Ein Beitrag zur Digital History 2023: Digitale Methoden in der geschichtswissenschaftlichen Praxis: Fachliche Transformationen und ihre epistemologischen Konsequenzen, Berlin, 23.-26.5.2023. Abstract: This paper draws on sentiment analysis methods to examine the relationship between technological change and political discourse, using telecommunications as an example. Web scraping enables the construction of a digital corpus of all Reichstag speeches between 1867 and 1942 that mention telegraphs or telephones. On this basis, three popular German-language sentiment dictionaries - SentiWS, German Polarity Clues, and German Sentiment Dictionary - are compared regarding their effectiveness in evaluating historical and complex texts. In addition, a more refined metric for extracting sentiment information is proposed, based on machine learning techniques such as part-of-speech tagging and dependency parsing. Through intensive close reading, the resulting sentiment scores are validated, with mixed results. Sentiment analysis cannot provide precise quantification of historical feelings about telecommunication, but peaks and relative changes in the underlying scores over time help to identify interesting turning points in the debate as well as technology-sceptical outsider positions.