Logical Underdetermination under Maximal Intervention Access in Audit-Fixed Neutral Inference
Year: 2026
Type: article
Abstract: Preprint: This work is a preprint, has not undergone peer review, and is made publicly available to establish a public scientific record. This manuscript establishes a structural closure result within the audit-fixed intervention framework. It shows that even under maximal admissible intervention access with a fixed, non-adaptive evaluator, neutrality does not generically entail unique reconstruction on a chosen description space. The audit-fixed observational content is fully captured by the induced outcome partition. Maximal intervention access enlarges the intervention domain but does not, by itself, introduce a neutral selection principle on the description space. Consequently, in the absence of explicitly assumed additional structure, uniqueness is structurally excluded as a neutral default. The result is purely protocol-level. It introduces no dynamics, no ontological commitments, and no mechanism reconstruction. It complements the prior result on invariant outcome structure and completes the logical closure of neutral reconstruction under intervention extension. related to: Foundations of Intervention Consistency: https://doi.org/10.5281/zenodo.18495497 Definition of the GIT Test: https://doi.org/10.5281/zenodo.18499822 No Neutral Default: https://doi.org/10.5281/zenodo.18664221 Correspondence regarding this work may be directed to:[email protected]
Authors: Kaya, Cueneyt, Buck, Detlev W.
Language: English
Subfield: History and Philosophy of Science
Field: Arts and Humanities
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7129080368
FWCI:
Cites:
Cited by:
Related to: