Linguistic spotting of arguments
Artificial intelligence and political discourse. The case of Emmanuel Macron (2017-2021)
Computers - in this case, artificial intelligence - can be of great service to social sciences and humanities - in this case, the analysis of political discourse - provided they abandon a neo-positivist, probative approach in favor of a hermeneutic, exploratory stance.
The aim of digital technology, in fact, is not to objectify the meaning of texts - since meaning is not a positive given, but a construct - but to objectify reading paths capable of nourishing interpretation. Applied to Emmanuel Macron's corpus between 2017 and 2021, compared with that of his predecessors (de Gaulle, Pompidou, Hollande), deep learning can help us recognize the salient linguistic elements on which the President bases his arguments and (un)veils his ideology, and which we still have to interpret.
