Lecture at the Semantic Technology Institute (STI) Innsbruck

The transformation of unsorted verbal data into knowledge systems by GABEK®

The problem

I presuppose we have lots of unsorted verbal data in colloquial language by employees of a complex social institution, by individuals of conflicting groups, or by clients of a company ect. and we want to understand the whole situation. Although we have the feeling to understand the individual statements we ask: which of the statements are relevant? Which are accepted by many persons? Which assumptions or pragmatical strategies and goals are held in the backgrounds of the verbal data?


To understand assumptions, strategies or goals of an organisation, it is necessary to explain them on the basis of opinions, values, experiences or norms of the concerned individuals. Therefore we group the data base with software support through WinRelan® in text groups, which show how statements are interrelated with each other. Using texts of employees, clients, experts or other individuals we build small but meaningful text groups, which substantiate the assumptions or justify core strategies and goals. The text groups must be consistent, must express meaningful connections between the topics expressed and should be founded through experiences of the concerned individuals. In such a way – using GABEK (GAnzheitliche BEwältigung von Komplexität) as a method of knowledge organisation – unsorted verbal data are structured in order to get an overview on the total situation as well as special anwers of questions for details.


However, this is not yet sufficient to understand also the complex situation as a whole. Sure, understanding presupposes to be able to explain details, but additionally it is necessary that the explained details can be set in a larger context. This means that decision makers should be able to point to a larger context for each statement or each decision they make, a context showing the surrounding, in which the statement is necessary and can be used for further patterns of explanation, orientation or action.

Linguistic gestalt building

In Analogy to the theory of perception of gestalten by Carl Stumpf a theory of linguistic gestalt building was developed with GABEK@. He defined perceptional gestalten as a kind of meaningful net between sensual data. We see linguistic gestalten as nets between statements. Linguistic gestalten are consistent text groups fulfilling particular formal and semantic rules. The statements are structured logically and hierarchically and build together a meaningful whole. The resulting gestaltentree is a holistic representation of the total situation – as far as verbal data are available, such, that all statements of the gestaltentree can be explained through textgroups (with exception of the verbal raw data) and all the statements can be seen within a wider context (with exception of the highest texts of the gestaltentree). The texts are structured logically and hierarchically to form a „Gestaltentree“.


The gestaltentree represents complex situations, as far, as there are verbal data available. Explications of partial texts give answer to the question for foundation or for argumentation (why you beleive that?), the widened context gives answer to the question for relevance (wherefore this is important and on which possible applications do you think?). Therefore the gestaltentree can be used for learning in an organisation. Decision makers are in the position to anwer questions for arguments and to communicate strategies in a transparent mode. Employees and those concerned can comprehend decisions and are likely ready to co-operate in the implementation process.

The lecture

gives a short overview on the GABEK-rules of gestalt-building.
Then the way to analyze verbal data in order to form a linguistic gestalten tree is presented.
Finally some examples of application of GABEK@ in praxis for problem solving, learning, organisation development etc. are discussed.





Comments are closed.


Skip to top 

Copyright © 1994 - 2018 GABEK® | Log in | Impressum | Kontakt |