Why Conclusions Diverge from the Same Observations: Formalizing World-Model Non-Identifiability via an Inference Profile θ \theta

Toru Takahashi · 2026

Read Why Conclusions Diverge from the Same Observations: Formalizing World-Model Non-Identifiability via an Inference Profile θ \theta by Toru Takahashi (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-12255.txt.

Introduction

Toru Takahashi argues that clashing answers from evidence reflect non-identifiability, not broken thinking. He sorts the puzzle on two levels. At theta-level, the same model yields different answers as settings differ. At W-level, repeat use of a setting biases data contact and updates, so the model splits. He defines a profile of Reference, Exploration, Stabilization, and Horizon, showing how outputs split for the same sight. Disputes project onto bases: concrete versus abstract, outward sharing, and order versus freedom, from limits on math, sight, and teamwork.

Catalog record

CREATORToru Takahashi
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE28331
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.12255

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Why Conclusions Diverge from the Same Observations: Formalizing World-Model Non-Identifiability via an Inference Profile θ \theta Toru Takahashi Abstract When people share the same documents and observations yet reach different conclusions, the disagreement often shifts into a judgment that the other party is cognitively defective, irrational, or acting in bad faith. This paper argues that such divergence is better described as a form of non-identifiability inherent in inference and learning, rather than as a defect of the other party. We organize the phenomenon into two levels: (i) θ \theta -level non-identifiability, where conclusions diverge under the same world model W W because inference settings differ; and (ii) W W -level non-identifiability, where repeated use of an inference setting θ \theta biases data exposure and update rules, causing the learned world model W W itself to diverge. We introduce an inference profile θ = ( R , E , S , D ) \theta=(R,E,S,D) , consisting of Reference, Exploration, Stabilization, and Horizon, and show how outputs can split even for the same observation o o and the same W W . We further explain why disagreements tend to project onto a small number of bases—abstract versus concrete, externalizability, and order versus freedom—as a consequence of general constraints on learning systems: computational, observational, and coordination constraints. Finally, we relate the framework to deep representation learning, including representation hierarchy, latent-state estimation, and regularization–exploration trade-offs, and illustrate the framework through a case study on AI regulation debates. 1 Introduction People often examine the same documents, statistics, logs, or incidents and nevertheless reach different conclusions. This phenomenon appears repeatedly in policy making, organizational decision making, research and development, product safety, and public debates on emerging technologies. What is especially problematic is not only that conclusions diverge, but that such divergence is often reinterpreted as a defect of the other party: lack of ability, lack of sincerity, ideological corruption, or moral deficiency. Once disagreement is framed in this way, the discussion rarely moves toward productive activities such as additional observation design, model comparison, or refinement of inference procedures. The aim of this paper is to separate conclusion divergence from personality evaluation and to formulate it as a problem of non-identifiability in world-model estimation. A world model is understood here as a statistical and computational device that forms internal representations from observation sequences in order to support prediction and decision making. Non-identifiability refers to a situation in which, under finite data, partial observability, and representational constraints, multiple models or inference policies remain compatible with the same observations, so that the conclusion is not uniquely determined by the observations alone [ 1 , 2 ] . This paper distinguishes two levels of non-identifiability. • \theta -level non-identifiability : even under the same world model W , conclusions may diverge if inference settings differ, including which grounds are adopted, how widely alternatives are explored, when updating stops, and which temporal horizon is emphasized. • W -level non-identifiability : repeated use of an inference setting \theta biases exposure to data and update rules over time, causing the learned world model W itself to diverge. Once this happens, the same observation is processed through different causal and representational structures. This two-level structure makes it possible to treat both short-term misalignment and long-term epistemic fragmentation within a single framework. Figure 1 gives a schematic overview. Shared observations o_{\leq t} Inference profile \theta=(R,E,S,D) Conclusion y World model W_{\phi} repeated use biases updates same input divergent output Figure 1: Two levels of non-identifiability. At the \theta level, different inference profiles can produce different conclusions even under the same world model. At the W level, repeated inference operations bias data exposure and updates, causing the world model itself to diverge. The contribution of this paper is threefold. First, it introduces an inference profile \theta=(R,E,S,D) as a compact representation of inference operation. Second, it explains why this four-dimensional operational freedom tends to be projected onto three recurrent bases of disagreement: abstract/concrete, externalizability, and order/freedom. Third, it connects these bases to structural features of deep representation learning, thereby grounding the framework in computational terms rather than treating disagreement merely as a matter of values or communication style. The central claim is not that all disagreements are equally valid, nor that inference differences should be normatively celebrated. Rather, the claim is that disagreement often has an identifiable computational structure. By locating divergence in inference operation and world-model learning, we can shift the problem from moralized evaluation to designable coordination. 2 Background: SIA and MIA 2.1 The Single Intelligence Assumption When people disagree despite sharing observations, many discussions implicitly rely on what this paper calls the Single Intelligence Assumption (SIA). SIA is not a single formal doctrine but a composite of three assumptions. 1. Centralization : the essence of intelligence is assumed to lie in abstract symbol manipulation or logical reasoning, while other modes of inference are treated as secondary. 2. Normativization : the centralized form of intelligence is treated as the uniquely correct norm, so deviations from it are interpreted as insufficient effort, insufficient education, or ethical failure. 3. Commutability : if the same input is provided, properly rational agents are expected to arrive at the same conclusion. SIA can be useful in education, evaluation, and institutional operation because it provides a common standard. However, the stronger the commutability assumption becomes, the more likely divergence is to be attributed to defects of the other party rather than to differences in inference operation, observation design, or model structure. This attribution blocks inquiry into observation design, inference settings, and model mismatch. 2.2 The Multiple Inference Assumption As an alternative, this paper adopts the Multiple Inference Assumption (MIA). MIA treats diversity in inference operation and learning not as an exception to be eliminated, but as a consequence of non-identifiability in world-model estimation. The key point is not to normatively celebrate diversity, but to describe divergence as a set of adjustable design degrees of freedom rather than as bad faith or cognitive deficiency. MIA therefore functions as a methodological assumption: when the same observation leads to different conclusions, the first question should be where the inference process differs, and whether the difference belongs to an operational setting \theta or to a learned world model W . 2.3 Externalizability and Internalization This paper characterizes grounds and states in terms of the cost of sharing and verification across agents. • Externalizability 1 1 1 An alternative phrasing is objectifiability : the degree to which a state can be described as an object that exists beyond any particular subject. We use externalizability throughout this paper to emphasize the operational aspect of inter-agent sharing and verification, rather than the metaphysical question of whether the state has subject-independent existence. : the degree to which a ground or state can be described in a form that is shareable, reproducible, and auditable across agents. Measurements, logs, legal texts, and benchmark scores have high externalizability. • Internalization : the degree to which a state is inferred as an internal state of an agent, such as confidence, anxiety, expectation, tacit knowledge, or a sense of danger. Such states cannot be uniquely reconstructed from observation o and typically require high description cost to be communicated. This distinction is not an epistemological opposition

Read the complete text