From digital humanities to a humane digitization

For decades, there was little interaction between the fields of computer science and the humanities. This has changed dramatically in the last years because of the wide application of computer-based tools in the humanities. The application of these tools goes well beyond mere data management in large data bases; it rather uses newest technologies from artificial intelligence to identify implicit relationships in meaningfully represented data from the humanities. One such representation is a complex network that makes the interactions between data entities analyzable. However, the analysis of such complex networks is often based on two kinds of models, which are rarely discussed explicitly, which makes it difficult to understand the limits of the applicability of these models.  Because of this lack of transparency we often see misinterpretations of the results when readily available analysis methods from some software package are carelessly applied to  complex networks. Successful analysis thus requires literacy in the models and algorithms applied to the data.

The first impression might now be that computer science is much more useful to the humanities than vice versa. But the above analysis also indicates that the transparent and humane use of algorithms is not trivial: Today, algorithms are used to identify „criminal personalities“, to evaluate peoples' creditworthiness, or to determine an insurance rate. How can we ensure that algorithms do not have a bias, that they do not discriminate between people that we do not want to be treated differently or that they are not used to manipulate results? What does a sensible balance between privacy and security, business models and transparency look like? To solve these and other pressing questions, I claim that computer science requires the insights from the humanities more than ever to enable a humane digitization. We need analogies from historical cases as provided by historians, the best insights into the evolution of society by sociologists and economists, fundamental knowledge of the human mind based on psychological research, and last but not least a well-balanced regulation of algorithm-based activities by legal scholars. It could actually become the century of the humanities.