Scientific Judgments Concerning Properties That Cannot Be Directly Observed

Abstract

Scientific judgments about properties that cannot be directly observed require distinguishing identifiability from observation laws, estimation from finite data, and theoretical premises linking indicators to target properties. This paper connects established ideas of observational equivalence and partial identification with a claim-level audit of arguments in published scientific papers. It specifies rules for selecting and segmenting claims, separates tests of observational predictions from attributions of properties, and compares the scope of statements in titles, abstracts, bodies, and conclusions. Bohm (1952) and Cogitate Consortium (2025) serve as limited reference contrasts, while a conceptual argument by Porębski and Figura (2025) is examined in greater detail. In the latter case, the audit distinguishes challenges to the evidential force of self-reports from categorical assertions of the absence of consciousness, and examines the inferential bridges and scope restrictions needed to move from multiple plausibility-based arguments to the published conclusion. The findings concern the structure of the selected arguments and their outstanding justificatory conditions; they neither determine the presence or absence of consciousness nor establish whether the original authors’ conclusions are true or false.

Keywords