Identifier et traiter les traumatismes psychiques : quelle place pour l’intelligence artificielle ? Actualisation des travaux en langue française et appel à une étude nationale

Abstract

Anchored in the recent clinical and therapeutic advancements stemming from the rise of psychotraumatology over the past decade, this article explores how cutting-edge developments in psycholinguistics enhance the identification of post-traumatic symptoms. The linguomarkers that define what we propose as a syndrome psycholinguistic traumatic (SPLIT) appear to be more specific and fine-grained than the current description of post-traumatic stress disorder (PTSD) in international nosographic classifications – particularly in the characterization of dissociative phenomena. In parallel with the rapid progress in artificial intelligence over the past two years, we present findings from a French software tool designed to support diagnostic strategies and to assess the effectiveness of evidence-based psychotherapies. These psychotherapeutic processes are conceived as acts of language emerging within the therapeutic dialogue – acts that activate and reorganize our networks of mental representations. By seeking to elucidate the structure of trauma and, consequently, the paths toward recovery, psychotherapy promotes the refunctionalization and re-harmonization of these representational networks, through the homeostatic reduction of excessive internal tensions. Drawing on language sciences, our approach advocates for the coexistence of objectivity – the foundation of quantifiable, replicable scientific inquiry – with subjectivity, as an essential marker of each individual's singular human experience. Might we soon be able to identify the specific words, phrases, and discourses that facilitate healing? We call for a nationwide psycholinguistic study to better detect and address post-traumatic psychological suffering among our fellow citizens.