Study: Personality and Real Events Shape What We Dream
Get the Health newsletter
Daily health & science — research, biotech, public health, the studies worth knowing. Free.
- IMT School for Advanced Studies Lucca researchers analyzed 3,700+ dream and waking-life reports from 287 participants aged 18–70 over two weeks, finding dream content reflects personality, cognitive style, and external events rather than being random.
- Natural language processing tools decoded the semantic structure of written dream reports at accuracy levels matching independent human evaluators, validating a scalable method for dream research that has historically resisted large-scale study.
- Dreams do not replay daily experiences directly—familiar workplaces, hospitals, and classrooms were transformed by merging with unrelated places, shifting viewpoints, and combining different parts of a person's life into single vivid scenes.
- Mind-wandering tendencies correlated with more fragmented, rapidly shifting dreams, while participants who valued dreams and believed in their personal meaning reported richer, more vivid, and more immersive experiences.
- COVID-19 lockdown dream reports contained stronger emotions and more references to restrictions, barriers, and confinement—themes that gradually faded in later months, suggesting dreams track psychological adaptation to major disruptions over time.
- Lead author Valentina Elce said the findings show dreams are "a dynamic process shaped by who we are and what we live through," not just reflections of past experiences.
Why it matters: By validating NLP for dream analysis at human-evaluator accuracy on 3,700 reports, the IMT Lucca team gives researchers a scalable method for studying memory, emotion, and mental health—fields where dream content has long been hard to investigate beyond small samples.
