«The entities of linguistic form are of “algebraic” nature and have no natural designation; they can therefore be designated arbitrarily in many different ways.»

(Louis Hjelmslev)

Since Claude Shannon‘s Mathematical Theory of Communication (1936), the notion of information in its technically treatable sense is often distinguished from its linguistic sense by ascribing to the former, as opposed to the latter, a purely quantitative treatment. Yet since the founding documents of a general linguistics in the late 19th century, it is clear for every linguist who affirms the break with the traditional way of studying language as philology, that the notion of the sign is to be treated purely quantitatively as well. Ferdinand de Saussure‘s structuralist paradigm for understanding processes of signification views the linguistic sign as a quantitative value, yet as a negative one which cannot, in itself, be positivized. As a negative value, it can only be specified by „profiling“ it through infinitary lists and their net of contrasts: a ,this‘ can never be signified directly, Saussure held, but only by labelling it as ,not-that‘ and not-that‘ and ,not-that‘ etc. In short, a linguistic sign can only be determined structurally and differentially, within a framework of place-value distributions.

From a logical point of view, de Saussure‘s paradigm of negative determination obviously entails problems regarding methodological feasibility, since it holds, by principle, that the necessary infinitary lists can never exhaustively be made explicit. This is the decisive reason why de Saussure himself considered his own structural approach, which attempted to conceive of language as a system, as having ultimately failed. Surely the post-structuralist critiques on such a notion of general linguistics are well known; yet from the point of view of algebraic computability (rather than that of logics), the situation looks different and is hardly explored today. Louis Hjelmslev is one of the very few linguists who continued the „differentiability within negativity“ approach initiated by de Saussure, by extending it mathematically. He considered Saussure‘s ,negative values‘ in a generalized sense as ,algebraic invariants‘. Like this, the structuralist paradigm is open for taking probabilistic procedures like Markov Chains and other algorithms, with which the diverse programming languages ordinarily work today, into account. From the logical point of view, this can hardly count as a forward pointing path, since it does not clarify how a notion of system could be objectified. Yet with regard to the logistic networks, such fixation is (arguably) neither necessary nor desireable. Here, Hjelmslev‘s algebraic approach offers a powerful alternative to the pre-dominant approaches in terms of semantic or object-oriented (informational) database logic and ontologies, because it is capable of abstracting from the distinction between natural language vs artifical/formal language and needs not subject one to the other: communication and signification can be treated as mutually complementary aspects.

In this kolloquium we will work through Hjelmselv‘s Prolegomena to a Theory of Language (1943), and appropriate it methods in practice. We want to explore if and how structural linguistics as glossematics (in the sense of Hjelmslev) can be extended towards an alphabet of things that were capable of integrating the operability of generative linguistics (Chomsky etc), and hence could provide a powerful method of pre-specific modeling.