A language thus has this curious and striking feature. It has no immediately perceptible entities. And yet one cannot doubt that they exist, or that the interplay of these units is what constitutes linguistic structure. That is undoubtedly a characteristic which distinguishes languages from all other semiological institutions.
Ferdinand de Saussure (Course in General Linguistics, 105)
After having investigated the physiological mechanisms of speech, Saussure turns to consider the nature of linguistic signs. In his analysis we find a clear formulation of the first principle of his “new” linguistics, namely that the sign is arbitrary.
At first glance this statement seems to suggest an argument for an evolutionary linguistics, but Saussure draws our attention to the fact that “this very same factor tends to protect a language from any attempt to change it.” (73) As a system of arbitrary signs, there is no rational basis for language; paradoxically this makes language itself “inaccessible” to reason.
There is no logical reason we should prefer “sister” over “hermana,” and thus no grounds for argument about change. The great number of signs necessary to constitute a language (and the complex character of the linguistic system itself) also contribute to the force of collective inertia resisting linguistic innovation:
“At any time a language belongs to all its users. It is a facility unreservedly available throughout a whole community… [it] is something in which everyone participates all the time, and that is why it is constantly open to the influence of all. This key fact is by itself sufficient explain why a linguistic revolution is impossible.” (74)
A community naturally exerts a restrictive, conservative influence upon a language. Nonetheless, the passage of time also allows linguistic signs to be changed with “some rapidity” — hence both variability (diachrony) and invariability (synchrony) are characteristic of the linguistic sign. These two characteristics are “intimately connected” — a sign is only subject to change because it “continues through time.”
Thus the principle of change is based upon the principle of continuity; infidelity to the past is always relative. A misunderstanding is possible here: the changes Saussure is speaking of always “result in a shift in the relationship between signal and signification.” (75) Many words no longer denote, or only indirectly, the original relationship between the idea and the sign — a new correlation between the phonic substance and the idea emerged and obscured, or even supplanted, the old correlations.
Unlike any other human institution, language is unbound, or unrestricted in its “choice of means”: “For there is nothing at all to prevent the association of any idea whatsoever with any sequence of sounds whatsoever.” (76) Language is a social institution, but we must go further than this and consider that the arbitrary character of linguistic signs reveals something peculiar to language.
What distinguishes language from all other institutions can be seen in the way a language evolves: “A language is situated socially and chronologically by reference to a certain community and a certain period of time… the fact that it signs are arbitrary implies theoretically a freedom to establish any connection whatsoever between sounds and ideas.” (76)
No known language — even a constructed language — is immune from evolution, from shifting, from inevitable change owing to the fact that language leads “a semiological life of its own.” (76) The transmission of language follows laws having nothing to do with those of deliberate manufacture; rather, the principle of general semiology is “the continuity of signs through time,” a principle Saussure believes confirmed by systems of writing, deaf languages, and so forth.
Closely related to this idea is his second general principle, the linear character of the signal. Auditory linguistic signs have a temporal aspect and two important temporal characteristics. First, they occupy temporal space, and second, this space is only measured in one dimension, that is, it is measured linearly. Saussure claims that the importance of this principle equals that of the first, as
“The whole mechanism of linguistic structure depends upon it. Unlike visual signals (e.g., ships’ flags) which can exploit more than one dimension simultaneously, auditory signals have available to them only the linearity of time. The elements of such signals are presented one after another: they form a chain. This feature appears immediately when they are represented in writing, and a spatial line of graphic signs is substituted for a succession of sounds in time.” (70)
But this is not always so clear to see: what about inflection, wherein a number of significant features appear to be presented simultaneously? Saussure contends this is simply an illusion, that the accentuation and the syllable really only constitute a single “act of phonation”: “There is no duality in this act, although there are various contrasts with what precedes and follows.” (70)
In a broader sense, then, linguistic change should not only be understood in terms of the evolution of the relation between a signal and its signification, but also in terms of the “evolution of the whole system.” (179) But such a comprehensive analysis of diachronic development runs quickly into the problem of the diachronic “unit” — that which in each event is the element directly “subject to change.” These units cannot be the same for both synchronic and diachronic analyses, and “[i]n any case, [the unit] will not be fully elucidated as long as it has not been studied from both static and evolutionary points of view.” (180) However, only the solution given through the diachronic unit can allow us the penetrate beneath the “superficial appearance of linguistic evolution,” permitting us to “grasp its essence.” (180)