Semiotics The Basics - Daniel Chandler-2

Download as pdf or txt
Download as pdf or txt
You are on page 1of 40

INTRODUCTION

If you go into a bookshop and ask an assistant where to find a book


on semiotics, you are likely to meet with a blank look. Even worse,
you might be asked to define what semiotics is – which would be a
bit tricky if you were looking for a beginner’s guide. It’s worse still
if you do know a bit about semiotics because it can be hard to offer a
simple definition that is of much use in the bookshop. If you’ve ever
been in such a situation, you’ll probably agree that it’s wise not to ask.
Semiotics could be anywhere. The shortest definition is that it is the
study of signs. But that doesn’t leave enquirers much wiser. ‘What do
you mean by a sign?’ people usually ask next. The kinds of signs that
are likely to spring immediately to mind are those that we routinely
refer to as ‘signs’ in everyday life, such as road signs, pub signs, and
star signs. If you were to agree with them that semiotics can include
the study of all these and more, people will probably assume that semi-
otics is about ‘visual signs’. You would confirm their hunch if you said
that signs can also be drawings, paintings, and photographs, and by
now they’d be keen to direct you to the art and photography sections.
But if you are thick-skinned and tell them that it also includes words,
sounds, and ‘body language’, they may reasonably wonder what all
these things have in common and how anyone could possibly study
such disparate phenomena. If you get this far, they’ve probably already
‘read the signs’ that suggest that you are either eccentric or insane, and
communication may have ceased.

DOI: 10.4324/9781003155744-1
2 I N T R OD UC T I O N

DEFINITIONS
In semiotics, a sign is traditionally defined as something which ‘stands
for’ (or represents) something else. It can take any form – a word, an
image, a sound, an odour, a flavour, an action, an event, an object,
or whatever. Anything has the potential to be interpretable, but noth-
ing has an intrinsic meaning; for human beings, something becomes
a sign when it is interpreted as signifying something. Conventional
‘symbols’ are obvious examples (such as when a flag is used to ‘stand
for’ a particular country), but we also recognize ‘natural signs’, such as
when we infer that dark clouds are a ‘sign of’ rain. We experience the
world as meaningful. As a species we are above all meaning-makers
(Homo significans), driven to ‘find’ meanings ‘in’ things. The under-
lying metaphor that things are ‘containers’ of meanings exemplifies
the subtle potency of our foremost semiotic system – the language
that we speak. We have no direct, unmediated access to ‘things in
themselves’. We dwell in a symbolic world of human meanings; our
everyday reality is a web of signs. As we will see, communication,
culture, community, and cognition depend upon them.
Semiotics is concerned with how meanings are made and how
reality is represented (and indeed constructed) through signs, sign
systems, and processes of signification. Theories of signs (or sym-
bols) appear throughout the history of philosophy from ancient times
onwards. The early medieval theologian and philosopher Augustine
of Hippo (354–430 ce) is widely regarded as the founder of medieval
semiotics, but it wasn’t until 1690 that the English philosopher John
Locke (1632–1704) coined the term ‘semiotics’ to describe a field of
study that had yet to emerge.
The two primary traditions in contemporary semiotics stem respec-
tively from the Swiss linguist Ferdinand de Saussure (1857–1913)
and the American philosopher Charles Sanders Peirce (pronounced
‘purse’) (1839–1914). They are widely regarded as the co-founders
of what is now generally known as semiotics – despite the fact that
neither of them actually wrote a book on the subject. The first edition
of Saussure’s Course in General Linguistics, published posthumously
in 1916, contains the declaration that he could envisage, and staked
a claim for, ‘a science that studies the life of signs within society’,
which he called semiology, from the Greek sēmeîon, ‘sign’ (CLG
33/16: i.e. Cours de linguistique générale, page 33/Course in General
IN T RO D U C T I ON 3

Linguistics, page 16). His use of the term sémiologie dates originally
from a manuscript of 1894. Although Saussure was a linguist, he saw
linguistics as a branch of the ‘general science’ of semiology, which
was in turn an offshoot of (social) psychology. Across the Atlantic,
to the philosopher Charles Peirce the field of study which he calls
‘semeiotic’ (or ‘semiotic’) is the ‘formal doctrine of signs’ that is
closely related to logic (2.227). Working quite independently from
Saussure, Peirce borrowed his term from Locke. Saussure’s term
‘semiology’ is sometimes used to refer to the Saussurean tradition,
while the term ‘semiotics’ sometimes refers to the Peircean tradition
(see Figure 0.1; for a useful discussion, see Daylight 2014). However,
nowadays the term ‘semiotics’ is widely used as an umbrella term
to embrace the whole field (Nöth 1990, 14). We will outline and dis-
cuss both the Saussurean and Peircean models of the sign in the next
chapter.
Some commentators adopt a definition of semiotics by the American
philosopher Charles W. Morris (1901–79) as ‘the science of signs’
(1938, 1–2). The term ‘science’ (used also by Saussure) is misleading.
Semiotics is perhaps best thought of as a way of looking at the produc-
tion of meaning from a particular critical perspective. Its scope and
general principles are still a site of struggle, but semiotics continues to
expand into new areas. The Association Internationale de Sémiotique
(International Association of Semiotics) was founded in Paris in 1969,
and its journal Semiotica first appeared in the same year. The first
world congress of IASS (the International Association for Semiotic
Studies) was held in Milan in 1974. Semiotics is now served by mul-
tiple journals and regular conferences at national and international

F I G U RE 0 .1 Semiotic traditions
Source: © 2021 Daniel Chandler
4 I N T R OD UC T I O N

levels (see the appendix, ‘Going Further’). Scholars involved in the


field include linguists, philosophers, psychologists, sociologists,
anthropologists, literary, aesthetic, and media theorists, psychoana-
lysts, market researchers, and educationalists. Indeed, the impact of
semiotics within existing disciplines could be argued to be its greatest
contribution to the advancement of knowledge.

RELATION TO PHILOSOPHY AND LINGUISTICS


As the study of signification, semiotics is intrinsically transdis-
ciplinary, but its strongest disciplinary ties are to philosophy and
linguistics. Within both of those disciplines the study of meaning
is most closely associated with semantics, so the question often
arises: ‘What is the relationship between semiotics and seman-
tics?’ Charles Morris (1946, 217–19), one of the pioneers of modern
semiotics, defined it in 1938 as having three branches (Figure 0.2):
semantics (the meanings of signs), pragmatics (the use of signs), and
syntactics or syntax (the relations between signs). This framework
is still generally accepted as encompassing the scope of semiotics.
It features among the many branches of philosophy and also rep-
resents the principal branches of linguistics (along with phonetics/
phonology). However, no clear distinction can be made between
semantics and pragmatics because the meaning of signs cannot be
divorced from the context of use (a topic to be explored at length
in Chapter 6).

F I G U RE 0 .2 The relation of semiotics to philosophy and linguistics


Source: © 2021 Daniel Chandler
IN T RO D U C T I ON 5

The theory of signs is mentioned most often in the subdisciplines of


logic and theoretical linguistics. The first explicit reference to semiot-
ics, ‘the doctrine of signs’, as a branch of philosophy, appeared in John
Locke’s Essay Concerning Human Understanding (1690, IV.xxi.4).
Peirce refers to it as ‘another name for’ logic (CP 2.227: i.e. Collected
Papers vol. 2, para. 227). However, a broader link between semiot-
ics and philosophy is with the philosophy of language (Eco 1984), a
philosophical sub-discipline concerned primarily with meaning and
reference, and especially with relations between language, thought,
and the world. It shares with semiotics a concern with ways of meaning.
Within the philosophy of language, semiotics is also closely related
to hermeneutics – broadly, the theory of interpretation (particularly
the contextual interpretation of texts) – which is sometimes classi-
fied as part of semiotics (e.g. Honderich 1995, 937). Arguably, in their
shared concern with the pragmatics of meaning-making, semiotics
tends to be more system-oriented while hermeneutics tends to be more
process-oriented. From a semiotic perspective, the territory of the phi-
losophy of language forms only part of theoretical semiotics, which
concerns itself with all kinds of signs and sign systems, whether or not
they are linguistic.
Within the philosophical domain of aesthetics, or the philosophy of
art, ‘semiotics’ can also refer to a school of thought associated with
the American philosopher Nelson Goodman (1906–98), who argues
controversially that (contrary to traditional ‘resemblance theory’)
pictures are as conventional as linguistic signs (1968). Philosophical
issues in semiotic theory will be explored at length in Chapter 2 and
they will surface throughout in relation to issues of representation and
‘the social construction of reality’ through the mediation of signs and
sign systems. Points of contact between semiotics and the philoso-
phy of mind (primarily via phenomenology) will be encountered in
Chapters 6 and 7.

STRUCTURALISM
The development of the European tradition of semiology is closely
associated with structuralism – a transdisciplinary academic perspec-
tive that arose in the late 1920s and reached its zenith in the late 1960s
(when it began to transform into poststructuralism). Inspired primar-
ily by Ferdinand de Saussure’s conception of languages as systems of
6 I N T R OD UC T I O N

F I G URE 0 .3 The emergence of European structuralism


Source: © 2021 Daniel Chandler

relations, leading figures in structural linguistics include the Danish


linguist Louis Hjelmslev (1899–1966) and the Russian-born linguist
Roman Jakobson (1896–1982), who first coined the term ‘structural-
ism’ in 1929 (Jakobson 1990, 6). The structural paradigm in linguistics
provided the primary impetus for the broader structuralist ‘movement’,
key figures within which include the Frenchmen Claude Lévi-Strauss
(1908–2009) in anthropology (the standard-bearer of the movement),
Roland Barthes (1915–80) in literary criticism and cultural analysis,
and Jacques Lacan (1901–81) in psychoanalysis (Figure 0.3). Two
other Frenchmen, both philosophers – Michel Foucault (1926–84) and
Louis Althusser (1918–90) – are regularly classified as structuralists,
but both rejected the label (Althusser 1968, 7; Foucault 1969, 15), and
neither was inspired by structural linguistics, so they are untypical of
‘mainstream’ structuralism.
Structuralists seek to apply a linguistic model to a much wider range
of social phenomena. Jakobson writes that

Language is … a purely semiotic system … The study of signs, however,


… must take into consideration also applied semiotic structures,
IN T RO D U C T I ON 7

as for instance, architecture, dress, or cuisine … Any edifice is


simultaneously some sort of refuge and a certain kind of message.
Similarly, any garment responds to definitely utilitarian requirements
and at the same time exhibits various semiotic properties.
(Jakobson 1968a, 703)

He adds (ibid., 698) that semiotics ‘deals with those general princi-
ples which underlie the structure of all signs whatever’. Structuralists
search for ‘deep structures’ underlying the ‘surface features’ of sign
systems: Claude Lévi-Strauss in myth, kinship rules and totemism;
Jacques Lacan in the unconscious; Roland Barthes and Algirdas
Greimas in the ‘grammar’ of narrative. Julia Kristeva (1973, 1249)
declares that ‘what semiotics has discovered … is that the law govern-
ing or, if one prefers, the major constraint affecting any social practice
lies in the fact that it signifies; i.e. that it is articulated like a language’.
‘Although language is only one particular semiological system,’
Saussure sees it as the best example of a system based on the arbitrari-
ness of the sign – a principle we will explore shortly. He suggests that
linguistics has the potential to become ‘the master-pattern [le patron
general] for all branches of semiology’ (CLG 101/68). However,
Saussure does not limit all semiological systems to his model of the
language system [langue], as some of his critics claim (Thibault 1997,
21). Indeed, he subordinates linguistics to semiology as a discipline.
Subsequently, structuralist semiotics has drawn heavily on linguistic
concepts – partly because of his influence and also because linguistics
is a more established discipline than the study of other sign systems.
Jakobson (1970/1990, 455) insists that ‘language is the central and
most important among all human semiotic systems’, while Lévi-
Strauss notes that ‘language is the semiotic system par excellence; it
cannot but signify, and exists only through signification’ (1972, 48).
Furthermore, as the French linguist and semiotician Émile Benveniste
(1902–76) observes, ‘language is the interpreting system of all other
systems, linguistic and non-linguistic’ (1969, 239). Although sign sys-
tems can be non-verbal (as with signal flags, for instance), we cannot
read messages or specify meanings in such systems without recourse
to language (Chandler 2022).
As we have noted, Saussure saw linguistics as a branch of ‘semi-
ology’, subject to any general laws that might be discovered by the
new science that he envisaged. He consequently saw it as important
8 I N T R OD UC T I O N

to identify what language has in common with all other ‘systems


of expression’, such as rites and customs (CLG 35/16–17; 101/68).
Like Saussure, Jakobson was in no doubt that ‘language is a system
of signs, and linguistics is part and parcel of the science of signs or
semiotics’ (1949a, 50). However, it proved difficult to avoid adopting a
linguistic model in exploring other sign systems. Thus Barthes (1967b,
xi) declares that ‘perhaps we must invert Saussure’s formulation and
assert that semiology is a branch of linguistics’. The American linguist
Leonard Bloomfield (1939, 55) insists that ‘linguistics is the chief con-
tributor to semiotics’, and Jakobson (1963d, 289) defines semiotics as
‘the general science of signs which has as its basic discipline linguis-
tics, the science of verbal signs’.
Following Saussure’s example, Lévi-Strauss (1960, 17) argues that
anthropology is a branch of semiology. It ‘aims to be a semiological sci-
ence, and takes as a guiding principle that of “meaning” ’ (1972, 364).
Structuralist methods have subsequently been very widely employed
in the semiotic analysis of many cultural phenomena. This approach
has had considerable influence in contemporary cultural studies, where
films, television and radio programmes, advertising posters, and so on
are commonly referred to as ‘texts’ that require ‘reading’. It is a common
pedagogical strategy to undermine commonsense assumptions about
what features such media may have in common with a symbolic system
like language rather than with what we treat as reality in the everyday
world. However, there is a danger of overextending the linguistic meta-
phor and failing to address the affordances of different semiotic systems.
Semiotics concerns itself with all ways of meaning. Most contempo-
rary semioticians would argue that a linguistic model can, at best, only
ever form part of a general theory of signification that reaches beyond
intentional communication. Useful as it may be in the cultural domain,
semiotics cannot be limited to such a model.

WHY STUDY SEMIOTICS?


‘Why should we study semiotics?’ is a pressing question in part because
the writings of semioticians have a reputation for being dense with jar-
gon: one critic wryly remarks that ‘semiotics tells us things we already
know in a language we will never understand’ (Paddy Whannel, cited
in Seiter 1992, 31). The semiotic establishment may initially seem to
be a very exclusive club, but its concerns are not confined to members.
IN T RO D U C T I ON 9

This book is intended to be of particular value to readers who wish


to use semiotics as an approach to the analysis of texts and cultural
practices. However, even within the cultural domain, semiotics is
far more than a method of textual analysis. For instance, it involves
the philosophical exploration of issues of representation and reality.
Studying semiotics can make us less likely to take reality for granted
as something that is wholly independent of interpretive systems. It can
assist us to become more aware of the mediating role of signs and of
the roles played by ourselves and others in constructing social realities.
The technical concept of ‘mediation’ (from the medieval Latin medi-
are, ‘to be in the middle’) is ubiquitous in semiotics and deserves a brief
explanation at the outset. It is a fundamental principle of semiotic the-
ory that all human experience is mediated by signs (and that language
is a sign system that mediates thought and reality). Communication is
the clearest example of the crucial role of semiotic mediation between
the parties involved, who must draw upon a common stock of signs
and conventions for their use. This process applies not only to speak-
ers and listeners (or writers and readers), but also to the makers and
perceivers of any form of expression (such as a work of art).
More broadly, mediation refers to any framework or process which
‘intervenes’ in our perception of the external world and hence contrib-
utes to the construction of our conceptual and experiential worlds. For
instance, opening a door requires us to approach it with ‘preconcep-
tions’: we need to recognize that it belongs to the linguistic category
of ‘doors’ (based on what these are for and how they work). It also
requires us to relate it to the social system, for instance with refer-
ence to whether you own it, whether you are entitled to open it, and
so on. Such prior knowledge forms part of a complex interpretational
framework that mediates the experience and routinely guides our
expectations and behaviour (usually beyond our conscious awareness).
If a door is slammed in your face you may care to reflect that we are
never dealing with a purely physical phenomenon: someone was send-
ing you a message!
The semiotic principle of mediatedness represents a challenge to
‘naïve realism’ as well as to subjective idealism, since sign systems
depend on shared frames of reference. It is not incompatible with ‘crit-
ical realism’, but many cultural semioticians adopt an overtly ‘social
constructionist’ stance. Social constructionism does not entail deny-
ing the existence of ‘external reality’ but it does assume that our sign
10 I N T R OD UC T I O N

systems (especially language) play a major part in ‘the social construc-


tion of reality’ (or at least ‘the construction of social reality’). From
this perspective, signs are the intersubjective medium that enables
social and psychological realities to take shape – rather than ‘coming
between’ perceivers and an objective reality that ‘lies behind’ them.
Meanings are intersubjective to the extent that we share common
understandings of things. We develop such understandings through
our use of signs in the context of social interaction (embedded, of
course, in a material context). The most prominent intersubjective
framework is a shared language, but all forms of expression contribute
to this function.
Embarking on a voyage of semiotic discovery for the first time is
not for the faint-hearted. Some of the concepts encountered here may
initially seem daunting, but those that matter most are revisited from
different angles in subsequent chapters. Perhaps the greatest challenge
is that one must be prepared to entertain ideas that are often deeply
counter-intuitive, contrary to some of our own taken-for-granted
assumptions. For instance, it is a struggle to step back from the cate-
gorical systems built into our own mother tongue, so as not to mistake
them for ‘the way things are’. Semiotics demands relational thinking,
and it is not easy for any of us to focus on the relationships between
concepts rather than on the ‘things’ to which our words seem to refer.
Adopting a semiotic perspective helps us to realize that information
or meaning is not ‘contained’ in the world, or in books, computers, or
other media. Meaning is not ‘transmitted’ to us – we actively inter-
pret texts and the world according to a complex and dynamic interplay
of frames of reference. A semiotic perspective helps to make more
explicit the interpretive systems and textual conventions that ordinar-
ily retreat to transparency.
Socially oriented semiotics is a subversive activity. In defining
realities, sign systems serve ideological functions. A critical semiotic
perspective involves investigating the construction and maintenance
of reality by particular social groups. Such an approach has much in
common with that of symbolic interactionism in sociology and social
psychology (Sandstrom et al. 2010). Deconstructing and contesting
reality maintenance systems can reveal whose realities are privileged
and whose are suppressed. To decline the study of signs is to leave
to others the control of the world of meanings that we inhabit. The
semiotic mission is to take apart what is taken for granted. This is a
IN T RO D U C T I ON 11

particularly challenging quest because, as the Austrian philosopher and


sociologist Alfred Schutz emphasized, in everyday life we routinely
suspend doubt about the reality of the world (Schutz and Luckmann
1974). Phenomenologists call this the ‘natural attitude’, following the
German philosopher Edmund Husserl (1859–1938). Adopting a semi-
otic perspective requires the cultivation of a critical approach to our
own everyday experience – asking ourselves what we are taking for
granted in making sense of a particular situation (Zerubavel 2018).
Indeed, how do you define the ‘situation’ that you are ‘in’? Learning to
‘stand back’ and reflect like this is not easy, but becoming aware of the
processes of mediation involved in constructing the realities of every-
day life is both inherently fascinating and intellectually empowering.
If you are undeterred and ready for an intellectual adventure, read on.
1

MODELS

Signs are commonly defined in terms of a relation between form


and meaning. However, this simplistic formulation raises the issue
of ‘the meaning of meaning’. Formal models of the sign often dis-
tinguish between two kinds of meaning: conceptual meaning (sense
or designation) and referential meaning (reference or denotation). The
reference of a sign is what it refers to beyond the sign system (known
as a referent or an object). Referents can include things, beings, or
events (real or imaginary) but also more abstract categories. The term
sense is sometimes used interchangeably with ‘meaning’, but it is often
used more specifically to refer to the distinctions made within a lan-
guage. Dictionaries define the various linguistic senses of individual
words (many of which have no reference to things to which we can
point in the world). If we ask what is meant by the word ‘semioti-
cian’, its sense is ‘someone who studies signs’ (as distinct from say,
‘someone who paints signs’), while its reference could be to any of its
practitioners in the world. If someone asks what you do and you reply
(rashly), ‘I am a semiotician’, you have provided a reference but they
will probably be none the wiser; the next thing they will expect (opti-
mistically) is some sense.
The traditional definition of a sign as ‘something that stands for
something else’ is a medieval one (in the scholastic Latin formula,
aliquid stat pro aliquo). The distinction between signs and what they

DOI: 10.4324/9781003155744-2
14 M OD E LS

are signs of is fundamental in semiotics (as we will see in Chapter 2,


‘the sign is not the thing’), and the traditional formulation foregrounds
the relation between a sign vehicle and its referent. Such a referential
relation is a common feature in models of the sign, but it cannot con-
stitute a viable model in which this is the only relation. ‘The object of
a sign is one thing; its meaning is another’ (Peirce CP 5.6). A purely
referential model reduces meaning to reference (as if meaning resides
‘in’ the world). Indicating what we are talking about (for instance, by
pointing to something) is obviously important, but (to the frustration
of monolingual travellers) it is insufficient for establishing meaning.
Equating what a sign means with what it stands for is unhelpfully
circular.
However, such a dyadic model is implicitly triadic insofar as it
presupposes an interpreter (for whom the sign is meaningful). The
meaning of a sign is not ‘contained’ within it, but arises in its interpre-
tation. The ‘standing for’ sign–object relation requires interpretation
by a conscious being. ‘ “Being a sign of” is a three-term relation’ (Price
1969, 92). A sign stands for (or is a sign of) something, to someone. We
can hardly discuss human meaning-making without reference to the
mind. Augustine’s model (397 ce) is primarily referential (Nöth 1990,
85), but he does acknowledge this third dimension: ‘a sign is a thing
which, over and above the impression it makes on the senses, causes
something else to come into the mind’ (On Christian Doctrine II.1.i).
Although Augustine acknowledges the role of the mind, he does
not focus on the key distinction between sense and reference that
characterizes the classical triadic model of the sign (Figure 1.1). This
traditional semiotic triangle features both the relation of reference and
that of signification (sometimes termed, respectively, the sign–object
relation and the sign–mind relation). However, these dyadic relations
are subordinated to the triadic process of mediation (interpretation
being represented here with arrows): sense (or conceptual meaning)
mediates the referent. The broken line at the base of the triangle signi-
fies that there is not necessarily any direct relationship between the sign
vehicle and the referent (Ogden and Richards 1923, 11). Both Aristotle
(c. 350 bce) and the Stoic philosophers (c. 250 bce) developed vari-
ants of this triadic model, in which signs signify referents by means of
mediating concepts. According to Aristotle, who has been described
as ‘the first thinker to theorize in a systematic way about meaning and
reference’ (Putnam 1988, 19), we understand the meaning of a sign
MO D E L S 15

F I G U RE 1.1 The classical triadic model of the sign


Source: © 2021 Daniel Chandler

(such as a word) when we associate it with a concept – a representa-


tion in the mind – that determines what it refers to. Aristotle’s model
of meaning, advanced in On Interpretation (350 bce), dominated
European thinking for over two millennia.
As will become apparent, a great deal hangs on how we define a sign.
In the medieval ‘language of flowers’, the herb rosemary stands for
‘remembrance’, but it requires someone such as Shakespeare’s Ophelia
to interpret it as such: rosemary growing in the kitchen herb garden has
no such signification. However, just as meaning cannot be reduced to
something ‘in the world’, neither can it be reduced to something ‘in the
mind’. The Aristotelian cognitive model does not account for the social
grounding of meaning. Rosemary cannot stand for remembrance in the
absence of a socially shared code for the symbolism of flowers.
Our natural languages are our primary socially shared sign systems.
We begin our exploration of the most influential contemporary sign mod-
els with a semiotic approach to language that involves a radical challenge
to the traditional ‘standing for’ relation or ‘representational’ model, and to
our common sense assumptions about the language–world relationship.

THE SAUSSUREAN MODEL


While the concept of ‘the sign’ is ancient, the notion of sign systems is
a modern one. For the Swiss linguist Ferdinand de Saussure, a leading
contender for the title of ‘founder of modern linguistics’, language is
a system of signs, and linguistic signs make sense only as part of a
16 M OD E LS

F I G URE 1. 2 Saussure’s dyadic model of the linguistic sign


Source: Adapted from Saussure 1916/1995, 158

language’s sign system. Within such a system, a sign has two aspects,
which Saussure termed a signifiant (usually rendered in English as
a ‘signifier’) and a signifié (a ‘signified’) (see Figure 1.2). Although
in contemporary discourse the term signified is often used to refer
generally to ‘meaning’, and in loose usage may involve reference,
Saussure makes it very clear that he is not dealing with the dimension
of reference: ‘The linguistic sign unites, not a thing and a name, but
a concept and an acoustic image’ (CLG 98/66). Thus, for Saussure,
words do not ‘stand for’ things, and his signifier and signified are not
to be understood dualistically as ‘sign’ (vehicle) and ‘referent’ – a
common misinterpretation.
Within the Saussurean linguistic model, the sign is the unified
whole that results from the association of a sound with a concept (ibid.,
99/67). This is a relationship in which the two layers are as inseparable
as the two sides of a piece of paper (157/113). A linguistic sign could
not consist of sound without sense or of sense without sound (144/102–
3). Although the signifier and the signified can be distinguished for
analytical purposes, Saussure defines them as wholly interdepend-
ent, neither pre-existing the other. As we will see, this radical concept
proved challenging not only for his ‘deconstructionist’ critics but even
for his structuralist followers. In Saussure’s semiology, the semiotic
articulation or correlation of the signifier and the signified is a délimi-
tation réciproque: a reciprocal or mutual delimitation or definition
(156/112). The signifier and the signified exist in a symbiotic or bidi-
rectional relation within a relational sign system. The two arrows in
the diagram represent their interaction.
MO D E L S 17

Any individual sign is a recognizable combination of a signifier


with a particular signified. For instance, the spoken word ‘duck’ is a
sign consisting of:

•• a signifier: a mental representation of a perceptible pattern of


sound, and
•• a signified: the relational concept of a species of waterbird –
not a pictorial ‘mental image’ but a linguistic ‘value’ (a notion
to be discussed shortly).

Both the signifier and the signified are purely psychological, united
in the mind by an associative link. For Saussure, the ‘acoustic image’
is ‘not the material sound, a purely physical thing’ but the impres-
sion made on our senses or its ‘psychological imprint’ (CLG 98/65–6).
Neither of these are material ‘things’; both consist of non-material
form rather than substance. As we will see, this immateriality derives
from Saussure’s radical conception of language as a system of signs
(a network of pairings of sounds with concepts). Note that in post-
Saussurean semiotics (originally in Hjelmslev), the signifier is
commonly interpreted as the material (or physical) form of the sign –
it is something that can be seen, heard, touched, smelled, or tasted – as
with Roman Jakobson’s signans, which he describes (more tradition-
ally) as the external and perceptible part of the sign (1963b/1990, 111;
1984, 98).
Saussure was a linguist, and his focus was understandably on lin-
guistic sign systems. As we have noted, he refers specifically to the
signifier as an ‘acoustic image’ (image acoustique). In the Cours,
writing is referred to as a separate, secondary, dependent, but com-
parable sign system (CLG 32/15, 46ff./24ff., 165–6/119–20). Within
the system of written signs, a signifier such as the written letter ‘t’
signifies a sound in the primary sign system of language (and thus
a written word would also signify a sound rather than a concept).
Jacques Derrida famously argues that, from this perspective, writing
relates to speech as signifier to signified and writing is ‘a sign of a
sign’ (1967a, 43).
Saussure’s signified is a concept in the mind – not a thing but the
notion of a thing. Some may wonder why his model of the sign refers
only to a concept and not to a thing (the ‘common sense’ view). The
18 M OD E LS

philosopher Susanne Langer (1957, 60) notes that symbolic signs


‘are not proxy for their objects but are vehicles for the conception of
objects’. Such signs perform the extraordinarily powerful function of
enabling us to ‘call to mind’, and communicate about, things that are
not materially present in the here and now.
For Saussure, linguistic signs are wholly immaterial (CLG 32/15).
The immateriality of the Saussurean sign is a feature that tends to be
neglected in many popular commentaries. If the notion seems strange,
we need to remind ourselves that words have no value in themselves –
that is their value. Saussure notes that it is not the metal in a coin that
fixes its value (ibid., 164/118). Several reasons could be offered for this.
For instance, if linguistic signs drew attention to their materiality this
would hinder their communicative transparency. Furthermore, being
immaterial, language is an extraordinarily economical medium, and
words are always ready to hand. Nevertheless, a principled argument
can be made for the revaluation of the materiality of the sign, as we
will see in due course.
Jacques Derrida criticizes Saussure for his ‘psychologism’, dis-
missing the Saussurean model as simply replacing with a mental
representation the referent associated with traditional dyadic models
of the sign (1981, 22–3) – a wilful misrepresentation of Saussure’s rad-
ical conception. Saussure’s model of the linguistic sign (Figure 1.2) is
indeed a psychological one. As such, it presupposes (although it does
not incorporate) a ‘subject’ who interprets the sign. Acknowledging the
interpretive role of the mind enables this ostensibly dyadic model to be
decomposed into a binary structure with two pairs of relata (signifier–
signified, sign–subject), and we may interpret it as implicitly triadic
(see Langer 1957, 57–8).
However, the Saussurean model is not reducible to a matter of
individual psychology. Saussure sees linguistics as closely related to
social psychology (CLG 21/6). Language has both a ‘social side’ and
an ‘individual side’ (ibid., 24/8). The linguistic sign system is socially
grounded and functions as an intersubjective mediator between indi-
viduals in society (Figure 1.3; cf. Figure 6.2). It is reflected as a cognitive
system in the minds of individuals, but it exists in its entirety only
in the masse parlante – the community of speakers (30/14, 112/77).
Language is a social institution that is ‘independent of the individual’
(37/18). It is a cooperative enterprise. Language in use presupposes a
speech community (sharing a common language), and a social (and
MO D E L S 19

F I G U RE 1.3 The linguistic sign and its interpreters


Source: © 2021 Daniel Chandler

material) context in which meaning is negotiated. As Thibault puts


it, ‘Meaning is always the social product of the language system’
(1997, 40). The psychosocial Saussurean model, in which meaning is
socially grounded, is thus radically different from the purely cognitive
Aristotelian model.

ARBITRARINESS
Traditionally, in the theory of signs (before and after Saussure), the
term ‘arbitrary’ refers to sign–object relations, where it is convention-
ally contrasted with ‘natural’ relations. In this context, the issue is
whether the form that the sign takes has some inherent connection
to a referent (as with a shadow) or whether the connection is purely
conventional (as with the word shadow). In Plato’s dialogue Cratylus,
set in the fifth century bce, the issue of ‘the correctness of names’
is debated. This debate produced the first known formulation of an
opposition between resemblance and conventionality as the basis of
signs. Although Cratylus defends the notion of a natural relationship
between words and what they represent, Hermogenes declares that ‘no
one is able to persuade me that the correctness of names is determined
20 M OD E LS

by anything besides convention and agreement … No name belongs


to a particular thing by nature’ (Plato 1998, 2). While Socrates rejects
the absolute arbitrariness of language proposed by Hermogenes, he
does acknowledge that convention plays a part in determining mean-
ing. In his work On Interpretation, Aristotle (2004, 2) went further,
asserting that there can be no natural connection between the sounds
of a language and the things signified. ‘By a noun [or name] we mean
a sound significant by convention … The limitation “by convention”
was introduced because nothing is by nature a noun or name — it is
only so when it becomes a symbol’. The issue even enters into every-
day discourse via Shakespeare: ‘That which we call a rose by any other
name would smell as sweet’ (Romeo and Juliet II.ii).
Saussure rejects the commonsense (naïve realist) assumption that
there is a natural relationship between words and things. Counter-
intuitively, words do not ‘refer to’, ‘correspond to’, or ‘stand for’
things. Language is not based on the mimetic representation of inde-
pendently identifiable ‘things’. There is nothing tree-like about the
word ‘tree’. Even onomatopoeic words (such as words for the sounds
made by familiar animals) are more conventional than is often
supposed, as is demonstrated by the variability between different
languages in their words for the same sounds (CLG 101–2/69). The
principle applies as much to writing as it does to speech: there is no
inherent connection between any letter and the sound that it denotes
(ibid., 165/119).
For Saussure, the arbitrariness of the sign (l’arbitraire du signe)
is the ‘first principle’ of language (ibid., 100ff./67ff.), but his dis-
tinctive linguistic focus is not on arbitrariness in the traditional
referential sense (concerning sign–object relations) but rather on
arbitrariness within the linguistic system. For Saussure, there is no
intrinsic, direct, or self-evident relationship between a linguistic
signifier and its signified (a concept rather than an extralinguistic
referent). The relation between them is an internal link within the
linguistic system. The connections between words and ideas have no
natural basis. The linguistic sign is ‘unmotivated’ (101–2/69). This
point had been anticipated by John Locke. Writing in 1689, Locke
noted that signification works ‘not by any natural connexion that
there is between particular articulate sounds and certain ideas …
but by a voluntary imposition whereby such a word is made arbitrar-
ily the mark of such an idea’ (1690, III.x.5). Nothing prevents any
MO D E L S 21

idea from being associated with any sequence of sounds whatsoever;


such associations are completely arbitrary (CLG 110/76, 157/113).
In language, the form of the signifier is not determined by what it
signifies. No specific signifier is naturally more suited to a signified
than any other signifier; in principle any signifier could represent
any signified.
Many critics have found Saussure’s principle of arbitrariness dif-
ficult if not impossible to accept (or even to understand). In their
book The Meaning of Meaning, Charles Ogden (1889–1957) and Ivor
Richards (1893–1979) criticize Saussure for ‘neglecting entirely the
things for which signs stand’ (1923, 8), but he did not regard linguis-
tic signs as ‘standing for things’ at all. Heretically, Benveniste (1939,
44) argues that a referential third term – ‘the thing itself, reality’ – is
an implicit element in Saussure’s model (and a necessary one in any
model of the sign), and that it is the link between the signifier and its
extralinguistic referent (rather than its signified concept) that is arbi-
trary. Saussure’s critics are quick to suggest that that reality seeps into
his system and that what he called un rudiment de lien naturel (the
vestige of a natural bond) is detectable within it (CLG 101/68). Lest
any be misled by any of Saussure’s critics or postmodernist interpret-
ers, let us be clear that although Saussure denies the referent any place
in his model of the language system this does not involve a denial of
the existence of external reality (or of the fact that people use language
to refer to reality). He insists that language cannot be divorced from
‘social reality’ (ibid., 112/77). Furthermore, he acknowledges that
we cannot conceive of things such as a street or a train outside their
material realization (151–2/109). While linguistic entities are abstract,
‘abstract entities are always based, in the last analysis, on concrete
entities’ (190/138). It is understandable that the Saussurean model is
criticized by realists as philosophically idealist (excluding the reality
of anything beyond what is created in the mind). However, like most
of us, Saussure himself takes the reality of the objective world for
granted, though he sees it as unavoidably mediated through languages
and their categories.
In reading and interpreting Saussure, it is all too easy for the
unwary to slip into treating his principle of arbitrariness as a matter of
reference. After all, how can we assess whether or not the word ‘tree’
is ‘tree-ish’ without reference to some kind of mental representation
of a tree, which (however mediated) must surely derive from some
22 M OD E LS

experience of trees in the living world? Surely such concepts cannot


be wholly divorced from reality? Indeed, in the Cours, the signified for
the word ‘tree’ is illustrated by its editors with a schematic drawing of
a tree (ibid., 99/67). Nevertheless, if we wish to understand Saussure’s
conception of the language system in accord with his own intentions,
we must attempt to exclude the concept of reference. This demanding
task may also assist us in recognizing the (counter-intuitive) mediated-
ness of all experience.
The Saussurean model, with its emphasis on internal structures
within a sign system, supports the notion that rather than reflecting
reality, language plays a major role in constructing it (a contention
challenged by naïve realists). Insofar as we live within the realities con-
structed by language, rather than excluding ‘reality’, the Saussurean
model embodies much of what is real and meaningful to those within a
speech community. What is signified is part of the collective conscious-
ness of such a community rather than part of the material world. The
ontological arbitrariness that the language system involves becomes
invisible to us as we learn to accept it as natural. As the anthropologist
Franz Boas notes, to native speakers of a language, none of its clas-
sifications appear arbitrary (Jakobson 1943, 482). We will return to the
relationship between language and reality in Chapter 2.
Saussure illustrates the principle of arbitrariness at the lexical
level – in relation to individual words as signs. He does not, for
instance, argue that syntax is arbitrary. However, he declares that
the entire system of language is based on ‘the irrational principle’ of
the arbitrariness of the sign. This provocative declaration is followed
immediately by the acknowledgement that this would result in fatal
complications if applied without restriction (CLG 182/133). If linguis-
tic signs were to be totally arbitrary in every way, language would
not be a system and its communicative function would be destroyed.
Saussure concedes that there is no language in which nothing is
motivated. He admits that language is not completely arbitrary but
has a certain logic (ibid., 183/133, 107/73). The principle of arbitrari-
ness does not mean that the form of a word is accidental or random,
of course. While the sign is not determined extralinguistically it is
subject to intralinguistic determination. For instance, combinations
of sounds must conform with existing patterns within the language
in question. Furthermore, we can recognize that a compound noun
such as ‘screwdriver’ is not wholly arbitrary since it is a meaningful
MO D E L S 23

combination of two existing linguistic signs. Saussure introduces a


distinction between degrees of arbitrariness. While repeating that
the linguistic sign is fundamentally arbitrary, he acknowledges that
(even within language) not all signs are ‘radically arbitrary’ (unmo-
tivated) and that the sign may be ‘relatively motivated’ (180–1/131).
Here, then, Saussure modifies his stance somewhat and accepts that
signs may be relatively arbitrary. Social semioticians have sub-
sequently argued that we should not underestimate the situational
motivation of signs in their social context of use, but how language
is used was not what the Cours was about.
It should be noted that, while the relationships between signifiers and
their signifieds are ontologically arbitrary (philosophically, it would
not make any difference to the status of these entities in ‘the order of
things’ if what we call ‘black’ had always been called ‘white’ and vice
versa), this is not to suggest that signifying systems are socially or
historically arbitrary. Natural languages are not, of course, arbitrar-
ily established (unlike inventions such as Morse code). Even in the
case of the colours of traffic lights, the original choice of red for ‘stop’
was not entirely arbitrary since it already carried relevant associations
with danger. The art historian Ernst Gombrich (1909–2001) notes that
‘Red, being the colour of flames and blood, offers itself as a metaphor
for anything that is strident or violent. It is no accident, therefore, that
it was selected as the code sign for “stop” in our traffic code and as a
label of revolutionary parties in politics’ (1963, 13).
After the sign has come into historical existence, it cannot be arbi-
trarily changed. As part of its social use within a sign system, every
sign acquires a history and connotations of its own that are familiar
to members of the sign-users’ culture. Saussure remarks that although
the signifier appears to be freely chosen, from the point of view of the
linguistic community, it is imposed – because a language is a legacy
of the past that its users have no choice but to accept. The arbitrariness
principle does not, of course, mean that an individual can arbitrarily
choose any signifier for a given signified. The relation between a sig-
nifier and its signified is not a matter of individual choice; if it were,
then communication would become impossible. Individuals have no
power to change a sign in any way once it has become established in
the linguistic community (CLG 101/69). From the point of view of
individual language-users, language is a ‘given’ – we don’t create the
system for ourselves.
24 M OD E LS

For Saussure, the arbitrariness of the linguistic sign (the whole


ensemble as well as the sign relations between signifier and signi-
fied) is related to the dependence of language on cultural convention.
Saussure notes that, in principle, every means of expression used in
society is based on collective behaviour or convention (ibid., 101–2/68).
A word means what it does to us only because we collectively agree to
let it do so. As long as there is such an agreement, in principle anything
can signify anything. In cultural studies, the Saussurean emphasis on
arbitrariness and conventionality has been widely employed to alert
us to ‘the familiar mistake of assuming that signs which appear natu-
ral to those who use them have an intrinsic meaning and require no
explanation’ (Culler 1975, 5). A valuable function of the Saussurean
concepts is to undermine our taken-for-granted, commonsense beliefs
in the naturalness of the links between our interpretive systems and
the world. Of course we use language to talk about our experiential
world, but Saussure’s radical stance is that the language system does
not need to refer to extralingual reality.
Saussure felt that the main concern of semiotics should be all sys-
tems that are grounded in the arbitrariness of the sign. For him, it is
signs that are wholly arbitrary that epitomize ‘the semiological pro-
cess’, and of course he sees this as characteristic above all of language
(CLG 101/68). The Cours does not in fact offer many examples of sign
systems other than spoken language and writing, mentioning only the
deaf-and-dumb alphabet, social customs, etiquette, religious and other
symbolic rites, legal procedures, military signals, and nautical flags
(ibid., 33/16, 35/17, 101/68, 107/73). I have discussed elsewhere his gen-
erally neglected observations about flag signalling codes (Chandler
2022). However, Saussure’s critics see his model as excluding signs
that are not as arbitrary, unmotivated, and conventional as words.

THE RELATIONAL SYSTEM


Whereas in traditional models of the sign, meaning is tied to repre-
sentation, for Saussure, meaning is not derived directly from things in
the world but from the way that each language imposes its relational
system on the continuum of thought.
Saussure refers to sound and thought as two distinct but correlated
planes (see Figure 1.4). Later, the Danish linguist Louis Hjelmslev
(1961, 59) refers to these as the ‘expression plane’ and the ‘content
plane’. Language, we are told in the Cours, ‘can … be pictured in its
42 M OD E LS

relations between signs) favours structuralist methodologies, while a


focus on processes and pragmatics (the use of signs) favours Peircean
concepts. Whether or not both approaches are adopted by the same
investigators, in the domain of communication and representation it
seems foolish not to explore their potential to perform complemen-
tary functions.

SIGN RELATIONS
In terms of a (‘referential’) relation between a sign vehicle and a
referent (the sign–object relation), the traditional dualistic distinc-
tion between ‘conventional signs’ (the names we give to people and
things) and ‘natural signs’ dates back to ancient Greece (Plato’s
Cratylus, 360 bce). Endorsing this distinction, Plato’s student Aristotle
regards words as the prototypical examples of conventional signs (De
Interpretatione, 350 bce). Writing in 397 ce, Augustine distinguishes
‘natural signs’ (signa naturalia) from conventional signs (signa data,
‘given signs’) on the basis that natural signs lack intentionality and
are interpreted as signs by virtue of an immediate link to what they
signified (he instances smoke indicating fire and footprints indicating
that an animal had passed by) (On Christian Doctrine, II.1). Both
‘natural’ and ‘conventional’ signs feature in Charles Peirce’s influential
tripartite classification (in which iconic signs straddle the traditional
dualistic distinction).
What Peirce regards as ‘the most fundamental’ division of signs
(first outlined in 1867) has been highly influential (2.275). Although it
is often misinterpreted as a classification of distinct ‘types of signs’,
it refers to differing interpretive relationships between a sign vehicle
and its referent. Whereas Saussurean ‘sign relations’ are between the
signifier and the signified (and internal to the language system), in the
Peircean model the concept is referential. Here then are the three dif-
ferent relations:

1 Symbolic: based on a relationship that is fundamentally


unmotivated, arbitrary, or purely conventional – so that it
must be agreed upon and learned: e.g. language in general
(plus specific languages, alphabetical letters, punctuation
marks, words, phrases, and sentences), numbers, Morse code,
traffic lights, national flags.
MO D E L S 43

2 Iconic: based on perceived resemblance or imitation (involv-


ing some recognizably similar quality, such as appearance,
sound, feeling, taste, or smell) – e.g. a portrait, a cartoon,
a scale-model, onomatopoeia, metaphors, realistic sounds in
‘programme music’, sound effects in radio drama, a dubbed
film soundtrack, imitative gestures.
3 Indexical: based on direct connection (physical or causal).
This link can be observed or inferred – e.g. ‘natural signs’
(smoke, thunder, footprints, echoes, non-synthetic odours
and flavours), medical symptoms (pain, a rash, pulse-rate),
measuring instruments (weathercock, thermometer, clock,
spirit-level), ‘signals’ (a knock on a door, a phone ringing),
pointers (a pointing ‘index’ finger, a directional signpost),
recordings (a photograph, a film, video, or television shot, an
audio-recorded voice), personal ‘trademarks’ (handwriting,
catchphrases).

These three well-established forms of relations between representamen


and object form part of Peirce’s triadic model of the sign. Once again, we
need to remind ourselves that, for Peirce, signification is a three-way rela-
tion (requiring an interpretant) rather than simply a two-way sign–object
relation. Nothing is intrinsically a symbol, an icon, or an index. The same
sign vehicle can, in different contexts, involve different sign relations.
We will not confine ourselves here to Peirce’s own definitions. In
structuralist and poststructuralist discourse these sign–object relations
are often misleadingly imported into a dyadic framework employing
the Saussurean terms signifier and signified. Indeed, ‘the fundamental
post-Saussurean confusion is between the signified of the sign and the
referent of an utterance or a piece of text’ (Tallis 1995, 94). If we adopt
the concepts of iconicity and indexicality, we need to remind ourselves
that neither of these sign relations can be accommodated within the
original Saussurean linguistic model because they depend on some
kind of referential context (beyond the sign system itself). Iconicity is
based on (at least perceived) ‘resemblance’, and indexicality is based
on (at least perceived) ‘direct connection’ (completing the relational
base of the semiotic triangle).
Whereas Jakobson absorbs the Peircean sign relations into his own
dyadic model, the late Italian semiotician, Umberto Eco, challenges
them. Eco was greatly influenced by Peirce, but he criticizes Peirce’s
44 M OD E LS

referential realism and, like Saussure, excludes referentiality from


his own communicational ‘theory of codes’ (1976, 58ff.). For Eco,
‘assuming that the “meaning” of a sign-vehicle has something to do
with its corresponding object’ is a ‘referential fallacy’ (ibid., 62).
‘The real object corresponding to the sign is absolutely irrelevant to
any semiotic purpose’ (1973, 69). He adds that ‘the referent of a sign
is … an abstract entity which moreover is only a cultural convention’
(1976, 66; wholly italicized in the source). Reference is to a cultural
world (ibid., 61–2). In relation to ‘iconism’, Eco insists that percep-
tion itself is coded and that so too is the relation of ‘resemblance’
between an image and its referent. As for indices (or indexes), the
‘conventional’ correlations are inferred from repeated experiences
associating the index with its object (222–3). Eco is a conventional-
ist, seeing all signs as conventionally coded and regarding meanings
as ‘cultural units’, which together form a semantic universe. He
consequently chooses to ‘disregard the difference between motivated
and arbitrary signs’ (121).
Peirce explains the irreducible triadic basis of signs in terms of
three fundamental ontological categories, which he calls firstness, sec-
ondness, and thirdness. All phenomena have these three aspects. Note
that beyond this brief summary, Peircean scholars differ greatly in
how they interpret and apply these categories (Goudge 1950, 85–95).
Firstness is reflected in a sign’s reference to itself, secondness in its ref-
erence to its object (as in the three sign relations), and thirdness in its
reference to the mediating role of its interpretant (CP 1.541). An icon
foregrounds firstness, an index secondness, and a symbol thirdness.
Firstness is exhibited in an inherent quality of a sign, independent of
its object and its interpretant (1.25, 8.329). Secondness is exhibited in
the hard ‘brutal fact’ of ‘the hæcceities [the “thisness”] of things, the
hereness and nowness of them’ (1.405; [my comment]), in their actual,
existential reality (8.330). Furthermore, this is not that: we conceive
of the existence of things in terms of their relations or ‘their reactions
against each other’ (1.324). Secondness thus also involves notness (a
very Saussurean notion). Thirdness is the process of mediation, which
is involved in all processes of representation (1.328, 1.532, 3.423, 5.72,
5.104, 5.105, 7.630). Whereas firstness and secondness are ‘given in
perception’ (5.194), thirdness is the conceptual realm of inference,
thought, and generality. While firstness is monadic, and secondness
is dyadic, thirdness is ‘the triadic relation existing between a sign, its
MO D E L S 45

object, and the interpreting thought’ (8.332). Thus, signs themselves


belong to the category of thirdness.
Within the category of firstness there are three kinds of signs: a
qualisign (a ‘mere quality’), a sinsign (an ‘actual existent’), and a legi-
sign (a ‘general law’). Within that of secondness are the three sign
relations (iconic, indexical, and symbolic). Within thirdness there are
three forms of representation: rheme (possibility), dicisign (fact), and
argument (reason). A qualisign can only be an icon; a sinsign can be
an icon or an index; only a legisign can be a symbol. A colour sample
is a qualisign; any indexical sign is a sinsign; conventional signs, such
as words and traffic lights, are legisigns. Limitations of space prohibit
the exploration of these Peircean refinements and of the ten classes of
sign to which they give rise (2.264).
Sign relations are typically ranked in terms of relative convention-
ality or motivation: primarily symbolic forms such as language are
highly conventional or ‘unmotivated’; iconic forms always involve
some degree of conventionality; indexical forms ‘direct the attention
to their objects by blind compulsion’ (2.306). Indexical and iconic sign
vehicles can be seen as constrained by their referents. Within each form,
particular signs also vary in their degree of conventionality or moti-
vation. Such characteristics have implications for their interpretation
(see Table 1.11). For instance, it is widely argued that while symbolic

sign based forms of knowledge needed


relations primarily on
representational situational
knowledge knowledge
symbolic convention interpreters need
to have learned the
iconic similarity relevant forms and interpreters can
conventions infer likely meanings
indexical connection on the basis of
analogous real-
world situations and
objects

TA B LE 1.11 Knowledge needed for interpreting sign relations


Source: © 2021 Daniel Chandler
46 M OD E LS

F I G URE 1. 12 Sign relations in road signs


Source: © 2021 Daniel Chandler

and (to a lesser extent) iconic signs require ‘decoding’ (as structural-
ists put it), indexical signs are not part of a system of signs and depend
primarily on inference, though Jakobson (1968b) notes that even the
interpretation of indices can involve learning conventional rules (as in
the case of medical symptoms). The overlaps between the three forms
of sign relations underline the fact that conventional, coded symbolism
cannot be tidily divorced from other modes of sign relations.
Before discussing these forms of sign relations in detail, three
British road signs will serve as examples that represent the rela-
tive dominance of each (see Figure 1.12). All three are official blue
informational signs with white borders. An example of the primacy
of symbolic relations is a round-cornered square with a bold white
letter P, signifying ‘Parking’ (viewers would need to understand the
convention in order to know what the ‘P’ stood for). A circular sign
with a generic bicycle shape is an example of iconic relations based on
perceived similarity and signifying a ‘recommended route for pedal
cycles’. A round-cornered square sign with generic crossed spoon and
fork shapes in white refers indexically to the associated concept of
‘eating place’ and indicates that there is a motorway café ahead. As
part of the sign system in the British Highway Code, all such road
signs also have a conventional symbolic dimension, and as individ-
ual objects in specific roadside contexts, they also have an indexical
dimension (connecting their intended meaning with their location).

SYMBOLIC RELATIONS
Language is a (predominantly) symbolic sign system and it is widely
seen as the pre-eminent symbolic form. Peirce declares that ‘All
words, sentences, books and other conventional signs are symbols’ (CP
MO D E L S 47

2.292), and we will follow his usage here. Saussure avoids referring to
linguistic signs as ‘symbols’ because of the danger of confusion with
popular usage, in which symbols are never wholly arbitrary, showing
the vestige of a ‘natural bond’ with what they signify (CLG 101/68,
106/73). For instance, if we joke that ‘a thing is a phallic symbol if it’s
longer than it’s wide’, this would allude to resemblance, making it at
least partly iconic – Jakobson (1968a, 702) suggests that such exam-
ples may be best classified as ‘symbolic icons’ (see also Figure 1.14
for hybrid signs). It is thus important to bear in mind that literary,
religious, mythical, and heraldic ‘symbols’ are not purely symbolic in
the semiotic sense.
Peirce notes that symbols (which he then called ‘tokens’) ‘are, for the
most part, conventional or arbitrary’ (CP 3.360). Arbitrariness refers
here, of course, to the sign–object relation. A symbol is a sign ‘whose
special significance or fitness to represent just what it does represent lies
in nothing but the very fact of there being a habit, disposition, or other
effective general rule that it will be so interpreted. Take, for example,
the word “man”. These three letters are not in the least like a man; nor is
the sound with which they are associated’ (4.447). Peirce adds elsewhere
that ‘a symbol … fulfils its function regardless of any similarity or anal-
ogy with its object and equally regardless of any factual connection
therewith’ (5.73). The lack of dependence on resemblance or direct con-
nection contributes to the power and flexibility of symbols in this sense.
‘The symbol,’ says Peirce, ‘is connected with its object by virtue of
the idea of the symbol-using mind, without which no such connection
would exist’ (CP 2.299). It ‘is constituted a sign merely or mainly by
the fact that it is used and understood as such’ (2.307). Symbolic signs
are usually understood to be intended to communicate something. Their
intended meanings are not intuitively obvious: understanding symbolic
signs depends wholly on our familiarity with the relevant conventions
(without which they may fail to signify or be misinterpreted). We inter-
pret symbols according to ‘a habitual connection’, ‘a rule’, or ‘a law,
usually an association of general ideas, which operates to cause the
symbol to be interpreted as referring to that object’ (2.292, 2.297, 1.369,
2.249). Consequently, the symbol needs to exist in ‘constant conjunctions
whereby it is linked to something other than itself’ (Price 1969, 213).
While arbitrariness is regarded as a key ‘design feature’ of lan-
guage, displacement is another key property that enables linguistic
signs to be used to represent objects in their physical absence,
48 M OD E LS

including entities that exist only in our imaginations (Hockett 1958).


The meaning of truly general symbols can thus transcend particular
contexts (the ‘here and now’ to which other animals are confined).
This is in strongest contrast to indexical signs (corresponding to a
distinction between semantic and pragmatic meaning). Such quali-
ties enable symbols to be good to think with. The symbolic mode is
the most effective in communicating abstract concepts and gener-
alizations. Peirce observes that ‘A genuine symbol is a symbol that
has a general meaning’ (CP 2.293), signifying a kind of thing rather
than a particular thing (2.301). Linguistic categories perform this key
function. ‘Though what we call reality is too rich and too varied to
be reproducible at will, symbols can be learned and recalled to a
surprising extent’ (Gombrich 1982, 16).
Symbols are widely regarded by semioticians as best exemplified
by words, but they are not limited to this form. They include any con-
ventional signs that signify concepts, and any form of signifier has the
potential to be used symbolically. It is possible for images to attain some
of the symbolic status of words, though their scope and flexibility are
limited. As the Irish philosopher George Berkeley notes (Principles,
1710/2004, §10; Three Dialogues, 1713), images cannot ‘resemble’
the referents of abstract ideas. Concepts are not directly picturable.
However, images can be used symbolically to represent objects asso-
ciated with abstract concepts. Hence, the familiar ideographic heart
symbol (Figure 1.13 L) is variously used to signify abstractions such

F I G U RE 1 .13 Representing the heart


Source: © 2021 Daniel Chandler
MO D E L S 49

as love, romance, emotion, or health. On the internet, simple house-


like images signifying a link to a website’s homepage are primarily
symbolic: they do not resemble a particular house but conventionally
signify the concept ‘home’. In some contexts an image of a lion may
function as a metaphorical symbol for abstract concepts such as power
or courage (which are of course conventional associations rather than
zoological features of lions).
When images have a well-established conventional meaning, they
function as iconic symbols. Public signage often depends on visual
symbols that are intended to achieve generality of reference (within
a limited universe of discourse) and to carry authority. For instance,
in a ‘no dogs allowed’ sign, a simple, schematic representation of a
prototypical dog is likely to be more effective as a general symbol
than an image of a particular type of dog – although, unlike the word
‘dog’, any depiction of a dog unavoidably resembles some kinds of
dogs more than others. Even a photograph can have a symbolic func-
tion (in an appropriate context), as when images of coinage are used to
signify ‘the economy’ when that is the topic of a TV news item.
Symbolic forms are often more indirect. Only context and conven-
tion can determine whether a sign featuring a generic image of a male
figure is intended to symbolize people in general (as on a ‘no admit-
tance’ sign), or males only (as on the door of a public toilet). In the
case of the British road sign where a simple schematic image of an
elephant on a directional sign signifies a nearby zoo the sign might
best be described as indexically symbolic, since the elephant is part of,
and here stands for, the animal kingdom as a whole.

ICONIC RELATIONS
The traditional basis of iconic sign -object relations lies in resemblance.
In semiotics, icons are not necessarily visual but may involve any
sensory mode. However, pictures are widely cited as prime examples
of iconicity. It is a necessary feature of a representational picture
(such as a portrait) that it is perceived as looking sufficiently like
what it depicts for that object to be recognized. However, from Plato
onwards, philosophers have pointed out that everything resembles
everything else in various ways, having multiple properties in
common. Pictures can be seen as visually resembling their subjects
50 M OD E LS

in countless ways (while being notably unlike them in others – such


as being typically flat and framed, properties that most resemble
those of other pictures). Although the most relevant representational
property is often deemed to be outline shape (Hopkins 1998, Hyman
2006), this is not always the case, as in the depiction of the sky,
the sea, clouds, smoke, and so on. In short, resemblance is not as
distinctive of pictorial representation as is often assumed; it is not a
sufficient condition as in traditional resemblance theories. ‘Like other
judgments, similarity depends on context and frame of reference’
(Tversky 1977, 340).
Nevertheless, since at least the ancient Greeks, resemblance has
been generally regarded as the ‘natural’ basis of depiction in contrast
to the conventionality of words. To the extent that we are able to rec-
ognize something familiar in a particular image (even cross-culturally
and across the centuries), it is understandable that iconic relations are
widely perceived as more ‘natural’ than symbolic relations. Iconic
forms do not draw our attention to their mediation, seeming to pre-
sent reality more directly or transparently than symbolic forms. Ernst
Gombrich refers to ‘the beholder’s capacity to read “iconicity” into …
[the] sign’ (1949/1987, 248), ‘seeing’ the referent in the sign vehicle –
even in a highly conventionalized iconic symbol such as the heart on
a playing card, which bears very little anatomical relation to a human
heart (Figure 1.13). We routinely underestimate the importance of our
active contribution to sense-making, but images in particular do not
usually seem to need ‘reading’ at all.
Few critics of resemblance theory deny that perceived resemblance
plays some role in pictorial representation (e.g. Schier 1986, 186; Lopes
1996, 4). However, its strongest opponents emphasize the primacy of
conventions. ‘The notion of matching to the real world is insufficient
to explain how pictures mean … Correspondence … is not … to “real-
ity”, but rather … to conventions’ (Worth 1981, 181). All artists employ
stylistic conventions, and these are, of course, culturally and histori-
cally variable. The most radical ‘conventionalist’ is the philosopher
Nelson Goodman (1968, 1972), who argues that what a picture repre-
sents depends only on the symbol system to which it belongs. As Flint
Schier, another philosopher, puts it, ‘The real contrast between … a
linguistic symbol system and an iconic one does not consist in the con-
ventional nature of the former and the non-conventional nature of the
latter but in the nature of the conventions involved in each’ (1986, 179).
MO D E L S 51

Peirce endorses the traditional basis of iconic relations in a


(perceived) resemblance of the sign vehicle to (or analogy with) the
referent. He originally termed such sign relations, ‘likenesses’ (e.g.
CP 1.558). For him, icons have qualities that ‘resemble’ those of the
objects they stand for, and they ‘excite analogous sensations in the
mind’ (2.299). An iconic sign represents its object ‘mainly by its
similarity’ (2.276). However, similarity is not a sufficient basis for
sign status. A sign is an icon ‘insofar as it is like that thing and used
as a sign of it’ (2.247; my emphasis).
Although iconic signs are often defined as natural, and contrasted
with conventional signs, Peirce regards iconic signs as both motivated
and conventional. Although ‘any material image’ (such as a painting)
may be perceived as looking like what it represents, it is ‘largely con-
ventional in its mode of representation’ (CP 2.276).

We say that the portrait of a person we have not seen is convincing.


So far as, on the ground merely of what I see in it, I am led to form
an idea of the person it represents, it is an icon. But, in fact, it is not
a pure icon, because I am greatly influenced by knowing that it is
an effect, through the artist, caused by the original’s appearance …
Besides, I know that portraits have but the slightest resemblance to
their originals, except in certain conventional respects, and after a
conventional scale of values, etc.
(CP 2.92)

Peirce nevertheless maintains that ‘every picture (however conven-


tional its method)’ is an icon (2.279).
For most semioticians, pictures are not as conventional as linguistic
signs. However, many structuralists agree that reference is mediated not
only by resemblance but also by representational conventions. Indeed,
it has been suggested that iconic signs resemble not their referents but
conventionalized representations of such referents within a culture (e.g.
Washabaugh 1980). In their own model of the iconic sign, the Belgian
Groupe μ (1995a) emphasizes the mediation of the referent by type (or
conceptual category). While Eco (1968) argues that what an iconic image
resembles is its mental representation, he later offers an influential critique
of ‘iconism’ from a conventionalist perspective (1976, 191ff.; 2017).
For Morris, iconicity is ‘a matter of degree’ (1971, 273). However
‘natural’ the resemblance may seem, pictures resemble what they
52 M OD E LS

represent only in some respects. What we tend to recognize in an


image are analogous relations of parts to a whole: as in a diagram
(Peirce CP 2.279, 2.282). Even the most ‘realistic’ image is not a replica
or even a copy of what is depicted. It is not often that we mistake a
representation for what it represents. Of course, pictures are often used
(as in books like this) to illustrate what is being referred to. However,
unlike the index, ‘the icon has no dynamical connection with the object
it represents’ (2.299) and consequently ‘gives no assurance that any
such object as it represents really exists’ (Peirce 1976, vol. 4, 242–3).
Iconicity is not an objective sign–object relation: resemblance is in the
eye (or rather the mind) of the beholder.
As art historians remind us, pictures do not necessarily stand for
what they depict (or refer to what they resemble): they may be ‘sym-
bolic’ as well as iconic. However, since abstractions are not directly
picturable, ‘iconic language’ has a very limited representational
capacity (as in highly conventionalized pictographic signage). The
linguist John Lyons (1977, I, 105) notes that iconicity depends on the
affordances of the medium. While the phonic medium can represent
characteristic sounds (albeit in a relatively conventionalized way), the
graphic medium can represent characteristic shapes (as in the case of
Egyptian hieroglyphs). Iconic gestures, which are ‘remarkably similar
cross-linguistically’ (McNeil 1995, 358), are highly effective in com-
municating shapes, locations, and movements. In sign languages for
the deaf, iconic signs have the same advantage over spoken languages
(Taub 2001).

INDEXICAL RELATIONS
Peirce’s category of indexical signs includes what have been rec-
ognized as ‘natural signs’ since ancient times. As already noted,
Augustine, who regards both words and pictures as conventional
signs, distinguishes from these as natural signs ‘those which,
apart from any intention or desire of using them as signs, do yet lead
to the knowledge of something else, as for example, smoke when it
indicates fire’ (2009, 32).
As the semiotician Winfried Nöth (1990, 86) points out, the case
of natural signs ‘raises the question whether it is still appropriate to
call this relation one of “standing for” ’. It has even been suggested
that ‘indices are not signs’ (e.g. Brandt 1999, 99; original italicized).
MO D E L S 53

Unlike iconic and symbolic signs, indexical signs do not represent or


symbolize things. Although we often interpret dark clouds as signs
of imminent rain, they do not mean rain (in the sense of reflecting
some intent). ‘There is nothing conventional or arbitrary about the
fact that cumulo-nimbus clouds are a sign of a thunderstorm’ (Price
1969, 173). Nor do they signify rain iconically by resembling it or by
similarity of attributes. However, they do signify something other
than themselves when we infer that they indicate impending rain-
storms in the vicinity.
Unlike symbolic and iconic signs, natural signs are not in them-
selves part of an interpretive system (although they may nevertheless
need to be interpreted according to conventions, as in weather fore-
casts). Of course, natural signs, which are directly referential, have no
place in the Saussurean model. It is somewhat ironic that the study of
medical symptoms as natural signs, which originated with Hippocrates
(460–377 bce) and was developed by Galen (c. 129–200 ce), had by the
nineteenth century become a branch of medicine known as semiology
(apparently unbeknown to Saussure). Linguistic indexicality, termed
deixis (Bühler 1934), has subsequently become a central issue in prag-
matics. The referential meaning of deictic words, such as ‘you’ and ‘I’,
‘this’ and ‘that’, and ‘here’ and ‘now’, is relative to the context of the
situation in which they are used. In social interaction, deixis thus links
the language system with specific contextual referents.
For Peirce, an index ‘indicates’ something: for example, ‘a sundial
or clock indicates the time of day’ (CP 2.285). Indexicality is quite
closely related to the way in which the index of a book or an ‘index’
finger point directly to what is being referred to (Latin index ‘forefin-
ger’, ‘pointer’). ‘The index, … like a pointing finger exercises a real
physiological force over the attention … and directs it to a particular
object’ (8.41). Indexical signs ‘direct the attention to their objects by
blind compulsion’ (2.306). An index typically points away from itself,
but in some instances it points to itself (as in the case of this self-
referential sentence).
Peirce’s choice of ‘a pointing finger’ as the prototypical example
of the indexical class (CP 3.361) may be a little misleading since,
unlike natural signs, the human act of pointing is (also) a symbolic
sign (further conventionalized in pictographic arrows and directional
signposts). Of course, pointing gestures do have an indexical dimen-
sion, although linguistically challenged travellers can testify to their
54 M OD E LS

limited utility as a form of communication. We can point only to par-


ticular things. Even if interpreters are able to identify the referent (at
the relevant level of abstraction), pointing shows something, but it
says nothing about it – it needs supplementing by other kinds of signs.
Indeed, indexicality is in the eye of the beholder to the extent that eve-
rything can be interpreted as pointing to something (cf. Goethe 1818,
cited in Urzidil 1948, 21). Just as with other sign relations, indexical-
ity requires interpretation. An ungraduated thermometer would be of
limited utility, and even a graduated one requires familiarity with the
relevant conventions.
Nevertheless, indexical relations offer the most direct connection
with a referent, in strongest contrast to symbolic relations. An indexi-
cal sign is symptomatic of a state of affairs. As Susanne Langer (1957,
57) puts it, a natural sign ‘indicates the existence — past, present, or
future — of a thing, event, or condition’. For Peirce, indexical relations
are found in signs where there is a direct, ‘genuine relation’ between
the ‘sign’ and the object that does not depend purely on ‘the interpret-
ing mind’ (CP 2.92, 2.98). The connection is ‘a matter of fact’ (4.447).
A windsock doesn’t lie. The object is ‘necessarily existent’ (2.310).
There is ‘a real connection’ (5.75), which may be a ‘direct physical
connection’ (1.372, 2.281, 2.299), in which case an index is like ‘a frag-
ment torn away from the object’ (2.231). Unlike an icon (the object of
which may be fictional), an index stands ‘unequivocally for this or that
existing thing’ (4.531).
By definition, indexical meaning depends on the contextual relation
between the sign vehicle and its object. Indexical signs lack the general-
ity of symbols (the meanings of which are not inherently dependent on
context in this way). Whereas iconicity is characterized by similarity,
indexicality is characterized by contiguity (2.306). Jakobson asserts that
‘the index, in contradistinction to the icon and symbol, is the only sign
which necessarily involves the actual copresence of its object’ (Jakobson
1963c, 335). In indexical relations, the sign vehicle and the referent ‘go
together’. Elizabeth Bruss (1978, 88) notes that indexicality is a rela-
tionship in which the sign vehicle has ‘a demonstrable connection to
something else. The most important of these connections are spatial co-
occurrence, temporal sequence, and cause and effect’.
Indexicality comes in many forms (some of which may be classed as
quasi-indexical): for instance, road signs exhibit indexical dimensions
when they signify appropriately only in the intended location (e.g. a
crossroads sign at an intersection), when they point directly to what they
MO D E L S 55

refer (directional signposts), or when they depict something closely con-


nected with what they stand for (an image of cutlery indicating nearby
dining facilities) – though it should be apparent by now that none of
these examples is purely indexical. Anything that ‘reflects’ an individu-
al’s personality, age, or sex is also indexical. This includes handwriting,
or a distinctive individual style (e.g. of an artist or photographer).

THE CASE OF PHOTOGRAPHY


While photographs are perceived as iconic (visually resembling their
subjects), Peirce notes that they are also indexical (leading some to
refer to them as iconic indices).

Photographs, especially instantaneous photographs, are very


instructive, because we know that in certain respects they are exactly
like the objects they represent. But this resemblance is due to the
photographs having been produced under such circumstances that
they were physically forced to correspond point by point to nature.
In that aspect, then, they belong to the … class of signs … by physi-
cal connection [the indexical class].
(CP 2.281)

The British philosopher Roger Scruton (1983, 122) argues that, strictly
speaking, ‘photography is not representation’, since ‘the camera is … not
being used to represent something but to point to it’ (ibid., 113). Since the
photographic image is an index of the effect of light, all unedited photo-
graphic and filmic images are indexical (although we should remember
that conventional practices are always involved in composition, focus-
ing, developing, and so on). Such images do of course ‘resemble’ visual
features of their subject, and some commentators suggest that the power
of the photographic and filmic image derives from the iconic character
of the medium (though a snapshot doesn’t necessarily guarantee ‘a good
likeness’). However, while digital imaging techniques have eroded the
indexicality of photographic images, it is the indexicality still routinely
attributed to the medium that is primarily responsible for interpreters
treating them as objective records of reality. Although a photograph
generally resembles its referent it no longer does so when the image is
dramatically over- or under-exposed.
As the sociologist Pierre Bourdieu (1990a, 164) puts it, ‘Photography
is ordinarily seen as the most perfectly faithful reproduction of the
56 M OD E LS

real’. Photographic media seem to involve a transparent relationship


between the sign vehicle and its referent. A photograph is based on a
direct causal relationship to an external object. The Pencil of Nature
was the title that Henry Fox gave to his book on photography in 1844,
and subsequent writers on the subject have referred to a photograph as
a ‘trace’ of its subject (e.g. Sontag 1979, 154; Berger 1980, 50). Of the
three types of relationship on which signs are based, only indexicality
can serve as evidence of an object’s existence. Peirce (CP 4.447)
observes that ‘a photograph … owing to its optical connection with
its object, is evidence that that appearance corresponds to a reality’. In
many contexts, photographs are regarded as evidence of events having
happened at the time they were taken, not least in legal contexts: video
surveillance cameras and speed cameras are of course widely used
in this way. However, in one of his essays on photographic history,
John Tagg, wary of ‘the realist position’, cautions that ‘the existence
of a photograph is no guarantee of a corresponding pre-photographic
existent … The indexical nature of the photograph – the causative link
between the pre-photographic referent and the sign … can guarantee
nothing at the level of meaning’ (1988, 2–3). Even prior to digital
photography, both ‘correction’ and montage were practised, but Tagg
argues that every photograph involves ‘significant distortions’ (ibid.,
2). Just as pictures resemble what they represent only in some respects,
in photography whatever is shown is unavoidably transformed into a
flat, decontextualized fragment that is typically much smaller in scale
and (except in films) stationary and silent. This is an issue to which we
will return in Chapter 4 when we discuss whether photography is ‘a
message without a code’.
As a medium, photography is primarily indexical; as a ‘message’
(which is dependent on how a photograph is used or interpreted), this
sign relation may not be dominant. In discussing symbolicity we noted
that photographs have this potential function. Like all ‘photographs that
changed the world’, Joe Rosenthal’s famous 1945 photograph, Raising
the Flag on Iwo Jima (‘iconic’ in the popular sense) is in semiotic terms
primarily symbolic (see Hariman and Lucaites 2007). Peirce argues that
all indices refer to single instances (e.g. ‘this man’), whereas symbolic
representation can refer to general classes of instances (e.g. ‘men’).
Despite their literal specificity (making them unsuitable for road signs)
it is possible for photographs to be used to refer to general classes.
Within the genre of advertising, the potency of the famous photographic
MO D E L S 57

‘cowboy’ ads for Marlboro cigarettes does not derive primarily from
their indexicality as images of particular individuals, but rather from
their symbolization of a stereotypical concept of masculinity (even to
the extent that we may need to remind ourselves that this is not ‘mascu-
linity’, because the form functions to suggest otherwise).

MIXED MODES
It is easy to slip into referring to Peirce’s three forms as ‘types of signs’,
but they are not necessarily mutually exclusive: a sign can have iconic,
symbolic, and indexical aspects (or any combination). For instance, a
map is indexical in pointing to the locations of things, iconic in rep-
resenting directional relations and distances between landmarks, and
symbolic in using conventional symbols (the significance of which
must be learned). We have already noted that Peirce did not regard a
portrait as a pure icon. A ‘stylized’ image might be more appropriately
regarded as a ‘symbolic icon’. Jakobson refers to such combined terms
as ‘transitional varieties’ (1968a, 700), though arguably these hybrid
forms actually represent the norm (see Figure 1.14).

FI G U R E 1.1 4 Hybridity in sign relations


Source: © 2016 Daniel Chandler
58 M OD E LS

Peirce insists that ‘it would be difficult if not impossible to instance


an absolutely pure index, or to find any sign absolutely devoid of the
indexical quality’ (CP 2.306). All signs could be regarded as having
an indexical dimension insofar as they give rise to inferences about
connections to the referent. The indexicality of a thermometer needs
a conventional label in order to function as a sign. Jakobson (1968a,
700–1) points out that many deliberate indices also have a symbolic
or indexical quality, instancing traffic lights as being both indexical
and symbolic, and noting that even the pointing gesture is not always
interpreted purely indexically in different cultural contexts. Nor are
words always purely symbolic – they can be ‘iconic symbols’ (such
as onomatopoeic words) or ‘indexical symbols’ (as with deictic words
such as ‘here’ and ‘now’) (see Jakobson 1966 on iconicity and indexi-
cality in language).
Jakobson (1966/1990, 411) notes that Peirce’s three forms of sign
relations coexist in a ‘relative hierarchy’ in which one form is domi-
nant, with dominance determined by context. Thomas Sebeok suggests
that ‘the sign is legitimately, if loosely, labelled after the aspect that
ranks predominant’ (1994a, 22). Whether a sign is symbolic, iconic, or
indexical depends primarily on the way in which the sign is used, so
textbook examples chosen to illustrate the various sign relations can
be misleading. The same sign may be used iconically in one context
and symbolically in another: a photograph of a woman may stand for
some broad category such as ‘women’ or may more specifically rep-
resent only the particular woman who is depicted. Signs cannot be
classified in terms of the three forms of relations without reference to
the purposes of their users within particular contexts. A sign may con-
sequently be treated as symbolic by one person, as iconic by another,
and as indexical by a third. Signs may also shift in mode over time.
For instance, a Rolex watch is an index of wealth because one must be
wealthy to own one, but social usage has led to its becoming a conven-
tional symbol of wealth.
Consistently with his advocacy of binary relations, Jakobson
(1968a, 700) boldly asserts that Peirce’s three forms of sign relations
are ‘actually based on two substantial dichotomies’ – an assertion
which understandably irritates a Peircean scholar (Bruss 1978, 92).
Combining four terms used by Peirce, Jakobson proposes a matrix
of his own with contiguity and similarity on one axis and the quali-
ties of being either ‘imputed’ or ‘factual’ on the other (Table 1.15).

You might also like