The Interactive Method for Language Science and Some Salient Results

The use of information technology in linguistic research gave rise in the 1950s to what is known as Natural Language Processing, but that framework was created without paying due attention to the need for logical reconstruction of linguistic concepts which were borrowed directly from barely (or even not at all) formalised structural linguistics. The Computer-aided Acquisition of Semantic Knowledge project (CASK) based on the Knowledge Discovery in Databases technology (KDD) enabled us to interact with computers while gathering and improving our knowledge about languages. Thus, with the help of data mining tools, as a result of revisiting two sorts of generally admitted linguistic theories (the Predicate Argument Structure theory and Information Structure theory), we succeeded in improving these local linguistic approaches by proposing to unify the Associative Semantics (AS) theory (in which we introduced the concept of ortho-information) with the Meta-Informative Centering (MIC) theory (in which we described the meta-informative layer of natural languages). The resulting Distributed Grammar (DG) program (sketched out in this paper) treats, in addition to the above types of information, the third one, para-information (concerning identity and alikeness with respect to context and language ontology) which – despite many studies – had no uniform theoretical background in general linguistics. This DG program aims to lay the foundations for creating the theoretical background of Conceptual Linguistics.