Semiotics for Beginners

Signs

We seem as a species to be driven by a desire to make meanings: above all, we are surely Homo significans - meaning-makers. Distinctively, we make meanings through our creation and interpretation of 'signs'. Indeed, according to Peirce, 'we think only in signs' (Peirce 1931-58, 2.302). Signs take the form of words, images, sounds, odours, flavours, acts or objects, but such things have no intrinsic meaning and become signs only when we invest them with meaning. 'Nothing is a sign unless it is interpreted as a sign', declares Peirce (Peirce 1931-58, 2.172). Anything can be a sign as long as someone interprets it as 'signifying' something - referring to or standing for something other than itself. We interpret things as signs largely unconsciously by relating them to familiar systems of conventions. It is this meaningful use of signs which is at the heart of the concerns of semiotics.

The two dominant models of what constitutes a sign are those of the linguist Ferdinand de Saussure and the philosopher Charles Sanders Peirce. These will be discussed in turn.

Saussure offered a 'dyadic' or two-part model of the sign. He defined a sign as being composed of:

• a 'signifier' (signifiant) - the form which the sign takes; and
• the 'signified' (signifié) - the concept it represents.

The sign is the whole that results from the association of the signifier with the signified (Saussure 1983, 67; Saussure 1974, 67). The relationship between the signifier and the signified is referred to as 'signification', and this is represented in the Saussurean diagram by the arrows. The horizontal line marking the two elements of the sign is referred to as 'the bar'.

If we take a linguistic example, the word 'Open' (when it is invested with meaning by someone who encounters it on a shop doorway) is a sign consisting of:

• a signifier: the word open;
• a signified concept: that the shop is open for business.

A sign must have both a signifier and a signified. You cannot have a totally meaningless signifier or a completely formless signified (Saussure 1983, 101; Saussure 1974, 102-103). A sign is a recognizable combination of a signifier with a particular signified. The same signifier (the word 'open') could stand for a different signified (and thus be a different sign) if it were on a push-button inside a lift ('push to open door'). Similarly, many signifiers could stand for the concept 'open' (for instance, on top of a packing carton, a small outline of a box with an open flap for 'open this end') - again, with each unique pairing constituting a different sign.

Nowadays, whilst the basic 'Saussurean' model is commonly adopted, it tends to be a more materialistic model than that of Saussure himself. The signifier is now commonly interpreted as the material (or physical) form of the sign - it is something which can be seen, heard, touched, smelt or tasted. For Saussure, both the signifier and the signified were purely 'psychological' (Saussure 1983, 12, 14-15, 66; Saussure 1974, 12, 15, 65-66). Both were form rather than substance:

A linguistic sign is not a link between a thing and a name, but between a concept and a sound pattern. The sound pattern is not actually a sound; for a sound is something physical. A sound pattern is the hearer's psychological impression of a sound, as given to him by the evidence of his senses. This sound pattern may be called a 'material' element only in that it is the representation of our sensory impressions. The sound pattern may thus be distinguished from the other element associated with it in a linguistic sign. This other element is generally of a more abstract kind: the concept. (Saussure 1983, 66; Saussure 1974, 66)

Saussure was focusing on the linguistic sign (such as a word) and he 'phonocentrically' privileged the spoken word, referring specifically to the image acoustique ('sound-image' or 'sound pattern'), seeing writing as a separate, secondary, dependent but comparable sign system (Saussure 1983, 15, 24-25, 117; Saussure 1974, 15, 16, 23-24, 119). Within the ('separate') system of written signs, a signifier such as the written letter 't' signified a sound in the primary sign system of language (and thus a written word would also signify a sound rather than a concept). Thus for Saussure, writing relates to speech as signifier to signified. Most subsequent theorists who have adopted Saussure's model are content to refer to the form of linguistic signs as either spoken or written. We will return later to the issue of the post-Saussurean 'rematerialization' of the sign.

As for the signified, most commentators who adopt Saussure's model still treat this as a mental construct, although they often note that it may nevertheless refer indirectly to things in the world. Saussure's original model of the sign 'brackets the referent': excluding reference to objects existing in the world. His signified is not to be identified directly with a referent but is a concept in the mind - not a thing but the notion of a thing. Some people may wonder why Saussure's model of the sign refers only to a concept and not to a thing. An observation from the philosopher Susanne Langer (who was not referring to Saussure's theories) may be useful here. Note that like most contemporary commentators, Langer uses the term 'symbol' to refer to the linguistic sign (a term which Saussure himself avoided): 'Symbols are not proxy for their objects but are vehicles for the conception of objects... In talking about things we have conceptions of them, not the things themselves; and it is the conceptions, not the things, that symbols directly mean. Behaviour towards conceptions is what words normally evoke; this is the typical process of thinking'. She adds that 'If I say "Napoleon", you do not bow to the conqueror of Europe as though I had introduced him, but merely think of him' (Langer 1951, 61).

Thus, for Saussure the linguistic sign is wholly immaterial - although he disliked referring to it as 'abstract' (Saussure 1983, 15; Saussure 1974, 15). The immateriality of the Saussurean sign is a feature which 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 noted that it is not the metal in a coin that fixes its value (Saussure 1983, 117; Saussure 1974, 118). Several reasons could be offered for this. For instance, if linguistic signs drew attention to their materiality this would hinder their communicative transparency (Langer 1951, 73). 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 shall see in due course.

Saussure noted that his choice of the terms signifier and signified helped to indicate 'the distinction which separates each from the other' (Saussure 1983, 67; Saussure 1974, 67). Despite this, and the horizontal bar in his diagram of the sign, Saussure stressed that sound and thought (or the signifier and the signified) were as inseparable as the two sides of a piece of paper (Saussure 1983, 111; Saussure 1974, 113). They were 'intimately linked' in the mind 'by an associative link' - 'each triggers the other' (Saussure 1983, 66; Saussure 1974, 66). Saussure presented these elements as wholly interdependent, neither pre-existing the other (Silverman 1983, 103). Within the context of spoken language, a sign could not consist of sound without sense or of sense without sound. He used the two arrows in the diagram to suggest their interaction. The bar and the opposition nevertheless suggests that the signifier and the signified can be distinguished for analytical purposes. Poststructuralist theorists criticize the clear distinction which the Saussurean bar seems to suggest between the signifier and the signified; they seek to blur or erase it in order to reconfigure the sign or structural relations. Some theorists have argued that 'the signifier is always separated from the signified... and has a real autonomy' (Lechte 1994, 68), a point to which we will return in discussing the arbitrariness of the sign. Commonsense tends to insist that the signified takes precedence over, and pre-exists, the signifier: 'look after the sense', quipped Lewis Carroll, 'and the sounds will take care of themselves' (Alice's Adventures in Wonderland, chapter 9). However, in dramatic contrast, post-Saussurean theorists have seen the model as implicitly granting primacy to the signifier, thus reversing the commonsensical position.

Louis Hjelmslev used the terms 'expression' and 'content' to refer to the signifier and signified respectively (Hjelmslev 1961, 47ff). The distinction between signifier and signified has sometimes been equated to the familiar dualism of 'form and content'. Within such a framework the signifier is seen as the form of the sign and the signified as the content. However, the metaphor of form as a 'container' is problematic, tending to support the equation of content with meaning, implying that meaning can be 'extracted' without an active process of interpretation and that form is not in itself meaningful (Chandler 1995 104-6).

Saussure argued that signs only make sense as part of a formal, generalized and abstract system. His conception of meaning was purely structural and relational rather than referential: primacy is given to relationships rather than to things (the meaning of signs was seen as lying in their systematic relation to each other rather than deriving from any inherent features of signifiers or any reference to material things). Saussure did not define signs in terms of some 'essential' or intrinsic nature. For Saussure, signs refer primarily to each other. Within the language system, 'everything depends on relations' (Saussure 1983, 121; Saussure 1974, 122). No sign makes sense on its own but only in relation to other signs. Both signifier and signified are purely relational entities (Saussure 1983, 118; Saussure 1974, 120). This notion can be hard to understand since we may feel that an individual word such as 'tree' does have some meaning for us, but its meaning depends on its context in relation to the other words with which it is used.

Together with the 'vertical' alignment of signifier and signified within each individual sign (suggesting two structural 'levels'), the emphasis on the relationship between signs defines what are in effect two planes - that of the signifier and the signifier. Later, Louis Hjelmslev referred to the planes of 'expression' and 'content' (Hjelmslev 1961, 60). Saussure himself referred to sound and thought as two distinct but correlated planes. 'We can envisage... the language... as a series of adjoining subdivisions simultaneously imprinted both on the plane of vague, amorphous thought (A), and on the equally featureless plane of sound (B)' (Saussure 1983, 110-111; Saussure 1974, 112). The arbitrary division of the two continua into signs is suggested by the dotted lines whilst the wavy (rather than parallel) edges of the two 'amorphous' masses suggest the lack of any 'natural' fit between them. The gulf and lack of fit between the two planes highlights their relative autonomy. Whilst Saussure is careful not to refer directly to 'reality', Fredric Jameson reads into this feature of Saussure's system that 'it is not so much the individual word or sentence that "stands for" or "reflects" the individual object or event in the real world, but rather that the entire system of signs, the entire field of the langue, lies parallel to reality itself; that it is the totality of systematic language, in other words, which is analogous to whatever organized structures exist in the world of reality, and that our understanding proceeds from one whole or Gestalt to the other, rather than on a one-to-one basis' (Jameson 1972, 32-33).

What Saussure refers to as the 'value' of a sign depends on its relations with other signs within the system - a sign has no 'absolute' value independent of this context (Saussure 1983, 80; Saussure 1974, 80). Saussure uses an analogy with the game of chess, noting that the value of each piece depends on its position on the chessboard (Saussure 1983, 88; Saussure 1974, 88). The sign is more than the sum of its parts. Whilst signification - what is signified - clearly depends on the relationship between the two parts of the sign, the value of a sign is determined by the relationships between the sign and other signs within the system as a whole (Saussure 1983, 112-113; Saussure 1974, 114).

The notion of value... shows us that it is a great mistake to consider a sign as nothing more than the combination of a certain sound and a certain concept. To think of a sign as nothing more would be to isolate it from the system to which it belongs. It would be to suppose that a start could be made with individual signs, and a system constructed by putting them together. On the contrary, the system as a united whole is the starting point, from which it becomes possible, by a process of analysis, to identify its constituent elements. (Saussure 1983, 112; Saussure 1974, 113)

As an example of the distinction between signification and value, Saussure notes that 'The French word mouton may have the same meaning as the English word sheep; but it does not have the same value. There are various reasons for this, but in particular the fact that the English word for the meat of this animal, as prepared and served for a meal, is not sheep but mutton. The difference in value between sheep and mouton hinges on the fact that in English there is also another word mutton for the meat, whereas mouton in French covers both' (Saussure 1983, 114; Saussure 1974, 115-116).

Saussure's relational conception of meaning was specifically differential: he emphasized the differences between signs. Language for him was a system of functional differences and oppositions. 'In a language, as in every other semiological system, what distinguishes a sign is what constitutes it' (Saussure 1983, 119; Saussure 1974, 121). As John Sturrock points out, 'a one-term language is an impossibility because its single term could be applied to everything and differentiate nothing; it requires at least one other term to give it definition' (Sturrock 1979, 10). Advertising furnishes a good example of this notion, since what matters in 'positioning' a product is not the relationship of advertising signifiers to real-world referents, but the differentiation of each sign from the others to which it is related. Saussure's concept of the relational identity of signs is at the heart of structuralist theory. Structuralist analysis focuses on the structural relations which are functional in the signifying system at a particular moment in history. 'Relations are important for what they can explain: meaningful contrasts and permitted or forbidden combinations' (Culler 1975, 14).

Saussure emphasized in particular negative, oppositional differences between signs, and the key relationships in structuralist analysis are binary oppositions (such as nature/culture, life/death). Saussure argued that 'concepts... are defined not positively, in terms of their content, but negatively by contrast with other items in the same system. What characterizes each most exactly is being whatever the others are not' (Saussure 1983, 115; Saussure 1974, 117; my emphasis). This notion may initially seem mystifying if not perverse, but the concept of negative differentiation becomes clearer if we consider how we might teach someone who did not share our language what we mean by the term 'red'. We would be unlikely to make our point by simply showing them a range of different objects which all happened to be red - we would be probably do better to single out a red object from a sets of objects which were identical in all respects except colour. Although Saussure focuses on speech, he also noted that in writing, 'the values of the letter are purely negative and differential' - all we need to be able to do is to distinguish one letter from another (Saussure 1983, 118; Saussure 1974, 119-120). As for his emphasis on negative differences, Saussure remarks that although both the signified and the signifier are purely differential and negative when considered separately, the sign in which they are combined is a positive term. He adds that 'the moment we compare one sign with another as positive combinations, the term difference should be dropped... Two signs... are not different from each other, but only distinct. They are simply in opposition to each other. The entire mechanism of language... is based on oppositions of this kind and upon the phonic and conceptual differences they involve' (Saussure 1983, 119; Saussure 1974, 120-121).

Although the signifier is treated by its users as 'standing for' the signified, Saussurean semioticians emphasize that there is no necessary, intrinsic, direct or inevitable relationship between the signifier and the signified. Saussure stressed the arbitrariness of the sign (Saussure 1983, 67, 78; Saussure 1974, 67, 78) - more specifically the arbitrariness of the link between the signifier and the signified (Saussure 1983, 67; Saussure 1974, 67). He was focusing on linguistic signs, seeing language as the most important sign system; for Saussure, the arbitrary nature of the sign was the first principle of language (Saussure 1983, 67; Saussure 1974, 67) - arbitrariness was identified later by Charles Hockett as a key 'design feature' of language (Hockett 1958; Hockett 1960; Hockett 1965). The feature of arbitrariness may indeed help to account for the extraordinary versatility of language (Lyons 1977, 71). In the context of natural language, Saussure stressed that there is no inherent, essential, 'transparent', self-evident or 'natural' connection between the signifier and the signified - between the sound or shape of a word and the concept to which it refers (Saussure 1983, 67, 68-69, 76, 111, 117; Saussure 1974, 67, 69, 76, 113, 119). Note that Saussure himself avoids directly relating the principle of arbitrariness to the relationship between language and an external world, but that subsequent commentators often do, and indeed, lurking behind the purely conceptual 'signified' one can often detect Saussure's allusion to real-world referents (Coward & Ellis 1977, 22). In language at least, the form of the signifier is not determined by what it signifies: there is nothing 'treeish' about the word 'tree'. Languages differ, of course, in how they refer to the same referent. No specific signifier is 'naturally' more suited to a signified than any other signifier; in principle any signifier could represent any signified. Saussure observed that 'there is nothing at all to prevent the association of any idea whatsoever with any sequence of sounds whatsoever' (Saussure 1983, 76; Saussure 1974, 76); 'the process which selects one particular sound-sequence to correspond to one particular idea is completely arbitrary' (Saussure 1983, 111; Saussure 1974, 113).

The arbitrariness of the sign is a radical concept because it proposes the autonomy of language in relation to reality. The Saussurean model, with its emphasis on internal structures within a sign system, can be seen as supporting the notion that language does not 'reflect' reality but rather constructs it. We can use language 'to say what isn't in the world, as well as what is. And since we come to know the world through whatever language we have been born into the midst of, it is legitimate to argue that our language determines reality, rather than reality our language' (Sturrock 1986, 79). In their book The Meaning of Meaning, Ogden and Richards criticized Saussure for 'neglecting entirely the things for which signs stand' (Ogden & Richards 1923, 8). Later critics have lamented his model's detachment from social context (Gardiner 1992, 11). Robert Stam argues that by 'bracketing the referent', the Saussurean model 'severs text from history' (Stam 2000, 122). We will return to this theme of the relationship between language and 'reality' in our discussion of 'modality and representation'.

The arbitrary aspect of signs does help to account for the scope for their interpretation (and the importance of context). There is no one-to-one link between signifier and signified; signs have multiple rather than single meanings. Within a single language, one signifier may refer to many signifieds (e.g. puns) and one signified may be referred to by many signifiers (e.g. synonyms). Some commentators are critical of the stance that the relationship of the signifier to the signified, even in language, is always completely arbitrary (e.g. Lewis 1991, 29). Onomatopoeic words are often mentioned in this context, though some semioticians retort that this hardly accounts for the variability between different languages in their words for the same sounds (notably the sounds made by familiar animals) (Saussure 1983, 69; Saussure 1974, 69).

Saussure declares that 'the entire linguistic system is founded upon the irrational principle that the sign is arbitrary'. This provocative declaration is followed immediately by the acknowledgement that 'applied without restriction, this principle would lead to utter chaos' (Saussure 1983, 131; Saussure 1974, 133). If linguistic signs were to be totally arbitrary in every way language would not be a system and its communicative function would be destroyed. He concedes that 'there exists no language in which nothing at all is motivated' (ibid.). Saussure admits that 'a language is not completely arbitrary, for the system has a certain rationality' (Saussure 1983, 73; Saussure 1974, 73). The principle of arbitrariness does not mean that the form of a word is accidental or random, of course. Whilst the sign is not determined extralinguistically it is subject to intralinguistic determination. For instance, signifiers must constitute well-formed combinations of sounds which 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 combination of two existing signs. Saussure introduces a distinction between degrees of arbitrariness:

The fundamental principle of the arbitrary nature of the linguistic sign does not prevent us from distinguishing in any language between what is intrinsically arbitrary - that is, unmotivated - and what is only relatively arbitrary. Not all signs are absolutely arbitrary. In some cases, there are factors which allow us to recognize different degrees of arbitrariness, although never to discard the notion entirely. The sign may be motivated to a certain extent (Saussure 1983, 130; Saussure 1974, 131; original emphasis, see also following pages)

Here then Saussure modifies his stance somewhat and refers to signs as being 'relatively arbitrary'. Some subsequent theorists (echoing Althusserian Marxist terminology) refer to the relationship between the signifier and the signified in terms of 'relative autonomy' (Tagg 1988, 167; Lechte 1994, 150). The relative conventionality of relationships between signified and signifier is a point to which I return below.

It should be noted that whilst 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, arbitrarily established, unlike historical inventions such as Morse Code. Nor does the arbitrary nature of the sign make it socially 'neutral' or materially 'transparent' - for example, in Western culture 'white' has come to be a privileged signifier (Dyer 1997). Even in the case of the 'arbitrary' colours of traffic lights, the original choice of red for 'stop' was not entirely arbitrary, since it already carried relevant associations with danger. As Lévi-Strauss noted, the sign is arbitrary a priori but ceases to be arbitrary a posteriori - after the sign has come into historical existence it cannot be arbitrarily changed (Lévi-Strauss 1972, 91). As part of its social use within a code (a term which became fundamental amongst post-Saussurean semioticians), every sign acquires a history and connotations of its own which are familiar to members of the sign-users' culture. Saussure remarked that although the signifier 'may seem to be freely chosen', from the point of view of the linguistic community it is 'imposed rather than freely chosen' because 'a language is always an inheritance from the past' which its users have 'no choice but to accept' (Saussure 1983, 71-72; Saussure 1974, 71). Indeed, 'it is because the linguistic sign is arbitrary that it knows no other law than that of tradition, and [it is] because it is founded upon tradition that it can be arbitrary' (Saussure 1983, 74; Saussure 1974, 74). The arbitrariness principle does not, of course mean that an individual can arbitrarily choose any signifier for a given signified. The relation between a signifier and its signified is not a matter of individual choice; if it were then communication would become impossible. 'The individual has no power to alter a sign in any respect once it has become established in the linguistic community' (Saussure 1983, 68; Saussure 1974, 69). From the point-of-view of individual language-users, language is a 'given' - we don't create the system for ourselves. Saussure refers to the language system as a non-negotiable 'contract' into which one is born (Saussure 1983, 14; Saussure 1974, 14) - although he later problematizes the term (ibid., 71). The ontological arbitrariness which it involves becomes invisible to us as we learn to accept it as 'natural'.

The Saussurean legacy of the arbitrariness of signs leads semioticians to stress that the relationship between the signifier and the signified is conventional - dependent on social and cultural conventions. This is particularly clear in the case of the linguistic signs with which Saussure was concerned: a word means what it does to us only because we collectively agree to let it do so. Saussure felt that the main concern of semiotics should be 'the whole group of systems grounded in the arbitrariness of the sign'. He argued that: 'signs which are entirely arbitrary convey better than others the ideal semiological process. That is why the most complex and the most widespread of all systems of expression, which is the one we find in human languages, is also the most characteristic of all. In this sense, linguistics serves as a model for the whole of semiology, even though languages represent only one type of semiological system' (Saussure 1983, 68; Saussure 1974, 68). He did 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 (Saussure 1983, 15, 17, 68, 74; Saussure 1974, 16, 17, 68, 73). Saussure added that 'any means of expression accepted in a society rests in principle upon a collective habit, or on convention - which comes to the same thing' (Saussure 1983, 68; Saussure 1974, 68). However, whilst purely conventional signs such as words are quite independent of their referents, other less conventional forms of signs are often somewhat less independent of them. Nevertheless, since the arbitary nature of linguistic signs is clear, those who have adopted the Saussurean model have tended to avoid 'the familiar mistake of assuming that signs which appear natural to those who use them have an intrinsic meaning and require no explanation' (Culler 1975, 5).

At around the same time as Saussure was formulating his model of the sign, of 'semiology' and of a structuralist methodology, across the Atlantic independent work was also in progress as the pragmatist philosopher and logician Charles Sanders Peirce formulated his own model of the sign, of 'semiotic' and of the taxonomies of signs. In contrast to Saussure's model of the sign in the form of a 'self-contained dyad', Peirce offered a triadic model:

• The Representamen: the form which the sign takes (not necessarily material);
• An Interpretant: not an interpreter but rather the sense made of the sign;
• An Object: to which the sign refers.

'A sign... [in the form of a representamen] is something which stands to somebody for something in some respect or capacity. It addresses somebody, that is, creates in the mind of that person an equivalent sign, or perhaps a more developed sign. That sign which it creates I call the interpretant of the first sign. The sign stands for something, its object. It stands for that object, not in all respects, but in reference to a sort of idea, which I have sometimes called the ground of the representamen' (Peirce 1931-58, 2.228). The interaction between the representamen, the object and the interpretant is referred to by Peirce as 'semiosis' (ibid., 5.484). Within Peirce's model of the sign, the traffic light sign for 'stop' would consist of: a red light facing traffic at an intersection (the representamen); vehicles halting (the object) and the idea that a red light indicates that vehicles must stop (the interpretant).

Peirce's model of the sign includes an object or referent - which does not, of course, feature directly in Saussure's model. The representamen is similar in meaning to Saussure's signifier whilst the interpretant is similar in meaning to the signified (Silverman 1983, 15). However, the interpretant has a quality unlike that of the signified: it is itself a sign in the mind of the interpreter. Peirce noted that 'a sign... addresses somebody, that is, creates in the mind of that person an equivalent sign, or perhaps a more developed sign. The sign which it creates I call the interpretant of the first sign' (Peirce 1931-58, 2.228). Umberto Eco uses the phrase 'unlimited semiosis' to refer to the way in which this could lead (as Peirce was well aware) to a series of successive interpretants (potentially) ad infinitum (ibid., 1.339, 2.303). Elsewhere Peirce added that 'the meaning of a representation can be nothing but a representation' (ibid., 1.339). Any initial interpretation can be re-interpreted. That a signified can itself play the role of a signifier is familiar to anyone who uses a dictionary and finds themselves going beyond the original definition to look up yet another word which it employs. This concept can be seen as going beyond Saussure's emphasis on the value of a sign lying in its relation to other signs and it was later to be developed more radically by poststructuralist theorists. Another concept which is alluded to within Peirce's model which has been taken up by later theorists but which was explicitly excluded from Saussure's model is the notion of dialogical thought. It stems in part from Peirce's emphasis on 'semiosis' as a process which is in distinct contrast to Saussure's synchronic emphasis on structure (Peirce 1931-58, 5.484, 5.488). Peirce argued that 'all thinking is dialogic in form. Your self of one instant appeals to your deeper self for his assent' (Peirce 1931-58, 6.338). This notion resurfaced in a more developed form in the 1920s in the theories of Mikhail Bakhtin (Bakhtin 1981). One important aspect of this is its characterization even of internal reflection as fundamentally social.

Peirce, clearly fascinated by tripartite structures, made a phenomenological distinction between the sign itself [or the representamen] as an instance of 'Firstness', its object as an instance of 'Secondness' and the interpretant as an instance of 'Thirdness'. Such unfamiliar terms are relatively modest examples of Peircean coinages, and the complexity of his terminology and style has been a factor in limiting the influence of a distinctively Peircean semiotics.

Variants of Peirce's triad are often presented as 'the semiotic triangle' (as if there were only one version). Here is a version which is quite often encountered and which changes only the unfamiliar Peircean terms (Nöth 1990, 89):

• Sign vehicle: the form of the sign;
• Sense: the sense made of the sign;
• Referent: what the sign 'stands for'.

One fairly well-known semiotic triangle is that of Ogden and Richards, in which the terms used are (a) 'symbol', (b) 'thought or reference' and (c) 'referent' (Ogden & Richards 1923, 14). The broken line at the base of the triangle is intended to indicate that there is not necessarily any observable or direct relationship between the sign vehicle and the referent. Unlike Saussure's abstract signified (which is analogous to term B rather than to C) the referent is an 'object'. This need not exclude the reference of signs to abstract concepts and fictional entities as well as to physical things, but Peirce's model allocates a place for an objective reality which Saussure's model did not directly feature (though Peirce was not a naive realist, and argued that all experience is mediated by signs). Note, however, that Peirce emphasized that 'the dependence of the mode of existence of the thing represented upon the mode of this or that representation of it... is contrary to the nature of reality' (Peirce 1931-58, 5.323). The inclusion of a referent in Peirce's model does not automatically make it a better model of the sign than that of Saussure. Indeed, as John Lyons notes:

There is considerable disagreement about the details of the triadic analysis even among those who accept that all three components, A, B and C, must be taken into account. Should A be defined as a physical or a mental entity? What is the psychological or ontological status of B? Is C something that is referred to on a particular occasion? Or is it the totality of things that might be referred to by uttering the sign...? Or, yet a third possibility, is it some typical or ideal representative of this class? (Lyons 1977, 99)

The notion of the importance of sense-making (which requires an interpreter - though Peirce doesn't feature that term in his triad) has had a particular appeal for communication and media theorists who stress the importance of the active process of interpretation, and thus reject the equation of 'content' and meaning. Many of these theorists allude to semiotic triangles in which the interpreter (or 'user') of the sign features explicitly (in place of 'sense' or 'interpretant'). This highlights the process of semiosis (which is very much a Peircean concept). The meaning of a sign is not contained within it, but arises in its interpretation. Whether a dyadic or triadic model is adopted, the role of the interpreter must be accounted for - either within the formal model of the sign, or as an essential part of the process of semiosis. David Sless declares that 'statements about users, signs or referents can never be made in isolation from each other. A statement about one always contains implications about the other two' (Sless 1986, 6). Paul Thibault argues that the interpreter features implicitly even within Saussure's apparently dyadic model (Thibault 1997, 184).

Note that semioticians make a distinction between a sign and a 'sign vehicle' (the latter being a 'signifier' to Saussureans and a 'representamen' to Peirceans). The sign is more than just a sign vehicle. The term 'sign' is often used loosely, so that this distinction is not always preserved. In the Saussurean framework, some references to 'the sign' should be to the signifier, and similarly, Peirce himself frequently mentions 'the sign' when, strictly speaking, he is referring to the representamen. It is easy to be found guilty of such a slippage, perhaps because we are so used to 'looking beyond' the form which the sign happens to take. However, to reiterate: the signifier or representamen is the form in which the sign appears (such as the spoken or written form of a word) whereas the sign is the whole meaningful ensemble.

Whereas Saussure emphasized the arbitrary nature of the (linguistic) sign, most semioticians stress that signs differ in how arbitrary/conventional (or by contrast 'transparent') they are. Symbolism reflects only one form of relationship between signifiers and their signifieds. Whilst Saussure did not offer a typology of signs, Charles Peirce was a compulsive taxonomist and he offered several logical typologies (Peirce 1931-58, 1.291, 2.243). However, his divisions and subdivisions of signs are extraordinarily elaborate: indeed, he offered the theoretical projection that there could be 59,049 types of signs! Peirce himself noted wryly that this calculation 'threatens a multitude of classes too great to be conveniently carried in one's head', adding that 'we shall, I think, do well to postpone preparation for further divisions until there be a prospect of such a thing being wanted' (Peirce 1931-58, 1.291). However, even his more modest proposals are daunting: Susanne Langer commented that 'there is but cold comfort in his assurance that his original 59,049 types can really be boiled down to a mere sixty-six' (Langer 1951, 56). Unfortunately, the complexity of such typologies rendered them 'nearly useless' as working models for others in the field (Sturrock 1986, 17). However, one of Peirce's basic classifications (first outlined in 1867) has been very widely referred to in subsequent semiotic studies (Peirce 1931-58, 1.564). He regarded it as 'the most fundamental' division of signs (ibid., 2.275). It is less useful as a classification of distinct 'types of signs' than of differing 'modes of relationship' between sign vehicles and their referents (Hawkes 1977, 129). Note that in the subsequent account, I have continued to employ the Saussurean terms signifier and signified, even though Peirce referred to the relation between the 'sign' (sic) and the object, since the Peircean distinctions are most commonly employed within a broadly Saussurean framework. Such incorporation tends to emphasize (albeit indirectly) the referential potential of the signified within the Saussurean model. Here then are the three modes together with some brief definitions of my own and some illustrative examples:

	Symbol/symbolic: a mode in which the signifier does not resemble the signified but which is fundamentally arbitrary or purely conventional - so that the relationship must be learnt: e.g. language in general (plus specific languages, alphabetical letters, punctuation marks, words, phrases and sentences), numbers, morse code, traffic lights, national flags;
	Icon/iconic: a mode in which the signifier is perceived as resembling or imitating the signified (recognizably looking, sounding, feeling, tasting or smelling like it) - being similar in possessing some of its qualities: 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;
	Index/indexical: a mode in which the signifier is not arbitrary but is directly connected in some way (physically or causally) to the signified - 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, catchphrase) and indexical words ('that', 'this', 'here', 'there').

The three forms are listed here in decreasing order of conventionality. Symbolic signs such as language are (at least) highly conventional; iconic signs always involve some degree of conventionality; indexical signs 'direct the attention to their objects by blind compulsion' (Peirce 1931-58, 2.306). Indexical and iconic signifiers can be seen as more constrained by referential signifieds whereas in the more conventional symbolic signs the signified can be seen as being defined to a greater extent by the signifier. Within each form signs also vary in their degree of conventionality. Other criteria might be applied to rank the three forms differently. For instance, Hodge and Kress suggest that indexicality is based on an act of judgement or inference whereas iconicity is closer to 'direct perception' making the highest 'modality' that of iconic signs. Note that the terms 'motivation' (from Saussure) and 'constraint' are sometimes used to describe the extent to which the signified determines the signifier. The more a signifier is constrained by the signified, the more 'motivated' the sign is: iconic signs are highly motivated; symbolic signs are unmotivated. The less motivated the sign, the more learning of an agreed convention is required. Nevertheless, most semioticians emphasize the role of convention in relation to signs. As we shall see, even photographs and films are built on conventions which we must learn to 'read'. Such conventions are an important social dimension of semiotics.

Peirce and Saussure used the term 'symbol' differently from each other. Whilst nowadays most theorists would refer to language as a symbolic sign system, Saussure avoided referring to linguistic signs as 'symbols', since the ordinary everyday use of this term refers to examples such as a pair of scales (signifying justice), and he insisted that such signs are 'never wholly arbitrary. They are not empty configurations'. They 'show at least a vestige of natural connection' between the signifier and the signified - a link which he later refers to as 'rational' (Saussure 1983, 68, 73; Saussure 1974, 68, 73). Whilst Saussure focused on the arbitrary nature of the linguistic sign, a more obvious example of arbitrary symbolism is mathematics. Mathematics does not need to refer to an external world at all: its signifieds are indisputably concepts and mathematics is a system of relations (Langer 1951, 28).

For Peirce, a symbol is 'a sign which refers to the object that it denotes by virtue of a law, usually an association of general ideas, which operates to cause the symbol to be interpreted as referring to that object' (Peirce 1931-58, 2.249). We interpret symbols according to 'a rule' or 'a habitual connection' (ibid., 2.292, 2.297, 1.369). 'The symbol is connected with its object by virtue of the idea of the symbol-using animal, without which no such connection would exist' (ibid., 2.299). It 'is constituted a sign merely or mainly by the fact that it is used and understood as such' (ibid., 2.307). It 'would lose the character which renders it a sign if there were no interpretant' (ibid., 2.304). A symbol is 'a conventional sign, or one depending upon habit (acquired or inborn)' (ibid., 2.297). 'All words, sentences, books and other conventional signs are symbols' (ibid., 2.292). Peirce thus characterizes linguistic signs in terms of their conventionality in a similar way to Saussure. In a rare direct reference to the arbitrariness of symbols (which he then called 'tokens'), he noted that they 'are, for the most part, conventional or arbitrary' (ibid., 3.360). 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' (ibid., 4.447). He adds elsewhere that 'a symbol... fulfills its function regardless of any similarity or analogy with its object and equally regardless of any factual connection therewith' but solely because it will be interpreted as a sign (ibid., 5.73; original emphasis).

Turning to icons, Peirce declared that an iconic sign represents its object 'mainly by its similarity' (Peirce 1931-58, 2.276). A sign is an icon 'insofar as it is like that thing and used as a sign of it' (ibid., 2.247). Indeed, he originally termed such modes, 'likenesses' (e.g. ibid., 1.558). He added that 'every picture (however conventional its method)' is an icon (ibid., 2.279). Icons have qualities which 'resemble' those of the objects they represent, and they 'excite analogous sensations in the mind' (ibid., 2.299; see also 3.362). Unlike the index, 'the icon has no dynamical connection with the object it represents' (ibid.). Just because a signifier resembles that which it depicts does not necessarily make it purely iconic. The philosopher Susanne Langer argues that 'the picture is essentially a symbol, not a duplicate, of what it represents' (Langer 1951, 67). Pictures resemble what they represent only in some respects. What we tend to recognize in an image are analogous relations of parts to a whole (ibid., 67-70). For Peirce, icons included 'every diagram, even although there be no sensuous resemblance between it and its object, but only an analogy between the relations of the parts of each' (Peirce 1931-58, 2.279). 'Many diagrams resemble their objects not at all in looks; it is only in respect to the relations of their parts that their likeness consists' (ibid., 2.282). Even the most 'realistic' image is not a replica or even a copy of what is depicted. We rarely mistake a representation for what it represents.