In the last few years there has been a growing interest in using computers not only for quantitative but also for qualitative content analyses of various kinds of texts and unstructured interviews (Fielding and Lee 1993, Kelle 1998, Kuckartz 2001, Miles and Huberman 2005, Lewins and Silver 2007). This trend has given rise to the development of new software products such as MAXqda, NVivo, NUD. IST, and ATLAS. ti, which can be used for automatic coding, text retrieval, hyp- linking of related text segments, etc. Some of these programs such as ATLAS. ti or MAXqda even allow to represent the results of qualitative content analyses in graphical form as semantic networks of coded texts (Sowa 1984: 76 ff. , Lewins and Silver 2007: 179 ff. ). Such networks consist of 1. text segments or so-called quotations, which generally constitute a n- overlapping partition of the analyzed text corpus, 2. codes, which are classificatory attributes of the mentioned text segments, 3. links, which are the result of the content analytic coding and describe the attribute relations between the mentioned codes and quotations. Minestrone Soup Non-Eggs Ticinese Leek soup White wine Vegetables Romandie Figure 1: An example of a semantic network of a coded text: soup recipes from Latin Switzer- 1 land Fig.
