The Minimal Self: First-Person Individuation and the Limits of Functional Explanation - The AI Duplication Paradox and the Reductive Aspirations of Physicalism
Zenodo (CERN European Organization for Nuclear Research) June 17, 2026 Peer reviewed DOI: 10.5281/zenodo.20736658 via OpenAlex
Summary
The paper argues that first-person subjectivity cannot be fully explained using physical-functional terms, introducing the AI duplication paradox to illustrate the problem of subjective individuation. It suggests that even if consciousness can be explained, the question of what defines a unique perspective for a subject remains unresolved. The author critiques existing theories and proposes that the phenomenological concept of the minimal self offers a better understanding of individual consciousness.
Study at a glance
| Key finding | First-person subjectivity cannot be adequately explained in purely physical-functional terms, as it fails to address what makes a perspective unique to an individual subject. |
|---|
Abstract
Abstract AbstractThis paper asks whether first-person subjectivity can be adequately explained in purely physical-functional terms. I approach the question through what I call the AI duplication paradox—the ontological difficulty that arises when one and the same computational architecture is instantiated in parallel across multiple substrates—and argue that functionalism cannot account for the numerical individuation of a conscious subject. This difficulty, which I term the problem of subjective individuation, is distinct from the hard problem of consciousness: one could explain why consciousness exists and still face the question of what makes a given perspective belong to one subject and not to another. The paper supports this through the AI scenario and a parallel case of perfect biological cloning. Neither a single shared consciousness nor an indefinite multiplicity of functionally identical subjects resolves the issue, and spatial coordinates do not settle it, since they describe third-personal relations rather than subject-individuating features. Set within the current debate on individuating artificial minds, the argument shows that recent approaches, whether reductive or anti-reductive, conceptual or mechanistic, converge on the same limit: they describe how such systems are organised without fixing the subject for whom any experience is given. To motivate an alternative, I draw on the phenomenological notion of the minimal self and its constitutive mineness (Jemeinigkeit), which exhibits a non-aggregative unity and structural continuity that resist physicalfunctional decomposition. The argument is then brought into dialogue with contemporary illusionism, in particular Dennett’s Multiple Drafts model and Frankish’s quasi-phenomenal account. My claim is not that these views are confused, but that they leave mineness, binding, and numerical individuation unresolved at the level at which the hard problem is usually posed.