Icon, Index, Symbol
Relationship to the Object
Peirce proposed three forms of the sign related in different ways to the Object, and labelled these Icon, Index and Symbol. Prototypically, an Iconic sign is one that relates to its Object via a relationship of likeness (Peirce, 1931-1958: 2.276). It represents the Object by looking like it, as a portrait might its sitter, an engineering diagram its realised article or the Underground map of London its system of train lines. These representations may be different kinds of likeness, but their form is recognisably related to their Object, and all are labelled as Iconic by Peirce. Under the heading of Iconic signs, Peirce also placed the metaphor, on the grounds that metaphors posit a relationship of likeness between one entity and another. Peirce’s classification of a metaphor as an Iconic sign was not motivated by the fact that many metaphors are dead—that is they have ceased to rely on unexpected pairings of entities. Rather, he saw it as a logical outcome of the fact that targets resemble their sources in a relationship of likeness. Unfortunately, Pierce did not write extensively about metaphor, and there is little further guidance or support for his decision other than his mention of metaphor as one type of Icon. Other semioti- cians have recognised that the disparity between the source and target domains may require interpretation through codes and rules that suggest that there the metaphor can also be Symbolic. Some scholars (e.g. Haley, 1995; Ponzio, 2010; Sorensen, 2011) have argued that both sign forms are present in metaphor. Chandler (2001: np) writes:
The basis in resemblance suggests that metaphor involves the iconic mode.
However, to the extent that such a resemblance is oblique, we may think of
metaphor as symbolic. (Emphasis in original)
An Indexical sign is characterised as being in a relationship of contiguity with its Object (Peirce, 1931-1958: 2.276) rather than a relationship of resemblance, as is the case for Iconic signs. Peirce’s notion of Indexical signs can be elusive to grasp, as this relationship of contiguity can be manifest in a number of different ways (Eco, 1976; Grutman, 2010; Lock, 1997; Sorensen, 2011). Contiguity can take the form of cause-and-effect relations, such as that between smoke and fire or footprints and the presence of a person. It can be realised in metonymic and synecdochic relations, where a part of an entity stands for the whole, or a single instance can stand for an entire class, such as a sign showing a coffee cup with a red line through where the coffee is a metonym for “all drinks”. Whole- for-part relations are also synecdochic, and thus Indexical, shown in language by expressions such as “BP announced”, where the organisation here stands for a person/people within it. A sign can be Indexical when it points to or indicates the presence of something else, so that arrows and pub signs, for example, are Indexical. These prototypical instances are relatively straightforward, but some signs are more difficult to locate, or show complex relationships with their Object, for example, a Jaguar car can be seen as both an Index and a Symbol of wealth. A photograph may be regarded as an Icon because it shows a resemblance to its Object, or an Index because it represents a point-to-point correspondence with reality (Chandler, 2007: 38-39).
A Symbolic sign is one that is related to its Object only arbitrarily and by convention (Peirce, 1931-1958: 2.249). It neither looks like the Object, nor is it related diagrammatically, nor is it associated causally or through part-whole relations. It is understood only through social agreement; it has acquired meaning through the development of conventional systems, which have to be learned to make sense to the receiver. For Peirce, Symbols had meaning through rules and laws rather than through instinct and observation. Words are prototypical Symbolic signs, as they (generally) have no relation to their Object apart from that which has been conventionally agreed. Peirce’s view largely accords with that of de Saussure, which is that language is a set of arbitrary, rather than motivated signs. Other Symbolic signs include mathematical symbols or literary or artistic symbols (such as a lamb in a painting, intended to represent innocence).
I describe the definitions above as prototypical because signs need not be (indeed seldom are) pure versions of Icons, Indexes or Symbols. Most Symbols have at root an Icon and/or Index—for example, red is considered a symbol of danger through convention, but this convention is likely to be based on an association of red with blood, or fire, or the fact that red is a highly visible colour. These associations are primarily in Indexical relationship with the concept “danger”. Similarly, (alphabetically) written signs, which are seen as arbitrary and Symbolic by Peirce’s definition, and which de Saussure argued only have meaning in relation to each other, are often originally based on systems of pictograms that have an Iconic basis—for example, a tree or a snake (Singleton, 2000). A single sign can be argued as Iconic, Indexical and Symbolic. If I pursue the example of Jaguar, the bonnet ornament on Jaguar cars is Iconic in that it is a direct likeness of the animal; however, we attach Indexical associations of power, speed, sleekness and beauty, which we transfer from the animal to the car itself. But we also understand the ornament as a Symbol of the car through the cultural knowledge that the signifier “Jaguar” represents not only the signified “animal” but also “a car manufacturer”. Further, we might interpret the bonnet ornament as an Index of wealth, or indeed a Symbol of wealth. As mentioned above, writers developing Peirce’s trichotomy in the area of metaphor stress the interconnectivity of the three sign forms in creating a metaphor (Abrams, 2002; Haley, 1995; Ponzio, 2010; Sorensen, 2011). As Merrell (2001: 37) writes:
Now, everything I have written in this section suggests that a sign can be in varying degrees iconic, indexical, and symbolic, all at the same time. A sign’s evincing one sign type does not preclude its manifesting some other sign type as well. There are no all-or-nothing categories with respect to signs. As one sign type is, another sign type can become, and what that sign was may become of the nature of the first sign that the second sign now is.
Merrell illustrates that Peirce’s three sign modes are separate only in theory: not only can a sign manifest more than one sign type, but also one sign type can become another. This quality of mutability is potentially significant for my diachronic study of representation.