This work identifies agency failure in language-using systems, suggesting a unique vulnerability to self-destructive behaviors.
Living systems persist by filtering inputs relative to a reference state that defines self-continuation. In biological systems, this reference is chemical and energetically constrained. This work identifies a specific failure mode that emerges uniquely in language-using systems: the reference used in filtering can become symbolically encoded and informationally editable through language. When this occurs, filtering decisions may correctly optimize symbolic identities or narratives while imposing energetic and survival costs on the underlying chemical system that executes the filtering. Agency does not fail at the level of chemistry or filtering mechanics; it fails at the reference layer. This framework explains ideological self-harm, martyrdom, burnout, anorexia, and related phenomena without invoking preference, morality, or pathology, and predicts that this vulnerability is restricted to systems capable of symbolic/linguistic self-representation. Non-symbolic living systems—including animals with abstract self-concepts—remain protected because their reference cannot be informationally rewritten. This reveals a fundamental vulnerability in linguistic intelligence: the capacity for symbolic culture is also the capacity for symbolic self-destruction. This paper is downstream to The Filtering Framework and to filtering in symbolic space DavidShadi.org
No takes yet. Share an insight, caveat, or question.
David Shadi (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: