Psychosis-associated disruptions to semantic foraging in narrative recollections: Integrating language model embeddings and topic modeling.

Read full text Summarise with VIRAT
Authors: Luke Leckie, Nancy B. Lundin, Amanda R. Bolbecker, Joshua W. Brown, William B. Hetrick, Kyle S. Minor, Brian F. O’Donnell, Peter Martin Todd, Jacob G. Foster
Cited by: 0
References: 55

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Thought disorder as observed through language disturbance is a hallmark symptom of psychosis. Disruptions of verbal fluency task behavior (e.g., listing animals) among individuals experiencing psychosis have been explained as alterations to semantic foraging processes. Semantic foraging views semantic information retrieval as consisting of bouts of exploitation within clusters of related information punctuated by strategic exploratory jumps between clusters. However, the semantic foraging framework has not yet been applied to naturalistic speech. Here, we examine whether disruptions to semantic foraging operating at the level of spoken narrative can explain the disruptions to natural language that are coupled to psychotic symptom severity (e.g., tangential speech, poverty of content). To explore this question, we combined the language model Bidirectional Encoder Representations from Transformers (BERT) with discourse atom topic modeling to analyze how 93 participants with schizophrenia-spectrum disorders forage for topics as they narrate their life stories. Negative symptom severity was associated with distinct narrative content, less topical focus, and smaller cluster sizes. Positive symptom severity was associated with greater semantic distances between successive clusters. Cognitive and disorganized symptom severity was associated with distinct narrative content, larger cluster sizes, fewer topics retrieved per cluster, and shorter between-cluster semantic distances. These results were largely robust to modeling choices and transdiagnostic effects of participant demographics, general indices of psychopathology, and inferred substance use history. By using language models within a semantic foraging framework to analyze narratives, this study extends observations from verbal fluency tasks into naturalistic speech and offers mechanistic insights into psychosis-related speech disturbances. (PsycInfo Database Record (c) 2026 APA, all rights reserved).

Fields of Study & Subjects

Schizophrenia research and treatment Mental Health via Writing Neurobiology of Language and Bilingualism

Publication details

Journal / source
Journal of Psychopathology and Clinical Science
Publication date
29 October 2026
Language
EN