The structure of a person's story predicts their mental health better than the specific words they use.
Narrative architecture is a more reliable indicator of psychological states than vocabulary or word choice. Most automated mental health screenings focus on word counting, assuming that depressed or anxious people use specific negative terms. This analysis shows that how people organize their thoughts across a whole story tells a much deeper truth. People with specific conditions struggle with the macro-structural flow of a narrative rather than just the individual words. Focusing on the scaffolding of speech could help clinicians identify risks that traditional word-based AI misses entirely.
Multi-Level Narrative Evaluation Outperforms Lexical Features for Mental Health
arXiv · 2604.27846
How people narrate their experiences offers a window into how the mind organizes them. Computational approaches to therapeutic writing have evolved from lexical counting to neural methods, yet remain fragmented: dictionary tools miss discourse structure, while embeddings conflate local coherence with global organization. No existing framework maps these techniques onto the hierarchical processes through which narratives are constructed. Here we introduce a three-level framework - micro-level lex