Topics — What a work is about. Nucleation holds whatever reading the owner files here, whether or not an Anthology of the SOTA topic could also hold it; instruction for ML practice belongs to the anthology (ADR-048). The words are seeded from the clusters the reading-time triage of 2026-09-25 found, and grow by decision.
Epistemology — knowledge, belief and evidence — justification and rational credence, the value of knowledge, testimony, trust and expertise, and social and inductive-risk questions about evidence (group: philosophy).
10 of 100 TERM documents. Back to the full index.
| # | Title | Summary | Status |
|---|---|---|---|
| TERM-052 | crystallized argument | The owner’s term: the argument in its terminal state at a given point in a discussion, as matured as it has yet become. “Crystallization” is that maturing. Asking for the crystallized argument asks for a readout of the current terminal state, not for a new act of formation. | Active |
| TERM-056 | question drift: a provisional answer becoming part of the question an inquiry is pursuing, as distinct from query drift, in which the queries move away from a question that stays fixed | A21 §3’s distinction, used again at A23 VI.C. Classical retrieval names query drift (LIT-1079). An agent that rewrites its own task can drift in the question. The nucleating-argument line studies drift in concepts (CLAIM-177); this is drift in what is being asked. | Active |
| TERM-064 v4 | legibility, of a distinction to an observer for a task: whether the observer, with the rules, categories and resources it actually has, can recover the distinction from what it is given and bring it to bear on the task; a relation among a representation, an observer’s rule class, its resources and a task, bounded above by what the source makes possible for an ideal receiver, and directed, since what is legible to a governing institution and what of its governing is legible to others are different relations | The owner’s, U22 of the exchange of 2026-10-20: “a theme we seem to be touching on that I’m not sure is laid out clearly in our current account is the role of legibility”. A22 defined it and proposed three TERMs; one is filed, with the regime and the direction in its body. Its formal core is CLAIM-199’s usable information. CLAIM-081 holds the record’s earlier use. Adopted by the owner on 2026-10-21. It recurs through A23 to A27. | Active |
| TERM-069 v2 | observational equivalence, under a probe family: two models, states or distributions agreeing on the outcome law of every probe in a declared admissible family; coarser than identity unless the family is rich enough, and, when probes are run one at a time, weaker on their marginals than on the joint law across them | ”Probe family” is in some two dozen argument entries (CLAIM-139, 142, 149, 190, 193, 206, 209, 215, 217, 223, 241, 264, 266, 269, 302 and 322; QUESTION-005, 031, 055, 067 and 135; CASE-041, 070 and 071) with no TERM. CLAIM-264 (Active, demonstrated) holds the mathematics. Asked for at A13 and A14 of the exchange of 2026-10-20. TERM-033, fidelity as observational equivalence over response trajectories, is its case with response trajectories as the probes, and TERM-014’s communicative object is a class of it. Adopted by the owner on 2026-10-21. | Active |
| TERM-079 | belief, of a process: a maintained, truth-directed inferential commitment, held to the degree that the process uses a proposition as a premise in interpreting observations, generating expectations and evaluating counterfactuals across the contexts it bears on, together with the profile by which that commitment is revised; distinct from entertaining an interpretation, from asserting it, and from granting it authority over action | Proposed in the exchange of 2026-10-24 (workbench entry of 2026-10-24, U25), after the owner asked: “we have an operational characterization of feeling, wanting, liking… how about belief?” The record had accounts of what governs action and what is felt, and nothing comparable for belief. The candidate puts belief on the same footing, tested by its role in anticipation, inference and revision, not by report or by which motive wins. Not adopted. | Proposed |
| TERM-089 v2 | apparatus, in the operational sense: a realized organization that selectively couples differences in one system to distinguishable outcomes in another; by function it may differentiate, retain, enforce and revise, and the four come apart | The sense of “apparatus” the exchange of 2026-10-27 converged on from U19 to U23, stated so that no observer and no constitutive metaphysics is presupposed. An apparatus makes some differences in a system produce distinct outcomes in another (differentiation), and may keep them (retention), let them change what happens next (enforcement), and change its own criteria of discrimination (revision). Proposed: the record uses “apparatus” in CLAIM-476 and TERM-087 in Barad’s sense, and needed a weaker one that a thermometer, a bacterium, a sensor network and luria all fall under without argument. | Active |
| TERM-092 v3 | inquiry frontier, of an agent at a budget: the questions it can formulate, making them determinate enough to pursue, with the concepts, records, tools, time and computation it actually has; distinct from the questions its language can express, from those it can pursue to a stated standard, and from those that have answers | The object of CLAIM-475, from A17 §§2 and 7 of the exchange of 2026-10-27. A question is on an agent’s frontier if some procedure available to it, within its budget and horizon, makes the question determinate enough to pursue. Formulating counts against the budget as well as answering. The frontier moves when resources grow and when reusable specifications make formulation cheaper, and it can change shape as well as size. The term separates askable from expressible and from answerable, which CLAIM-474’s last failure turns on. | Active |
| TERM-095 v2 | inquiry: a process that pursues a question, written (Σ, A, 𝓘, 𝓞, 𝓚, 𝓑): a language in which it is posed, an answer space that says which differences would count as an answer, the interventions and observations that make those differences accessible, criteria for judging candidate answers, and the resources it runs on; it is reflexive when any of these becomes what it is pursuing | The owner, 2026-10-27: inquiry is “an important building block of our formal argument”, and the record had no TERM for it. CLAIM-470 is stated in it, TERM-092’s inquiry frontier and TERM-088’s descriptive individuation presuppose it, and TERM-056’s question drift is a change inside one. The definition is A15 and A16’s tuple from the exchange of 2026-10-27, which neither called standard: answer space, admissible interventions, observations and criteria, with the language and the resource bounds A16 added. It is a construct, not a thesis. That inquiry is a form of specification is CLAIM-470’s claim, stated in this term and not built into it. Proposed. | Active |
| TERM-097 v2 | coherence, of a family of descriptions, transports or judgements: their agreement with one another under a stated correspondence structure (on overlaps, around cycles, between routes), distinct from fidelity to a source, from integration by dynamics, and from knowledge of the world | The sense in which the record keeps saying coherence is not something else: “coherence without integration” and “without knowledge” (CASE-179), a coherent record as the product of a legibility regime (CLAIM-448), and agreement between translation routes as coherence, not fidelity (CLAIM-496). It is a relation among the members of a family, read through the maps that say what counts as agreeing (TERM-093). Filed on the owner’s question of 2026-10-29 whether the record had TERMs for fidelity and coherence: fidelity has several, coherence none, though the word is in some 145 entries. Adopted the same day. | Active |
| TERM-098 v2 | concept, as realized by an organization: a reusable, systematically consequential pattern of distinctions, applicable across a stated range of contexts with inferential or operational consequences; distinct from the feature or representation that carries it, from the organization that realizes it, and from what it is about, and distinct from conceptual corrigibility, the further capacity to find and revise its inadequacies | Filed at the owner’s decision of 2026-10-31, after the exchange “Analyze nucleation concepts” defined “concept” three times in three turns (A4, A5, A6). At the owner’s review, version 2 separates conceptual organization, which the term names, from conceptual corrigibility, a further achievement it does not require. A candidate, not adopted: QUESTION-360 asks what would settle it. | Proposed |