Where Charles Saunders Peirce
took a logical perspective in
what he called "semiotics", attempting to establish a "formal doctrine of
signs", Swiss linguist Ferdinand de Saussure
(1857-1913) viewed what he called "semiology" as "a science which studies
the role of signs as part of social life". The term "semiotics" is now generally
applied to Saussure's work as well, recognizing the substantial similarity
between what the two theorists were trying to achieve.
One contrast between Peirce and Saussure is in their models of the sign. Peirce's model is triadic while Saussure's is dyadic, having two parts instead of three:
Saussure thought of both the signifier and the signified as purely psychological phenomena, with any connections to the physical world presumably mediated through the sensorimotor system of the interpreter. Later thinkers have used the term "signifier" also to refer to the material form of a sign, like Peirce's representamen. The relationship between signifier and signified is called signification, and together they constitute a sign.
One point worth mentioning is that a particular sign consists of the combination of signifier and signified; when either part is changed, the whole becomes a different sign. Thus when the "Enter" key signifies the end of a line of text, that is one sign; but when it is used to confirm a dialog, that is a different sign.
Notice that the interpretation of signifiers as material forms conveniently allows us to speak of user interface devices and display elements directly as signifiers, rather than having to continually (and awkwardly) refer to their psychological effects on users. Saussure's way of thinking, though, usefully draws attention to the fact that signification ultimately depends on perceptions, not merely on physical form. We should keep this in mind when designing interfaces, since not all users will perceive things the same way. Some users may not have the experience or sensory capacity necessary to make certain perceptual distinctions, and some may not receive certain sensory input at all. A color- blind person, for example, may have no access to the connotation of "warning" implied by the use of red as opposed to green in an interface element.
The nature of Saussure's signified as a mental concept, not a thing in the world, points to another important distinction, between what a user thinks a sign means and what the designer intended it to mean. For example, a user entering a value into a text field might think a press of the "Enter" key signifies an action like "submit" when in fact it merely signifies the end of the line (or vice-versa). It is incumbent upon designers of user interfaces to prevent this kind of error when possible, and to anticipate and gracefully handle errors that aren't prevented.
Within Saussure's model it is easy to conceptualize chains of signification: the signified of one sign simply becomes the signifier of another. For example, an error message that reads "No such file" directly signifies that no file exists with the name the user just entered; this fact might signify that the user made a typographical error, or perhaps that the file has been deleted. Generally, the fewer levels of indirection the user must traverse, the better. Naturally, software is limited in its ability to determine what the user did wrong, because to do so generally requires knowing the user's intention. Nevertheless, in many cases sensitive designers can improve the situation by providing more useful feedback.
According to Saussure, chains of signification are the ordinary state of affairs: signs usually refer to other signs, not to physical objects. For him, meaning is relational as opposed to referential; that is, signs make sense only within a system of related signs, not because of their inherent properties or their reference to material things.
In a typical GUI, for example, a file name signifies a file, but realistically a file is merely another sign. Any reference to the physical world is quite indirect, through many other signs at different levels: file system organization, its underlying implementation, and so on down to bits, which are themselves abstract and represented variously as the states of transistors or magnetic storage media. Of course, the user doesn't need to know all these details to understand and use files, which is precisely the point: the relations within the sign system of the user interface are of primary concern, not the connection to material reality.
Structuralism is the tradition, beginning with Saussure, of viewing signs as functioning within systems, and analyzing the structure of these systems by identifying their constituent parts and the relations among them. This outlook has widely informed study in the humanities and social sciences, beginning with linguistics and extending to literature, anthropology, and sociology, and other fields. The influence of structuralism is so widespread that some view it as the basis for all of semiotics, or even as synonymous with it.