Basics of the visualisation of linguistic data
In science, visualisations are not only a means of presenting the results of analyses or complex interrelationships, but also an instrument of analysis: large amounts of data can often only be analysed if structures in them are visualised and then interpreted. The "Visual Linguistics" project investigates and systematises forms of visualising linguistic data as required for the analysis of "big data". The aim is to develop the theoretical foundations for a conscious and effective use of visual analysis methods in linguistics.
Visualisations of scientific analyses are omnipresent. They appear to us in the form of bar and pie charts that visualise conditions and developments, as maps that show geographical distributions or as complex diagrams that make connections and processes clear. However, visualisations are not only a means of presenting the results of an analysis. Whenever particularly large and complex volumes of data need to be analysed ("big data"), visual analysis instruments are also used: they visualise algorithmically certain structures and properties in the data, thus making interesting relationships visible in the first place.
In contrast to other disciplines, visual analysis methods of this kind are only used hesitantly for linguistic data. In addition, there is a lack of in-depth theoretical reflection: Does a visualisation really simply show what is present in the data? When is a visualisation "good", i.e. does it allow new insights to be gained from the data? What types and forms of visualisation are there and in what contexts do they relate to different research practices?
In the project "Visual Linguistics", the basics of visualisation in linguistics and language-oriented digital humanities will be developed, tested with three case studies and the findings will be continuously documented on this website.