Arbitrary Signs

A couple of weeks ago, I mused on whether Saussure would have agreed with Barr’s rejection of the idea that different mentalities are built into different languages. In his superb book on Saussure, Jonathan Culler provides further evidence that Saussure would not agree with Barr.

Culler’s argument, to put it in a nutshell, is that understanding Saussure requires us to understand what he means by the “arbitrary nature of the sign,” and specifically to understand that Saussure means the arbitrary nature of the sign , not merely the arbitrary nature of the signifier .

Saussure defines the sign as a unified duality, consisting of the signifier – a signifying auditory of visual form – and the signified – the concept associated with the signifier. Dog is a sign in English because it unites the sound d-o-g with the concept “dog.” (The signifier is not exactly the sound, but equating the two is sufficiently accurate for my purposes.) Saussure knows that some signs are imitations of nature [ buzz ] and that some signs are partially motivated by pre-existing signs in the linguistic system [ type-writer ]. But these are exceptions, and when Saussure talks about the arbitrary nature of the sign, he at least means that the assignment of the signifier dog to the signified concept “dog” is not based on any “natural or inevitable link between the signifier and the signified.” English-speakers could have chosen any number of signifiers to denote the concept “dog,” and in fact French and German and Swahili speakers employ different signifiers.

Culler points out that if this is all Saussure means by the “arbitrary nature of the sign,” his point is obvious, commonplace, and hardly worth the emphasis Saussure places on it. But this is not all he has in mind; to this point, we are only describing the arbitrary nature of the “signifier.” But the sign is a duality of signified/signifier, and Saussure argues that both aspects of the sign are arbitrary. That is, not only do English and French speakers differ in assigning different sound-combinations to designate a set of universal, common pre-existing concepts. The concepts of different languages differ. This doesn’t at all mean that there is no world “out there” to which our concepts refer; but it means that we slice and dice the world in different ways, articulating the world by signs, that is by arbitrarily chosen signified/signifier combinations.

Saussure denies that language is simply a “nomenclature,” that is, a set of names for items that are already articulated. If language were a nomenclature, then it would be easy to translate from one language to another. Aimer would move directly into English as love . In fact, however, in translating “aimer” to English, “one must choose between ‘to like’ and ‘to love.’” Change within a language also works not by a shift of a signifier from one to the next concept; rather, “the concept attached to the signifier silly was continually shifting its boundaries, gradually changing its semantic shape, articulating the world in different ways from one period to the next.”

Culler summarizes Saussure’s point: “the signified associated with a signifier can take any form; there is no essential core of meaning that it must retain in order to count as the proper signified for that signifier. The fact that the relation between signifier and signified is arbitrary means, then, that since there are no fixed universal concepts or fixed universal signifiers, the signified itself is arbitrary, and so is the signifier.”

Back to Barr: On one level Saussure does support Barr’s contention; Barr’s point is precisely that when the semantic load of a signifier changes over time, “there is no essential core of meaning that it must retain.” Silly does not retain its original meaning of “happy, blessed, and pious” as it moves to “simple, foolish, perhaps even stupid.” (I’m not sure about this; lexically, this may be the case, but in actual usage, it’s entirely possible that in some contexts some writers will want the original meaning to shine through its current meaning.) On another level, though, Saussure seems to disagree with Barr; Saussure does seem to support the notion that there is a basic connection between linguistic and conceptual apparatus. Barr is no doubt right to contest some of the ways that this connection has been articulated, but Saussure supports the notion that speakers of language A articulate the world (articulate it conceptually ) differently from speakers of language B.

Culler argues that Saussure’s distinction of langue and parole , which seems arbitrary and unfounded to some interpreters, is in fact a direct implication of the arbitrary nature of the sign. And he says that Saussure’s relational notion of the “value” of signs is also dependent on his prior arguments for the arbitrary nature of signs. I’ll summarize that argument in a later post.