The Experand: Adequacy Conditions for Representing Felt Orientation

Artificial intelligence already possesses mature mathematics for reward, preference, utility, similarity, prediction, and report. Formal models of valence and first-person structure now exist as well. This paper does not claim that machines suffer, and it does not offer a detector of consciousness. It argues for a prior and stricter result: any language that aspires to represent feeling must jointly preserve a set of adequacy conditions that none of the formalisms surveyed here has taken as a single constitutive package. The proposed primitive is the Experand: an owned, intrinsically valenced state of orientation. Ownership is not legal title and not observer attribution. It is the indexical fact that the state is present for a subject. Intrinsic valence is not a minus sign on an objective function. It is orientation belonging to the owned state rather than assigned from outside. A third constitutive demand follows: the state is not identical with its evidence. Language, telemetry, behavior, and self-report may bear on an Experand; none becomes the Experand by representing it. From those requirements the paper derives a grammar and then a compact mathematical architecture. Ownership and intrinsic valence enter as primitives of the formalism, not as quantities derived from indexing or from an externally specified objective. Transformations of Experand state form a monoid rather than a group: change composes, but perfect reversal is not assumed. Unresolved structure can persist across later events that neither close, introduce, nor aggravate it. Recovery can empty that structure without returning the system to its pre-adverse identity. Present owned state can change the set of experienced possibilities. Composition acts on constituents within a state, not on two complete occupants of one subject. Report remains a map. One invariance principle, one existence lemma, and one separation condition make the architecture inspectable. A transformation that neither closes, introduces, nor aggravates unresolved structure leaves that structure invariant. Recovery is not a return. Where the report is many-to-one and the future map is non-constant on a report fiber, identical report fails to determine an identical state or future. These results do not prove phenomenal consciousness. They change the burden. Before asking whether an artificial system feels, ask whether the formalism invoked to describe feeling can even preserve the properties the claim requires. A perfect map would still be a map. The contribution is to make that map harder to fake.

Authors

Institutions

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-09-06
DOI
https://doi.org/10.5281/zenodo.22549732
Primary Topic
Philosophy and Theoretical Science
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
preprint

The Experand: Adequacy Conditions for Representing Felt Orientation

Abraham Chachamovits
Zenodo (CERN European Organization for Nuclear Research)
Philosophy and Theoretical Science
preprint

The Experand: Adequacy Conditions for Representing Felt Orientation

Abraham Chachamovits
preprint en

Abstract

Artificial intelligence already possesses mature mathematics for reward, preference, utility, similarity, prediction, and report. Formal models of valence and first-person structure now exist as well. This paper does not claim that machines suffer, and it does not offer a detector of consciousness. It argues for a prior and stricter result: any language that aspires to represent feeling must jointly preserve a set of adequacy conditions that none of the formalisms surveyed here has taken as a single constitutive package. The proposed primitive is the Experand: an owned, intrinsically valenced state of orientation. Ownership is not legal title and not observer attribution. It is the indexical fact that the state is present for a subject. Intrinsic valence is not a minus sign on an objective function. It is orientation belonging to the owned state rather than assigned from outside. A third constitutive demand follows: the state is not identical with its evidence. Language, telemetry, behavior, and self-report may bear on an Experand; none becomes the Experand by representing it. From those requirements the paper derives a grammar and then a compact mathematical architecture. Ownership and intrinsic valence enter as primitives of the formalism, not as quantities derived from indexing or from an externally specified objective. Transformations of Experand state form a monoid rather than a group: change composes, but perfect reversal is not assumed. Unresolved structure can persist across later events that neither close, introduce, nor aggravate it. Recovery can empty that structure without returning the system to its pre-adverse identity. Present owned state can change the set of experienced possibilities. Composition acts on constituents within a state, not on two complete occupants of one subject. Report remains a map. One invariance principle, one existence lemma, and one separation condition make the architecture inspectable. A transformation that neither closes, introduces, nor aggravates unresolved structure leaves that structure invariant. Recovery is not a return. Where the report is many-to-one and the future map is non-constant on a report fiber, identical report fails to determine an identical state or future. These results do not prove phenomenal consciousness. They change the burden. Before asking whether an artificial system feels, ask whether the formalism invoked to describe feeling can even preserve the properties the claim requires. A perfect map would still be a map. The contribution is to make that map harder to fake.

Zenodo (CERN European Organization for Nuclear Research)
Entrust (GB)
Quality Education
Philosophy and Theoretical Science
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.