Re-examining Criteria for Consciousness: An Ontological Framework Based on "Selectivity"

A core puzzle across science and philosophy is determining whether a system possesses consciousness. Existing criteria—behavioral imitation, neural correlates, and integrated information—rely on specific theoretical assumptions, limiting their cross-framework persuasiveness. The paper argues the root problem is the search for an 'instantaneous static marker,' whereas consciousness may be a process unfolding over time. It proposes a de-subjectivized ontological criterion: the essence of consciousness is a system's possession of 'selectivity,' meaning its evolutionary behaviors cannot be fully explained by its complete history and internal parameters. Selectivity is not quantifiable by any instantaneous physical quantity; it is an emergent holistic property of long-duration sequential processes. This redirects research from hunting for instantaneous evidence toward monitoring long-term behavioral predictability, supported by an isomorphic mathematical model and structural resonance with spacetime physics.

Study at a glance

Abstract

Determining whether a system possesses consciousness stands as a core cross-disciplinary puzzle spanning science and philosophy. Mainstream existing criteria—including behavioral imitation, neural correlates, and integrated information index—all rely on specific theoretical presuppositions, undermining their persuasiveness across different frameworks. This paper argues that the root of this dilemma lies in the universal pursuit of an "instantaneous static marker" within old paradigms, whereas consciousness may not be a static snapshot at all, but a process unfolding across time. To this end, we propose a de-subjectivized ontological criterion: the essence of consciousness is a system’s possession of "selectivity"—meaning the evolutionary behaviors of the system cannot be exhaustively explained by its complete historical trajectory and internal parameters. We further demonstrate that selectivity is not quantifiable via any instantaneous physical quantity; it is fundamentally an emergent holistic property of long-duration sequential processes. This reasoning resolves inherent flaws of prior paradigms and redirects consciousness research from hunting for instantaneous evidence toward monitoring long-term behavioral predictability. Finally, we show that this theoretical direction gains concrete methodological support from an isomorphic mathematical model built around the integral constant C, and forms structural resonance with parallel explorations in spacetime physics.