hall_Ch-01.indd 9 18/04/2013 12:23:50 PM

 Representation

difference in English between know and understand correspond exactly to and capture exactly the
same conceptual distinction as the French make between savoir and connaître? Perhaps; but can
we be sure?
The main point is that meaning does not inhere in things, in the world. It is constructed, produced.
It is the result of a signifying practice – a practice that produces meaning, that makes things mean.

1.4 Theories of representation

There are, broadly speaking, three approaches to explaining how representation of meaning through
language works. We may call these the reflective, the intentional and the constructionist or con-
structivist approaches. You might think of each as an attempt to answer the questions, ‘Where do
meanings come from?’ and ‘How can we tell the “true” meaning of a word or image?’
In the reflective approach, meaning is thought to lie in the object, person, idea or event in
the real world, and language functions like a mirror, to reflect the true meaning as it already
exists in the world. As the poet Gertrude Stein once said, ‘A rose is a rose is a rose’. In the
fourth century BC, the Greeks used the notion of mimesis to explain how language, even draw-
ing and painting, mirrored or imitated nature; they thought of Homer’s great poem, The Iliad,
as ‘imitating’ a heroic series of events. So the theory which says that language works by simply
reflecting or imitating the truth that is already there and fixed in the world is sometimes called
‘mimetic’.
Of course there is a certain obvious truth to mimetic theories of representation and language. As
we’ve pointed out, visual signs do bear some relationship to the shape and texture of the objects
which they represent. But, as was also pointed out earlier, a two-dimensional visual image of a rose
is a sign – it should not be confused with the real plant with thorns and blooms growing in the garden.
Remember also that there are many words, sounds and images which we fully well understand but
which are entirely fictional or fantasy and refer to worlds which are wholly imaginary – including,
many people now think, most of The Iliad! Of course, I can use the word ‘rose’ to refer to real,
actual plants growing in a garden, as we have said before. But this is because I know the code
which links the concept with a particular word or image. I cannot think or speak or draw with
an actual rose. And if someone says to me that there is no such word as ‘rose’ for a plant in her
culture, the actual plant in the garden cannot resolve the failure of communication between us.
Within the conventions of the different language codes we are using, we are both right – and for
us to understand each other, one of us must learn the code linking the flower with the word for it
in the other’s culture.
The second approach to meaning in representation argues the opposite case. It holds that it is
the speaker, the author, who imposes his or her unique meaning on the world through language.
Words mean what the author intends they should mean. This is the intentional approach. Again,
there is some point to this argument since we all, as individuals, do use language to convey or
communicate things which are special or unique to us, to our way of seeing the world. However,

10

01-Hall_Ch-01.indd 10 18/04/2013 12:23:50 PM

 The Work of Representation

as a general theory of representation through language, the intentional approach is also flawed. We
cannot be the sole or unique source of meanings in language, since that would mean that we could
express ourselves in entirely private languages. But the essence of language is communication and
that, in turn, depends on shared linguistic conventions and shared codes. Language can never be
wholly a private game. Our private intended meanings, however personal to us, have to enter into
the rules, codes and conventions of language to be shared and understood. Language is a social sys-
tem through and through. This means that our private thoughts have to negotiate with all the other
meanings for words or images which have been stored in language which our use of the language
system will inevitably trigger into action.
The third approach recognizes this public, social character of language. It acknowledges that nei-
ther things in themselves nor the individual users of language can fix meaning in language. Things
don’t mean: we construct meaning, using representational systems – concepts and signs. Hence
it is called the constructivist or constructionist approach to meaning in language. According to
this approach, we must not confuse the material world, where things and people exist, and the
symbolic practices and processes through which representation, meaning and language operate.
Constructivists do not deny the existence of the material world. However, it is not the material
world which conveys meaning: it is the language system or whatever system we are using to repre-
sent our concepts. It is social actors who use the conceptual systems of their culture and the linguis-
tic and other representational systems to construct meaning, to make the world meaningful and to
communicate about that world meaningfully to others.
Of course, signs may also have a material dimension. Representational systems consist of the
actual sounds we make with our vocal chords, the images we make on light-sensitive paper with
cameras, the marks we make with paint on canvas, the digital impulses we transmit electroni-
cally. Representation is a practice, a kind of ‘work’, which uses material objects and effects.
But the meaning depends not on the material quality of the sign, but on its symbolic function. It
is because a particular sound or word stands for, symbolizes or represents a concept that it can
function, in language, as a sign and convey meaning – or, as the constructionists say, signify
(sign-i-fy).

1.5 The language of traffic lights

The simplest example of this point, which is critical for an understanding of how languages function
as representational systems, is the famous traffic lights example. A traffic light is a machine which
produces different coloured lights in sequence. The effect of light of different wavelengths on the
eye – which is a natural and material phenomenon – produces the sensation of different colours.
Now these things certainly do exist in the material world. But it is our culture which breaks the spec-
trum of light into different colours, distinguishes them from one another and attaches names – Red,
Green, Yellow, Blue – to them. We use a way of classifying the colour spectrum to create colours
which are different from one another. We represent or symbolize the different colours and classify

11

01-Hall_Ch-01.indd 11 18/04/2013 12:23:50 PM

 Representation

them according to different colour-concepts. This is the conceptual colour system of our culture.
We say ‘our culture’ because, of course, other cultures may divide the colour spectrum differently.
What’s more, they certainly use different actual words or letters to identify different colours: what
we call ‘red’, the French call ‘rouge’, and so on. This is the linguistic code – the one which cor-
relates certain words (signs) with certain colours (concepts), and thus enables us to communicate
about colours to other people, using ‘the language of colours’.
But how do we use this representational or symbolic system to regulate the traffic? Colours
do not have any ‘true’ or fixed meaning in that sense. Red does not mean ‘Stop’ in nature, any
more than Green means ‘Go’. In other settings, Red may stand for, symbolize or represent
‘Blood’ or ‘Danger’ or ‘Communism’; and Green may represent ‘Ireland’ or ‘The Countryside’
or ‘Environmentalism’. Even these meanings can change. In the ‘language of electric plugs’, Red
used to mean ‘the connection with the positive charge’ but this was arbitrarily and without expla-
nation changed to Brown! But then for many years the producers of plugs had to attach a slip of
paper telling people that the code or convention had changed, otherwise how would they know?
Red and Green work in the language of traffic lights because ‘Stop’ and ‘Go’ are the meanings
which have been assigned to them in our culture by the code or conventions governing this lan-
guage, and this code is widely known and almost universally obeyed in our culture and cultures
like ours – though we can well imagine other cultures which did not possess the code, in which
this language would be a complete mystery.
Let us stay with the example for a moment, to explore a little further how, according to the
constructionist approach to representation, colours and the ‘language of traffic lights’ work as a
signifying or representational system. Recall the two representational systems we spoke of earlier.
First, there is the conceptual map of colours in our culture – the way colours are distinguished from
one another, classified and arranged in our mental universe. Secondly, there are the ways words
or images are correlated with colours in our language – our linguistic colour codes. Actually, of
course, a language of colours consists of more than just the individual words for different points
on the colour spectrum. It also depends on how they function in relation to one another – the sorts
of things which are governed by grammar and syntax in written or spoken languages, which allow
us to express rather complex ideas. In the language of traffic lights, it is the sequence and position
of the colours, as well as the colours themselves, which enable them to carry meaning and thus
function as signs.
Does it matter which colours we use? No, the constructionists argue. This is because what
signifies is not the colours themselves but (a) the fact that they are different and can be distin-
guished from one another; and (b) the fact that they are organized into a particular sequence –
Red followed by Green, with sometimes a warning Amber in between which says, in effect,
‘Get ready! Lights about to change.’ Constructionists put this point in the following way. What
signifies, what carries meaning – they argue – is not each colour in itself nor even the concept
or word for it. It is the difference between Red and Green which signifies. This is a very impor-
tant principle, in general, about representation and meaning, and we shall return to it on more
than one occasion in the chapters that follow. Think about it in these terms. If you couldn’t

12

01-Hall_Ch-01.indd 12 18/04/2013 12:23:50 PM

 The Work of Representation

differentiate between Red and Green, you couldn’t use one to mean ‘Stop’ and the other to
mean ‘Go’. In the same way, it is only the difference between the letters P and T which enable
the word SHEEP to be linked, in the English language code, to the concept of ‘the animal with
four legs and a woolly coat’, and the word SHEET to ‘the material we use to cover ourselves
in bed at night’.
In principle, any combination of colours – like any collection of letters in written language or of
sounds in spoken language – would do, provided they are sufficiently different not to be confused.
Constructionists express this idea by saying that all signs are arbitrary. ‘Arbitrary’ means that there
is no natural relationship between the sign and its meaning or concept. Since Red only means ‘Stop’
because that is how the code works, in principle any colour would do, including Green. It is the code
that fixes the meaning, not the colour itself. This also has wider implications for the theory of rep-
resentation and meaning in language. It means that signs themselves cannot fix meaning. Instead,
meaning depends on the relation between a sign and a concept which is fixed by a code. Meaning,
the constructionists would say, is ‘relational’.

ACTIVITY 3
Why not test this point about the arbitrary nature of the sign and the importance of the code for yourself?
Construct a code to govern the movement of traffic using two different colours – Yellow and Blue – as in
the following:
When the yellow light is showing ...
Now add an instruction allowing pedestrians and cyclists only to cross, using Pink.

Provided the code tells us clearly how to read or interpret each colour, and everyone agrees to
interpret them in this way, any colour will do. These are just colours, just as the word SHEEP is
simply a jumble of letters. In French the same animal is referred to using the very different linguistic
sign, MOUTON. Signs are arbitrary. Their meanings are fixed by codes.
As we said earlier, traffic lights are machines, and colours are the material effect of light waves
on the retina of the eye. But objects – things – can also function as signs, provided they have been
assigned a concept and meaning within our cultural and linguistic codes. As signs, they work sym-
bolically – they represent concepts, and signify. Their effects, however, are felt in the material and
social world. Red and Green function in the language of traffic lights as signs, but they have real
material and social effects. They regulate the social behaviour of drivers and, without them, there
would be many more traffic accidents at road intersections.

1.6 Summary
We have come a long way in exploring the nature of representation. It is time to summarize what we
have learned about the constructionist approach to representation through language.

13

01-Hall_Ch-01.indd 13 18/04/2013 12:23:50 PM

 Representation

Representation is the production of meaning through language. In representation, construction-

ists argue, we use signs, organized into languages of different kinds, to communicate meaning-
fully with others. Languages can use signs to symbolize, stand for or reference objects, people
and events in the so-called ‘real’ world. But they can also reference imaginary things and fantasy
worlds or abstract ideas which are not in any obvious sense part of our material world. There is
no simple relationship of reflection, imitation or one-to-one correspondence between language
and the real world. The world is not accurately or otherwise reflected in the mirror of language.
Language does not work like a mirror. Meaning is produced within language, in and through vari-
ous representational systems which, for convenience, we call ‘languages’. Meaning is produced
by the practice, the ‘work’, of representation. It is constructed through signifying – i.e. meaning-
producing – practices.
How does this take place? In fact, it depends on two different but related systems of rep-
resentation. First, the concepts which are formed in the mind function as a system of mental
representation which classifies and organizes the world into meaningful categories. If we have a
concept for something, we can say we know its ‘meaning’. But we cannot communicate this mean-
ing without a second system of representation, a language. Language consists of signs organized
into various relationships. But signs can only convey meaning if we possess codes which allow
us to translate our concept into language – and vice versa. These codes are crucial for meaning
and representation. They do not exist in nature but are the result of social conventions. They are
a crucial part of our culture – our shared ‘maps of meaning’ – which we learn and unconsciously
internalize as we become members of our culture. This constructionist approach to language thus
introduces the symbolic domain of life, where words and things function as signs, into the very
heart of social life itself.

ACTIVITY 4
All this may seem rather abstract. But we can quickly demonstrate its relevance by an example from
painting.
Look at the painting of a still life by the Spanish painter, Juan Sanchez Cotán (1521–1627), entitled
Quince, Cabbage, Melon and Cucumber (Figure 1.3). It seems as if the painter has made every effort to
use the ‘language of painting’ accurately to reflect these four objects, to capture or ‘imitate nature’. Is
this, then, an example of a reflective or mimetic form of representation – a painting reflecting the ‘true
meaning’ of what already exists in Cotán’s kitchen? Or can we find the operation of certain codes, the
language of painting used to produce a certain meaning? Start with the question, what does the painting
mean to you? What is it ‘saying’? Then go on to ask, how is it saying it – how does representation work
in this painting?
Write down any thoughts at all that come to you on looking at the painting. What do these objects say
to you? What meanings do they trigger off?

14

01-Hall_Ch-01.indd 14 18/04/2013 12:23:50 PM

 The Work of Representation

FIGURE 1.3 Juan Cotán, Quince, Cabbage, Melon and Cucumber, c. 1602

READING A
Now read the edited extract from an analysis of the still life by the art critic and theorist, Norman Bryson,
included as Reading A at the end of this chapter. Don’t be concerned, at this stage, if the language seems
a little difficult and you don’t understand all the terms. Pick out the main points about the way representa-
tion works in the painting, according to Bryson.
Bryson is by no means the only critic of Cotán’s painting, and certainly doesn’t provide the only ‘correct’
reading of it. That’s not the point. The point of the example is that he helps us to see how, even in a still
life, the ‘language of painting’ does not function simply to reflect or imitate a meaning which is already
there in nature, but to produce meanings. The act of painting is a signifying practice. Take note, in par-
ticular, of what Bryson says about the following points:

1 The way the painting invites you, the viewer, to look – what he calls its ‘mode of seeing’; in part, the
function of the language is to position you, the viewer, in a certain relation to meaning.
2 The relationship to food which is posed by the painting.
3 How, according to Bryson, ‘mathematical form’ is used by Cotán to distort the painting so as to
bring out a particular meaning. Can a distorted meaning in painting be ‘true’?
4 The meaning of the difference between ‘creatural’ and ‘geometric’ space: the language of painting
creates its own kind of space.

If necessary, work through the extract again, picking up these specific points.

15

01-Hall_Ch-01.indd 15 18/04/2013 12:23:51 PM

 Representation

2 SAUSSURE’S LEGACY
The social constructionist view of language and representation which we have been discussing
owes a great deal to the work and influence of the Swiss linguist, Saussure, who was born in Geneva
in 1857, did much of his work in Paris, and died in 1913. He is known as the ‘father of modern
linguistics’. For our purposes, his importance lies not in his detailed work in linguistics, but in his
general view of representation and the way his model of language shaped the semiotic approach to
the problem of representation in a wide variety of cultural fields. You will recognize much about
Saussure’s thinking from what we have already said about the constructionist approach.
For Saussure, according to Jonathan Culler (1976, p. 19), the production of meaning depends on
language: ‘Language is a system of signs.’ Sounds, images, written words, paintings, photographs,
etc. function as signs within language ‘only when they serve to express or communicate ideas. ...
[To] communicate ideas, they must be part of a system of conventions ...’ (ibid.). Material objects
can function as signs and communicate meaning too, as we saw from the ‘language of traffic lights’
example. In an important move, Saussure analysed the sign into two further elements. There was,
he argued, the form (the actual word, image, photo, etc.), and there was the idea or concept in your
head with which the form was associated. Saussure called the first element, the signifier, and the
second element – the corresponding concept it triggered off in your head – the signified. Every time
you hear or read or see the signifier (e.g. the word or image of a Walkman, for example), it correlates
with the signified (the concept of a portable cassette-player in your head). Both are required to pro-
duce meaning but it is the relation between them, fixed by our cultural and linguistic codes, which
sustains representation. Thus, ‘the sign is the union of a form which signifies (signifier) ... and an
idea signified (signified). Though we may speak ... as if they are separate entities, they exist only as
components of the sign ... [which is] the central fact of language’ (Culler, 1976, p. 19).
Saussure also insisted on what in section 1 we called the arbitrary nature of the sign: ‘There is
no natural or inevitable link between the signifier and the signified’ (ibid.). Signs do not possess
a fixed or essential meaning. What signifies, according to Saussure, is not RED or the essence of
‘red-ness’, but the difference between RED and GREEN. Signs, Saussure argued, ‘are members of
a system and are defined in relation to the other members of that system’. For example, it is hard
to define the meaning of FATHER except in relation to, and in terms of its difference from, other
kinship terms, like MOTHER, DAUGHTER, SON, and so on.
This marking of difference within language is fundamental to the production of meaning, accord-
ing to Saussure. Even at a simple level (to repeat an earlier example), we must be able to distin-
guish, within language, between SHEEP and SHEET, before we can link one of those words to the
concept of an animal that produces wool, and the other to the concept of a cloth that covers a bed.
The simplest way of marking difference is, of course, by means of a binary opposition – in this
example, all the letters are the same except P and T. Similarly, the meaning of a concept or word
is often defined in relation to its direct opposite – as in night/day. Later critics of Saussure were to
observe that binaries (e.g. black/white) are only one, rather simplistic, way of establishing differ-
ence. As well as the stark difference between black and white, there are also the many other, subtler

16

01-Hall_Ch-01.indd 16 18/04/2013 12:23:51 PM

 The Work of Representation

differences between black and dark grey, dark grey and light grey, grey and cream and off-white,
off-white and brilliant white, just as there are between night, dawn, daylight, noon, dusk, and so on.
However, his attention to binary oppositions brought Saussure to the revolutionary proposition that a
language consists of signifiers, but in order to produce meaning, the signifiers have to be organized
into ‘a system of differences’. It is the differences between signifiers which signify.
Furthermore, the relation between the signifier and the signified, which is fixed by our cultural
codes, is not – Saussure argued – permanently fixed. Words shift their meanings. The concepts
(signifieds) to which they refer also change, historically, and every shift alters the conceptual map
of the culture, leading different cultures, at different historical moments, to classify and think about
the world differently. For many centuries, western societies have associated the word BLACK with
everything that is dark, evil, forbidding, devilish, dangerous and sinful. And yet, think of how the
perception of black people in America in the 1960s changed after the phrase ‘Black is Beautiful’
became a popular slogan – where the signifier, BLACK, was made to signify the exact opposite
meaning (signified) to its previous associations. In Saussure’s terms, ‘Language sets up an arbitrary
relation between signifiers of its own choosing on the one hand, and signifieds of its own choosing
on the other. Not only does each language produce a different set of signifiers, articulating and dividing
the continuum of sound (or writing or drawing or photography) in a distinctive way; each language
produces a different set of signifieds; it has a distinctive and thus arbitrary way of organizing the
world into concepts and categories’ (Culler, 1976, p. 23).
The implications of this argument are very far-reaching for a theory of representation and for our
understanding of culture. If the relationship between a signifier and its signified is the result of a sys-
tem of social conventions specific to each society and to specific historical moments, then all mean-
ings are produced within history and culture. They can never be finally fixed but are always subject
to change, both from one cultural context and from one period to another. There is thus no single,
unchanging, universal ‘true meaning’. ‘Because it is arbitrary, the sign is totally subject to history
and the combination at the particular moment of a given signifier and signified is a contingent result
of the historical process’ (Culler, 1976, p. 36). This opens up meaning and representation, in a radical
way, to history and change. It is true that Saussure himself focused exclusively on the state of the
language system at one moment of time rather than looking at linguistic change over time. However,
for our purposes, the important point is the way this approach to language unfixes meaning, breaking
any natural and inevitable tie between signifier and signified. This opens representation to the con-
stant ‘play’ or slippage of meaning, to the constant production of new meanings, new interpretations.
However, if meaning changes, historically, and is never finally fixed, then it follows that ‘taking
the meaning’ must involve an active process of interpretation. Meaning has to be actively ‘read’
or ‘interpreted’. Consequently, there is a necessary and inevitable imprecision about language. The
meaning we take, as viewers, readers or audiences, is never exactly the meaning which has been
given by the speaker or writer or by other viewers. And since, in order to say something meaning-
ful, we have to ‘enter language’, where all sorts of older meanings which pre-date us, are already
stored from previous eras, we can never cleanse language completely, screening out all the other,
hidden meanings which might modify or distort what we want to say. For example, we can’t entirely

17

01-Hall_Ch-01.indd 17 18/04/2013 12:23:51 PM

 Representation

prevent some of the negative connotations of the word BLACK from returning to mind when we
read a headline like, ‘WEDNESDAY – A BLACK DAY ON THE STOCK EXCHANGE’, even if
this was not intended. There is a constant sliding of meaning in all interpretation, a margin – some-
thing in excess of what we intend to say – in which other meanings overshadow the statement or
the text, where other associations are awakened to life, giving what we say a different twist. So
interpretation becomes an essential aspect of the process by which meaning is given and taken. The
reader is as important as the writer in the production of meaning. Every signifier given or encoded
with meaning has to be meaningfully interpreted or decoded by the receiver (Hall, 1980). Signs
which have not been intelligibly received and interpreted are not, in any useful sense, ‘meaningful’.

2.1 The social part of language

Saussure divided language into two parts. The first consisted of the general rules and codes of the
linguistic system, which all its users must share, if it is to be of use as a means of communication.
The rules are the principles which we learn when we learn a language and they enable us to use
language to say whatever we want. For example, in English, the preferred word order is subject–verb–
object (‘the cat sat on the mat’), whereas in Latin, the verb usually comes at the end. Saussure
called this underlying rule-governed structure of language, which enables us to produce well-
formed sentences, the langue (the language system). The second part consisted of the particular
acts of speaking or writing or drawing, which – using the structure and rules of the langue – are
produced by an actual speaker or writer. He called this parole. ‘La langue is the system of language,
the language as a system of forms, whereas parole is actual speech [or writing], the speech acts
which are made possible by the language’ (Culler, 1976, p. 29).
For Saussure, the underlying structure of rules and codes (langue) was the social part of lan-
guage, the part which could be studied with the law-like precision of a science because of its closed,
limited nature. It was his preference for studying language at this level of its ‘deep structure’ which
made people call Saussure and his model of language, structuralist. The second part of language,
the individual speech-act or utterance (parole), he regarded as the ‘surface’ of language. There were
an infinite number of such possible utterances. Hence, parole inevitably lacked those structural
properties – forming a closed and limited set – which would have enabled us to study it ‘scien-
tifically’. What made Saussure’s model appeal to many later scholars was the fact that the closed,
structured character of language at the level of its rules and laws, which, according to Saussure,
enabled it to be studied scientifically, was combined with the capacity to be free and unpredictably
creative in our actual speech acts. They believed he had offered them, at last, a scientific approach
to that least scientific object of inquiry – culture.
In separating the social part of language (langue) from the individual act of communication
(parole), Saussure broke with our common-sense notion of how language works. Our common-
sense intuition is that language comes from within us – from the individual speaker or writer; that
it is this speaking or writing subject who is the author or originator of meaning. This is what we
called, earlier, the intentional model of representation. But according to Saussure’s schema, each

18

01-Hall_Ch-01.indd 18 18/04/2013 12:23:51 PM

 The Work of Representation

authored statement only becomes possible because the ‘author’ shares with other language-users
the common rules and codes of the language system – the langue – which allows them to com-
municate with each other meaningfully. The author decides what she wants to say. But she cannot
‘decide’ whether or not to use the rules of language, if she wants to be understood. We are born into
a language, its codes and its meanings. Language is therefore, for Saussure, a social phenomenon.
It cannot be an individual matter because we cannot make up the rules of language individually, for
ourselves. Their source lies in society, in the culture, in our shared cultural codes, in the language
system – not in nature or in the individual subject.
We will move on in section 3 to consider how the constructionist approach to representation, and
in particular Saussure’s linguistic model, was applied to a wider set of cultural objects and practices,
and evolved into the semiotic method which so influenced the field. First, we ought to take account
of some of the criticisms levelled at his position.

2.2 Critique of Saussure’s model

Saussure’s great achievement was to force us to focus on language itself, as a social fact; on the pro-
cess of representation itself; on how language actually works and the role it plays in the production
of meaning. In doing so, he saved language from the status of a mere transparent medium between
things and meaning. He showed, instead, that representation was a practice. However, in his own
work, he tended to focus almost exclusively on the two aspects of the sign – signifier and signified.
He gave little or no attention to how this relation between signifier/signified could serve the purpose
of what earlier we called reference – i.e. referring us to the world of things, people and events outside
language in the ‘real’ world. Later linguists made a distinction between, say, the meaning of the word
BOOK and the use of the word to refer to a specific book lying before us on the table. The linguist,
Charles Sanders Pierce, while adopting a similar approach to Saussure, paid greater attention to the
relationship between signifiers/signifieds and what he called their referents. What Saussure called
signification really involves both meaning and reference, but he focused mainly on the former.
Another problem is that Saussure tended to focus on the formal aspects of language – how
language actually works. This has the great advantage of making us examine representation as a
practice worthy of detailed study in its own right. It forces us to look at language for itself, and
not just as an empty, transparent, ‘window on the world’. However, Saussure’s focus on language
may have been too exclusive. The attention to its formal aspects did divert attention away from the
more interactive and dialogic features of language – language as it is actually used, as it functions
in actual situations, in dialogue between different kinds of speakers. It is thus not surprising that,
for Saussure, questions of power in language – for example, between speakers of different status
and positions – did not arise.
As has often been the case, the ‘scientific’ dream which lay behind the structuralist impulse of his
work, though influential in alerting us to certain aspects of how language works, proved to be illusory.
Language is not an object which can be studied with the law-like precision of a science. Later cul-
tural theorists learned from Saussure’s ‘structuralism’ but abandoned its scientific premise. Language

19

01-Hall_Ch-01.indd 19 18/04/2013 12:23:51 PM

 Representation

remains rule-governed. But it is not a ‘closed’ system which can be reduced to its formal elements.
Since it is constantly changing, it is by definition open-ended. Meaning continues to be produced
through language in forms which can never be predicted beforehand and its ‘sliding’, as we described
it above, cannot be halted. Saussure may have been tempted to the former view because, like a good
structuralist, he tended to study the state of the language system at one moment, as if it had stood still,
and he could halt the flow of language-change. Nevertheless it is the case that many of those who
have been most influenced by Saussure’s radical break with all reflective and intentional models of
representation, have built on his work not by imitating his scientific and ‘structuralist’ approach, but by
applying his model in a much looser, more open-ended – i.e. ‘post-structuralist’ – way.

2.3 Summary
How far, then, have we come in our discussion of theories of representation? We began by contrasting
three different approaches. The reflective or mimetic approach proposed a direct and transparent rela-
tionship of imitation or reflection between words (signs) and things. The intentional theory reduced
representation to the intentions of its author or subject. The constructionist theory proposed a complex
and mediated relationship between things in the world, our concepts in thought and language. We have
focused at greatest length on this approach. The correlations between these levels – the material, the
conceptual and the signifying – are governed by our cultural and linguistic codes and it is this set of
interconnections which produces meaning. We then showed how much this general model of how sys-
tems of representation work in the production of meaning owed to the work of Ferdinand de Saussure.
Here, the key point was the link provided by the codes between the forms of expression used by lan-
guage (whether speech, writing, drawing, or other types of representation) – which Saussure called
the signifiers – and the mental concepts associated with them – the signifieds. The connection between
these two systems of representation produced signs; and signs, organized into languages, produced
meanings, and could be used to reference objects, people and events in the ‘real’ world.

3 FROM LANGUAGE TO CULTURE: LINGUISTICS TO

SEMIOTICS
Saussure’s main contribution was to the study of linguistics in a narrow sense. However, since his
death, his theories have been widely deployed, as a foundation for a general approach to language
and meaning, providing a model of representation which has been applied to a wide range of cultural
objects and practices. Saussure himself foresaw this possibility in his famous lecture notes, collected
posthumously by his students as the Course in General Linguistics (1960), where he looked forward
to ‘A science that studies the life of signs within society ... I shall call it semiology, from the Greek
semeion “signs” ’ (p. 16). This general approach to the study of signs in culture, and of culture as a
sort of ‘language’, which Saussure foreshadowed, is now generally known by the term semiotics.

20

01-Hall_Ch-01.indd 20 18/04/2013 12:23:51 PM

 The Work of Representation

FIGURE 1.4 Wrestling as a language of ‘excess’

The underlying argument behind the semiotic approach is that, since all cultural objects convey
meaning, and all cultural practices depend on meaning, they must make use of signs; and in so
far as they do, they must work like language works, and be amenable to an analysis which basi-
cally makes use of Saussure’s linguistic concepts (e.g. the signifier/signified and langue/parole
distinctions, his idea of underlying codes and structures, and the arbitrary nature of the sign). Thus,
when in his collection of essays, Mythologies (1972), the French critic, Roland Barthes, studied
‘The world of wrestling’, ‘Soap powders and detergents’, ‘The face of Greta Garbo’ or The Blue
Guides to Europe’, he brought a semiotic approach to bear on ‘reading’ popular culture, treating
these activities and objects as signs, as a language through which meaning is communicated. For
example, most of us would think of a wrestling match as a competitive game or sport designed for
one wrestler to gain victory over an opponent. Barthes, however, asks, not ‘Who won?’ but ‘What
is the meaning of this event?’ He treats it as a text to be read. He ‘reads’ the exaggerated gestures of
wrestlers as a grandiloquent language of what he calls the pure spectacle of excess.

READING B
You should now read the brief extract from Barthes’s ‘reading’ of ‘The world of wrestling’, provided as
Reading B at the end of this chapter.

21

01-Hall_Ch-01.indd 21 18/04/2013 12:23:51 PM

 Representation

In much the same way, the French anthropologist, Claude Lévi-Strauss, studied the customs,
rituals, totemic objects, designs, myths and folk-tales of so-called ‘primitive’ peoples in Brazil,
not by analysing how these things were produced and used in the context of daily life among the
Amazonian peoples, but in terms of what they were trying to ‘say’, what messages about the culture
they communicated. He analysed their meaning, not by interpreting their content, but by looking at
the underlying rules and codes through which such objects or practices produced meaning and, in
doing so, he was making a classic Saussurean or structuralist ‘move’, from the paroles of a culture
to the underlying structure, its langue. To undertake this kind of work, in studying the meaning of
a television programme like EastEnders, for example, we would have to treat the pictures on the
screen as signifiers, and use the code of the television soap opera as a genre, to discover how each
image on the screen made use of these rules to ‘say something’ (signifieds) which the viewer could
‘read’ or interpret within the formal framework of a particular kind of television narrative (see the
discussion and analysis of TV soap operas in Chapter 6).
In the semiotic approach, not only words and images but objects themselves can function as signi-
fiers in the production of meaning. Clothes, for example, may have a simple physical function – to
cover the body and protect it from the weather. But clothes also double up as signs. They construct a
meaning and carry a message. An evening dress may signify ‘elegance’; a bow tie and tails, ‘formal-
ity’; jeans and trainers, ‘casual dress’; a certain kind of sweater in the right setting, ‘a long, romantic,
autumn walk in the wood’ (Barthes, 1967). These signs enable clothes to convey meaning and to
function like a language – ‘the language of fashion’. How do they do this?

ACTIVITY 5
Look at the example of clothes in a magazine fashion spread (Figure 1.5). Apply Saussure’s model
to analyse what the clothes are ‘saying’? How would you decode their message? In particular, which
elements are operating as signifiers and what concepts – signifieds – are you applying to them? Don’t just
get an overall impression – work it out in detail. How is the ‘language of fashion’ working in this example?

The clothes themselves are the signifiers. The fashion code in western consumer cultures like ours
correlates particular kinds or combinations of clothing with certain concepts (‘elegance’, ‘formal-
ity’, ‘casualness’, ‘romance’). These are the signifieds. This coding converts the clothes into signs,
which can then be read as a language. In the language of fashion, the signifiers are arranged in a cer-
tain sequence, in certain relations to one another. Relations may be of similarity – certain items ‘go
together’ (e.g. casual shoes with jeans). Differences are also marked – no leather belts with evening
wear. Some signs actually create meaning by exploiting ‘difference’: e.g. Doc Marten boots with a
flowing long skirt. These bits of clothing ‘say something’ – they convey meaning. Of course, not
everybody reads fashion in the same way. There are differences of gender, age, class, race. But all
those who share the same fashion code will interpret the signs in roughly the same ways. ‘Oh, jeans
don’t look right for that event. It’s a formal occasion – it demands something more elegant.’
You may have noticed that, in this example, we have moved from the very narrow linguistic
level from which we drew examples in the first section, to a wider, cultural level. Note, also, that

22

01-Hall_Ch-01.indd 22 18/04/2013 12:23:51 PM

 The Work of Representation

FIGURE 1.5 Advertisement for Gucci, in Vogue, September 1995

two linked operations are required to complete the representation process by which meaning is
produced. First, we need a basic code which links a particular piece of material which is cut and
sewn in a particular way (signifier) to our mental concept of it (signified) – say a particular cut of
material to our concept of ‘a dress’ or ‘jeans’. (Remember that only some cultures would ‘read’ the
signifier in this way, or indeed possess the concept of (i.e. have classified clothes into) ‘a dress’, as
different from ‘jeans’.) The combination of signifier and signified is what Saussure called a sign.
Then, having recognized the material as a dress, or as jeans, and produced a sign, we can progress
to a second, wider level, which links these signs to broader, cultural themes, concepts or meanings
– for example, an evening dress to ‘formality’ or ‘elegance’, jeans to ‘casualness’. Barthes called
the first, descriptive level, the level of denotation; the second level, that of connotation. Both, of
course, require the use of codes.
Denotation is the simple, basic, descriptive level, where consensus is wide and most people
would agree on the meaning (‘dress’, ‘jeans’). At the second level – connotation – these signi-
fiers, which we have been able to ‘decode’ at a simple level by using our conventional conceptual
classifications of dress to read their meaning, enter a wider, second kind of code – ‘the language
of fashion’ – which connects them to broader themes and meanings, linking them with what we
may call the wider semantic fields of our culture: ideas of ‘elegance’, ‘formality’, ‘casualness’ and

23

01-Hall_Ch-01.indd 23 18/04/2013 12:23:52 PM

 Representation

‘romance’. This second, wider meaning is no longer a descriptive level of obvious interpretation. Here
we are beginning to interpret the completed signs in terms of the wider realms of social ideology – the
general beliefs, conceptual frameworks and value systems of society. This second level of significa-
tion, Barthes suggests, is more ‘general, global and diffuse. … It deals with “fragments of an ideol-
ogy...”. These signifieds have a very close communication with culture, knowledge, history and it
is through them, so to speak, that the environmental world [of the culture] invades the system [of
representation]’ (Barthes, 1967, pp. 91–2).

3.1 Myth today

In his essay ‘Myth today’, in Mythologies, Barthes gives another example which helps us to see
exactly how representation is working at this second, broader cultural level. Visiting the barbers’ one
day, Barthes is shown a copy of the French magazine Paris Match, which has on its cover a picture
of ‘a young Negro in a French uniform saluting with his eyes uplifted, probably fixed on the fold of
the tricolour’ (the French flag) (1972, p. 116). At the first level, to get any meaning at all, we need to
decode each of the signifiers in the image into their appropriate concepts: e.g. a soldier, a uniform,
an arm raised, eyes lifted, a French flag. This yields a set of signs with a simple, literal message or
meaning: a black soldier is giving the French flag a salute (denotation). However, Barthes argues that
this image also has a wider, cultural meaning. If we ask, ‘What is Paris Match telling us by using this
picture of a black soldier saluting a French flag?’, Barthes suggests that we may come up with the
message: ‘that France is a great Empire, and that all her sons, without any colour discrimination,
faithfully serve under her flag, and that there is no better answer to the detractors of an alleged colo-
nialism than the zeal shown by this Negro in serving his so-called oppressors’ (connotation) (ibid.).
Whatever you think of the actual ‘message’ which Barthes finds, for a proper semiotic analysis
you must be able to outline precisely the different steps by which this broader meaning has been
produced. Barthes argues that here representation takes place through two separate but linked pro-
cesses. In the first, the signifiers (the elements of the image) and the signifieds (the concepts – soldier,
flag and so on) unite to form a sign with a simple denoted message: a black soldier is giving the
French flag a salute. At the second stage, this completed message or sign is linked to a second set
of signifieds – a broad, ideological theme about French colonialism. The first completed meaning
functions as the signifier in the second stage of the representation process and, when linked with
a wider theme by a reader, yields a second, more elaborate and ideologically framed message or
meaning. Barthes gives this second concept or theme a name – he calls it ‘a purposeful mixture
of “French imperiality” and “militariness”’. This, he says, adds up to a ‘message’ about French
colonialism and her faithful Negro soldier-sons. Barthes calls this second level of signification the
level of myth. In this reading, he adds, ‘French imperiality is the very drive behind the myth. The
concept reconstitutes a chain of causes and effects, motives and intentions. ... Through the concept
... a whole new history ... is implanted in the myth ... the concept of French imperiality ... is again
tied to the totality of the world: to the general history of France, to its colonial adventures, to its
present difficulties’ (Barthes, 1972, p. 119).

24

01-Hall_Ch-01.indd 24 18/04/2013 12:23:52 PM

 The Work of Representation

READING C
Turn to the short extract from ‘Myth today’ (Reading C at the end of this chapter), and read Barthes’s account
of how myth functions as a system of representation. Make sure you understand what Barthes means by
‘two staggered systems’ and by the idea that myth is a ‘meta-language’ (a second-order language).

For another example of this two-stage process of signification, we can turn now to another of
Barthes’s famous essays.

ACTIVITY 6
Now, look carefully at the advertisement for Panzani products (Figure 1.6) and, with Barthes’s analysis in
mind, do the following exercise:
1 What signifiers can you identify in the ad?
2 What do they mean? What are their signifieds?
3 Now, look at the ad as a whole, at the level of ‘myth’. What is its wider, cultural message or theme?
Can you construct one?

FIGURE 1.6 ‘Italian-ness’ and the Panzani ad

25

01-Hall_Ch-01.indd 25 18/04/2013 12:23:52 PM

 Representation

READING D
Now read the second extract from Barthes, in which he offers an interpretation of the Panzani ad for
spaghetti and vegetables in a string bag as a ‘myth’ about Italian national culture. The extract from
‘Rhetoric of the image’, in Image-Music-Text (1977), is included as Reading D at the end of this chapter.

Barthes suggests that we can read the Panzani ad as a ‘myth’ by linking its completed message (this
is a picture of some packets of pasta, a tin, a sachet, some tomatoes, onions, peppers, a mushroom,
all emerging from a half-open string bag) with the cultural theme or concept of Italianicity’ (or as we
would say, ‘Italian-ness’). Then, at the level of the myth or meta-language, the Panzani ad becomes
a message about the essential meaning of Italian-ness as a national culture. Can commodities really
become the signifiers for myths of nationality? Can you think of ads, in magazines or on television,
which work in the same way, drawing on the myth of ‘Englishness’? Or ‘Frenchness’? Or ‘American-
ness’? Or ‘Indian-ness’? Try to apply the idea of ‘Englishness’ to the ad reproduced as Figure 1.7.

FIGURE 1.7 An image of ‘Englishness’ – advertisement for Jaguar

4 DISCOURSE, POWER AND THE SUBJECT

What the examples above show is that the semiotic approach provides a method for analysing how
visual representations convey meaning. Already, in Roland Barthes’s work in the 1960s, as we have

26

01-Hall_Ch-01.indd 26 18/04/2013 12:23:53 PM