Graphs in linguistics: Diagrammatic features and data models

Abstract

This paper examines the use of diagrams in linguistics, in order to analyze their diagrammatic features, as well as their data structures, in relation to the modeling process. It starts from defining the comparison parameters, from several seminal works on diagrams. The analysis begins from the classical IPA Consonants chart and its relative vowel graph; both are instances of quite orthodox use of visual representations, showing themselves as a mere visual account of relational data. In the syntax section, there is a strong presence of trees with their hierarchical structure. Semantics offers a far more varied collection of diagrams. Based on many possible foci and on different tasks, there are several important diagrams in use. It will be considered Wordnet, Framenet, Generative Lexicon, DRT and UML-based semantics. All these examples will lead us to stress how data models play an essential role in determining the modelling potential of graphs and the possible tasks performed by means of them. The object-oriented paradigm and its graphical development will be asserted as the most interesting use of diagrams; it will display an outstanding potential in modelling tasks.