(This expository post first came out in (mass)think! in 6/2007.)
In contrast to linguistics in his time, which made language secondary to some other object of inquiry, in the Course in General Linguistics (1916; published posthumously), Ferdinand de Saussure treats language itself as the object of study by “tak[ing] the study of linguistic structure as his primary concern” (16, 9). Linguistic structure, “only one part of language,” albeit essential, is, according to Saussure, both “a social product of our language faculty” and “a body of necessary conventions adopted by society to enable [its] members […] to use their language faculty” (9-10). It is “language minus speech, […] the whole set of linguistic habits which enables the speaker to understand and to make himself understood” (77). That is, linguistic structure refers to the rules of (a) language, minus its specific articulation (speech). A language, in turn, is a “well-defined entity, […] locali[zable],” i.e. “ha[ving] a particular place in the realm of human affairs,” “in that particular section of the speech circuit where sound patterns [signifiers] are associated with concepts [signifieds]” (14, 15). Language, in other words, comprises the whole system of signifiers and signifieds and the linguistic structure, the rules of their association.
Saussure asserts that it is “the social part of language, external to the individual, […] exist[ing] in virtue of a kind of contract agreed between the members of a community,” in which the individual needs apprenticeship (if she wants to be able to use it) (14, 15). This social part is a specific, hence homogeneous, compartment/region of language in general (or “the totality of facts of language”), which is heterogeneous (14). Linguistic signs are therefore “not abstractions. The associations, ratified by collective agreement, which go to make up the language are realities localized in the brain” (15). Moreover, they are “tangible,” fixable by writing in “conventional images” since “there is only the sound pattern, and this can be represented by one constant visual image” (15). A language, while not the same as language in general, is also not simply speech, which “is an individual act of the will and the intelligence,” i.e. the particular application of an individual’s apprenticeship, a particular articulation of language (14). In laying out all these components and relations, Saussure portrays language as a social institution.
Rather than a nomenclature (i.e. language as the naming of things/ideas), for Saussure, language is a sign system. Linguistics (Saussurian linguistics = semiotics) is thus but a part of the study of signs (their nature, the laws governing them) in general, semiology; inversely, semiology is the application of the techniques of semiotics to other cultural domains, treating them as a system of signs (15). A linguistic sign, Saussure claims, is a link between the signifier and the signified. The signifier refers to the sound pattern, “not actually a sound [… but] the hearer’s psychological impression of a sound, as given to him by the evidence of his senses” (66). That is, the signifier is the word, or, more precisely, the sound one hears, or the sound image that registers in one’s brain, when a word (such as tree) is uttered. The signified, on the other hand, refers to the concept or the idea linked to (not just conveyed by and not that which causes) the sound pattern, i.e. the idea of the tree one forms in his head. These two are yet different from the referent, i.e. the thing linked to the signifier and/or the signified, e.g. the “actual” tree one can see, touch … The signifier and the signified together make up the sign.
Perhaps the most groundbreaking assertion that Saussure makes has to do with the arbitrariness of the sign. Saussure claims that “the link between [signifier] and [signified] is arbitrary,” i.e. there is no internal connection between the two (67). This means, first, that there is no (natural) reason why a particular signifier is related to a particular signified. The signifier, in other words, is unmotivated (69). There is no reason, for example, why we call a tree (or, more precisely: why we refer to the idea of a tree) tree. “This is demonstrated by differences between languages, and even by the existence of different languages” (68). Secondly, this means that signifieds themselves are arbitrary. Certain things, or, more accurately, certain signifieds (e.g. colors, or the signifieds of fleuve/rivière in French, two different things, depending on the direction of the flow) exist in some languages, but not in others (in English, there is only the signified for river). In other words, there is no given universal set of ideas. The linguistic system itself creates the “meaning.”