Introductory remarks

In the Introduction, I posed a basic question concerning network models: what phenomena (at what level of reality) are they models of? Do they represent psychological reality (putative semantic structure stored in the mind, which underlies an actualization in the context), historical reality (a diachronic social process of meaning extension), or logical reality (relatedness of ideas)? Or some combination of these? The first two options presuppose a developmental account of the relatedness in question. This assumption was famously rejected by Saussure in his attempt to establish linguistics as a discipline on solid foundations. I have already hinted at his view, when “the unity of the sign” was mentioned (1.0.). It is time to see in greater detail what his position involves and to what extent it shapes contemporary approaches to language.

Saussure: synchrony vs. diachrony

What are the grounds for a linguist to focus on language as a state? First, since a succession of historical developments “does not exist insofar as the speaker is concerned”, a linguist “can enter the mind of speakers only by completely suppressing the past” (Saussure 1959: 81). Saussure clearly assumes that the “reality” of language is in the present state of the mind of its speakers and that it must be captured without reference to past states.