Grounding psychosis research: why observable signs should anchor biological investigations.

Summary

Biological psychiatry struggles to find valid targets for mechanistic research because both diagnostic categories and individual symptoms are abstract symbols defined circularly within a closed interpretive system, creating a symbol grounding problem that blocks discovery of biomarkers. The author argues that progress requires separating ungrounded symptoms like delusions and hallucinations, which are co-constructed through personal and clinical interpretation, from grounded signs—directly observable features anchored in shared sensorimotor reality. A Minimal Grounding Set (MGS) can be recovered from common psychosis criteria, exemplified by disorganization and impoverishment. This MGS offers a privileged pathway for neuroscientific inquiry, with predictions that biological correlates will be most replicable for MGS, that MGS will serve as modular anchors in symptom networks, and that precision psychiatry programs depend on separating MGS from ungrounded symptoms.

Study at a glance

Abstract

Biological psychiatry faces a significant epistemic challenge in identifying valid objects for mechanistic research. Both diagnostic constructs and individual symptoms are abstract symbols defined circularly within a closed hermeneutic system, creating a symbol grounding problem that hinders the discovery of biophysical substrates (biomarkers). I argue that progress requires an epistemological separation between the ungrounded symptoms such as delusions and hallucinations, which are co-constructed through personal and clinical interpretation from grounded signs that are directly observable features anchored in shared sensorimotor reality. I propose that a Minimal Grounding Set (MGS) can be recovered from the commonly used criteria for psychosis. This MGS, exemplified by disorganization and impoverishment, offers a privileged pathway for the neuroscientific inquiry of psychosis. In this case, (1) biological correlates will be most replicable for MGS than other symptoms; (2) MGS will serve as modular anchors in symptom networks; and (3) progress in precision medicine programs like Computational Psychiatry and quantitative psychopathology frameworks such as hierarchical taxonomy (HiTOP) will depend on explicitly separating MGS from ungrounded symptoms. This approach of sign-first psychiatry will provide a non-circular foundation to tether abstract constructs affiliated with psychosis to biological realities that have eluded us for long.