{ "_meta": { "title": "Atlas Gap-Fill Corpus v1", "date": "2026-07-08", "rationale": "Disciplines absent from BOTH the base atlases (657 nodes) and the expansion corpus (235 nodes). Derived from the Atlas Omission Audit, Tiers 2 and 3, plus disciplines the audit itself missed. Every entry carries a status tag; none is presumed true.", "status": "OPEN \u2014 all 20 entries gate-checked 2026-07-09 with evidence recorded per entry. The checks are self-application (Ouroboros caveat); entries remain OPEN pending independent panel review \u2014 Heliaia docket item 1." }, "domains": { "sts": { "label": "Science & Technology Studies", "cluster": "soc" }, "educ": { "label": "Education & Learning Sciences", "cluster": "soc" }, "lis": { "label": "Library & Information Science", "cluster": "cs" }, "arch": { "label": "Architecture, Design & the Built Environment", "cluster": "hum" }, "trans": { "label": "Translation & Interpreting Studies", "cluster": "hum" }, "demog": { "label": "Demography & Population Studies", "cluster": "soc" }, "polad": { "label": "Public Policy & Administration", "cluster": "soc" }, "dh": { "label": "Digital Humanities & Computational Text", "cluster": "hum" } }, "open": [ { "id": "sts_social_construction_science", "dom": "sts", "domLabel": "Science & Technology Studies", "cluster": "soc", "kind": "open", "form": "framework", "status": "open", "mat": "perennial", "name": "How far does the social shaping of scientific knowledge go?", "stmt": "The Strong Programme holds that the content of scientific belief, not merely its context, admits symmetric sociological explanation for true and false beliefs alike. Critics hold that this collapses into relativism or is trivially compatible with realism. No agreed criterion separates the substantive claim from the trivial one.", "prog": "Bloor's symmetry principle; Latour and Woolgar's laboratory ethnography; the Sokal affair as a polemical inflection; recent 'engaged programme' work seeking a non-relativist reading.", "barrier": "The disputants disagree about what would settle it, which makes this a persistent conceptual problem rather than an evidential gap.", "belief": "Contested. The strong reading is widely rejected in the sciences and widely defended in STS, which is itself the phenomenon.", "review": { "gate": "J\u00fcrge", "date": "2026-07-09", "verdict": "HOLDS", "evidence": "Bloor symmetry principle confirmed as stated: same types of social cause for true and false belief alike (Knowledge and Social Imagery, 1976); the relativism / self-refutation critique line confirmed as the standing objection; the long critic list (Laudan, Hollis, Newton-Smith, Boghossian) confirms \"no agreed criterion\" is the honest status. Sub-note OPEN: the term \"engaged programme\" not separately re-verified this run." } }, { "id": "sts_expertise_boundary", "dom": "sts", "domLabel": "Science & Technology Studies", "cluster": "soc", "kind": "open", "form": "analysis", "status": "open", "mat": "active", "name": "What makes someone an expert, and who may judge?", "stmt": "A criterion distinguishing genuine expertise from credentialed authority, applicable by non-experts. Directly load-bearing for any account of higher-order evidence: if laypeople cannot identify experts, deference has no rational basis.", "prog": "Collins and Evans's periodic table of expertises; Goldman's five criteria for the novice-expert problem; the replication crisis as a natural experiment in credential failure.", "barrier": "Any proposed criterion must be applicable by someone who lacks the expertise being assessed, which is close to a circularity.", "belief": "Open, and increasingly urgent as generation outruns expert review.", "review": { "gate": "J\u00fcrge", "date": "2026-07-09", "verdict": "HOLDS", "evidence": "Collins & Evans periodic table of expertises confirmed (Rethinking Expertise, 2007; interactional vs contributory). Goldman (2001) confirmed as the right cite for the novice/expert problem; the count \"five criteria\" not independently re-confirmed this run \u2014 sub-note OPEN." } }, { "id": "educ_transfer", "dom": "educ", "domLabel": "Education & Learning Sciences", "cluster": "soc", "kind": "open", "form": "mechanism", "status": "open", "mat": "active", "name": "Does learning transfer, and under what conditions?", "stmt": "Whether skills acquired in one domain improve performance in structurally distinct domains. Far transfer is the premise of general education; the evidence for it is persistently weak.", "prog": "Thorndike's identical elements; near-transfer effects are robust, far-transfer effects rarely survive replication; the brain-training literature is a cautionary case.", "barrier": "Transfer is defined by the absence of shared structure, but any measured transfer invites the reply that shared structure was present after all.", "belief": "Near transfer: well supported. Far transfer: repeatedly claimed, rarely replicated.", "review": { "gate": "J\u00fcrge + Caliber", "date": "2026-07-09", "verdict": "HOLDS \u2014 strongest verification of the batch", "evidence": "Gobet & Sala (2023, Persp. Psych. Sci.): overall far-transfer effect is null with little true variability across training types; second-order meta-analysis (Collabra 2019) finds unbiased WM far transfer g\u22480.01; near transfer supported. Entry claims exactly this." } }, { "id": "educ_expertise_practice", "dom": "educ", "domLabel": "Education & Learning Sciences", "cluster": "soc", "kind": "open", "form": "mechanism", "status": "open", "mat": "active", "name": "How much of expertise is deliberate practice?", "stmt": "The share of variance in expert performance explained by accumulated deliberate practice, versus prior aptitude, starting age, and domain predictability.", "prog": "Ericsson's original estimates; subsequent meta-analyses report far smaller shares, varying sharply by domain; the popularised '10,000 hours' claim exceeds what the source study asserted.", "barrier": "Practice is not randomly assigned; those who improve faster practise more.", "belief": "The strong version is refuted; the weak version is uninformative. The interesting range is in between.", "review": { "gate": "J\u00fcrge", "date": "2026-07-09", "verdict": "HOLDS", "evidence": "Macnamara, Hambrick & Oswald (2014, Psych. Science): deliberate practice explains 26% of performance variance in games, 21% music, 18% sports, 4% education, <1% professions \u2014 exactly the \"far smaller shares, varying sharply by domain\" our entry asserts. A contested Ericsson-side reanalysis exists; the entry already places the truth \"in between.\"" } }, { "id": "lis_relevance", "dom": "lis", "domLabel": "Library & Information Science", "cluster": "cs", "kind": "open", "form": "analysis", "status": "open", "mat": "perennial", "name": "What is relevance?", "stmt": "Information retrieval optimises a relation nobody has satisfactorily defined. Topical, cognitive, situational, and affective relevance come apart, and evaluation metrics silently commit to one.", "prog": "Saracevic's stratified model; Cranfield-paradigm evaluation assumes topical relevance and static judgments; large language models retrieve against an implicit relevance notion that has never been stated.", "barrier": "Relevance is a three-place relation between item, question, and questioner, but the questioner's information need is the thing they cannot articulate.", "belief": "Open. The field builds systems that work without an account of what they optimise.", "review": { "gate": "J\u00fcrge", "date": "2026-07-09", "verdict": "HOLDS", "evidence": "Saracevic stratified model confirmed (1975 JASIS; 2007 JASIST Parts II\u2013III); manifestation typology (system/topical/cognitive/situational/motivational) confirmed; Cranfield-paradigm assumptions confirmed as topical, binary, static \u2014 precisely the silent commitment the entry names." } }, { "id": "lis_classification_bias", "dom": "lis", "domLabel": "Library & Information Science", "cluster": "cs", "kind": "open", "form": "normative", "status": "open", "mat": "active", "name": "Can a classification scheme be neutral?", "stmt": "Whether any subject classification can order knowledge without encoding the priorities of its makers, and what follows for the design of knowledge graphs and training corpora.", "prog": "Documented biases in Dewey and Library of Congress subject headings; Bowker and Star on classification as infrastructure; the same question now arises for embedding spaces and retrieval indices.", "barrier": "Any ordering foregrounds some distinctions and suppresses others; neutrality may not be the right ideal, but no replacement ideal is agreed.", "belief": "Neutrality is widely doubted. What should replace it is open.", "review": { "gate": "J\u00fcrge", "date": "2026-07-09", "verdict": "HOLDS", "evidence": "Berman, Prejudices and Antipathies (Scarecrow, 1971) confirmed as the opening documentation of LCSH racism, sexism, and Christocentrism; Knowlton (2005) tracks 39% of Berman\u2019s 225 proposed changes implemented exactly and 24% partially \u2014 bias documented, partially remediated, persistent. Dewey bias documentation confirmed. Bowker & Star standard cite, not separately re-run." } }, { "id": "arch_environment_wellbeing", "dom": "arch", "domLabel": "Architecture, Design & the Built Environment", "cluster": "hum", "kind": "open", "form": "mechanism", "status": "open", "mat": "active", "name": "Does the built environment cause wellbeing, or select for it?", "stmt": "Whether design features of streets, housing, and public space exert causal effects on health, crime, and social trust, net of who chooses to live where.", "prog": "Space syntax offers quantitative predictions; natural experiments in housing mobility find smaller effects than cross-sectional studies; the broken-windows literature is a contested case; age at move matters \u2014 childhood relocations show long-run gains that adolescent and adult moves do not, and the interpretation of the mobility experiments is itself a live dispute.", "barrier": "Residential self-selection confounds nearly every observational design, and randomised assignment of neighbourhoods is rarely possible.", "belief": "Some effects are real; the observational literature almost certainly overstates them.", "review": { "gate": "Caliber", "date": "2026-07-09", "verdict": "HOLDS, enriched", "evidence": "MTO randomized mobility: no significant overall adult employment/earnings effects \u2014 natural experiments finding less than cross-sectional studies, as the entry states; enrichment applied: age-at-move heterogeneity (childhood moves show long-run gains where adolescent moves do not) and the live Ludwig/Massey-era interpretation dispute. Harcourt & Ludwig on broken windows confirmed as the contested case." } }, { "id": "trans_indeterminacy", "dom": "trans", "domLabel": "Translation & Interpreting Studies", "cluster": "hum", "kind": "open", "form": "analysis", "status": "open", "mat": "perennial", "name": "Is there a fact of the matter about correct translation?", "stmt": "Whether rival translation manuals, each compatible with all behavioural evidence, can nonetheless differ in correctness. If not, meaning is indeterminate; if so, what fixes it.", "prog": "Quine's indeterminacy thesis and the argument from radical translation; Davidson's charity-based response; machine translation now produces the rival manuals at scale, which sharpens rather than settles the question.", "barrier": "Any proposed determinant of meaning must itself be specified in some language, and the regress is the argument.", "belief": "The indeterminacy argument is respected; its scope is disputed. Practising translators reject it in practice and cannot say why.", "review": { "gate": "J\u00fcrge", "date": "2026-07-09", "verdict": "HOLDS", "evidence": "Quine confirmed: rival manuals fitting all speech dispositions yet mutually incompatible (Word and Object, 1960, ch. 2). Davidson-charity response is the standard reply; standard cite, not separately re-run." } }, { "id": "demog_fertility_decline", "dom": "demog", "domLabel": "Demography & Population Studies", "cluster": "soc", "kind": "open", "form": "mechanism", "status": "open", "mat": "active", "name": "Why does fertility fall?", "stmt": "The mechanism of the fertility transition. Income, child mortality, female education, urbanisation, and diffusion of norms all correlate; none alone predicts timing across cases, and recent below-replacement fertility in rich countries defies the standard models.", "prog": "Classical demographic transition theory is a description, not a mechanism; the Princeton European Fertility Project found diffusion effects that income could not explain; current sub-replacement fertility was not predicted; later re-analyses contest the Princeton methodology on aggregation grounds, so the diffusion reading is itself contested rather than settled.", "barrier": "The transition happens once per society, so the unit of analysis is small and the confounds are the whole of modernity.", "belief": "The description is settled. The mechanism is not, and current forecasts inherit that weakness.", "review": { "gate": "Caliber Deep", "date": "2026-07-09", "verdict": "STANDS after one correction", "evidence": "Princeton European Fertility Project (Coale & Watkins eds., 1986): associations between standard economic indicators and fertility decline were modest; near-simultaneous transitions across varied environments read as diffusion. CORRECTED: entry now notes later re-analyses contesting the Princeton aggregate methodology \u2014 same one-sidedness class as the settler-mortality fix." } }, { "id": "polad_institutions_geography", "dom": "polad", "domLabel": "Public Policy & Administration", "cluster": "soc", "kind": "open", "form": "mechanism", "status": "open", "mat": "active", "name": "Do institutions or geography explain development?", "stmt": "Whether cross-national differences in prosperity are primarily attributable to institutional quality, geographic endowment, or culture, and whether these are separable at all.", "prog": "Settler-mortality instrumental variables; the reversal-of-fortune finding; sustained critique of the instruments' exclusion restriction; the debate has not converged despite three decades of data; Acemoglu, Johnson and Robinson published a formal reply (AER 2012) defending the results as robust to outlier trimming \u2014 the exchange, not either side, is the state of the field.", "barrier": "Institutions, geography, and culture co-evolved. No instrument is credibly exogenous to all three.", "belief": "Institutions matter. Whether they are the deepest cause is unresolved.", "review": { "gate": "Caliber Deep", "date": "2026-07-09", "verdict": "STANDS after one correction", "findings": [ "CORRECTED: original entry cited the Albouy critique without the AJR 2012 reply \u2014 one-sided; fixed in place", "OPEN (load-bearing): \"no instrument credibly exogenous\" survives as an epistemic claim; falsifier = an instrument passing adversarial exclusion review" ] } }, { "id": "dh_distant_reading", "dom": "dh", "domLabel": "Digital Humanities & Computational Text", "cluster": "hum", "kind": "open", "form": "framework", "status": "open", "mat": "active", "name": "Can distant reading produce interpretation, or only description?", "stmt": "Whether quantitative analysis of large text corpora yields interpretive claims of the kind literary criticism trades in, or whether it can only supply description that a human must then interpret.", "prog": "Moretti's programme; corpus-scale results on genre and influence that no reader could have obtained; sustained objections that the interpretive step is smuggled in at the coding stage; the strongest published critique (Da 2019) is statistical rather than hermeneutic \u2014 that the field\u2019s numbers themselves fail replication.", "barrier": "This is the humanities instance of the verification problem: the computational output is research-level content its readers cannot check by reading.", "belief": "Open, and directly relevant to whether AI-generated humanities scholarship can be evaluated at all.", "review": { "gate": "J\u00fcrge + Caliber", "date": "2026-07-09", "verdict": "HOLDS, enriched", "evidence": "Moretti (\"Conjectures on World Literature,\" NLR 2000) confirmed; Underwood\u2019s corpus-scale genre findings confirmed as the flagship results; enrichment applied: the strongest published attack (Da, Critical Inquiry 2019) is statistical \u2014 replication failures and unhelpful results \u2014 alongside the coding-stage objection the entry already carried." } } ], "results": [ { "id": "educ_testing_effect", "dom": "educ", "domLabel": "Education & Learning Sciences", "cluster": "soc", "kind": "result", "form": "mechanism", "status": "established", "mat": "settled", "name": "The testing and spacing effects", "stmt": "Retrieval practice produces better long-term retention than restudy, and distributed practice beats massed practice. Both replicate broadly across materials, ages, and settings, with smaller and less conclusive effects in mathematics.", "prog": "Among the most robust findings in the learning sciences; survives large-scale replication where much of the adjacent literature does not.", "barrier": "None to the effect itself. Adoption in classrooms remains poor, which is a different problem.", "belief": "Established.", "review": { "gate": "J\u00fcrge + Caliber", "date": "2026-07-09", "verdict": "HOLDS, one softening", "evidence": "Adesope, Trevisan & Sundararajan (2017, Rev. Educ. Research) g=0.61 for practice testing; Rowland (2014) g=0.50; spaced retrieval g=0.74 (Latimier et al. 2021). Softening applied: effects in mathematics are smaller and less conclusive than in other domains." } }, { "id": "educ_bloom_two_sigma", "dom": "educ", "domLabel": "Education & Learning Sciences", "cluster": "soc", "kind": "result", "form": "analysis", "status": "established", "mat": "contested", "name": "Bloom's two-sigma claim, corrected", "stmt": "The claim that one-to-one tutoring raises performance by two standard deviations over conventional instruction has not replicated at that magnitude. Tutoring helps substantially; the two-sigma figure came from small studies with mastery-learning conditions and generous outcome measures.", "prog": "A textbook case of a widely cited number outrunning its evidence, and a standing warning for anyone citing effect sizes from a single influential paper.", "barrier": "None. This entry exists as a corrected claim, not an open one.", "belief": "Overclaimed. Filed here as an established correction \u2014 a contested-as-settled case.", "review": { "gate": "Caliber", "date": "2026-07-09", "verdict": "HOLDS \u2014 3 claims VERIFIED, 1 OPEN", "evidence": "Bloom (1984) Educational Researcher 13(6) confirmed; magnitude non-replication confirmed via Nickow, Oreopoulos & Quan pooled ES 0.37 (WP) / 0.288 (AERJ 2024); Anania & Burke small-study mastery basis confirmed; \"generous outcome measures\" left OPEN." } }, { "id": "sts_lab_studies", "dom": "sts", "domLabel": "Science & Technology Studies", "cluster": "soc", "kind": "result", "form": "framework", "status": "established", "mat": "settled", "name": "Laboratory ethnography as method", "stmt": "Close observation of working scientists established that published papers systematically omit the contingency, negotiation, and craft by which results are produced. This descriptive finding is not seriously disputed, and is separable from the relativist theses sometimes drawn from it.", "prog": "Latour and Woolgar; Knorr Cetina; the descriptive result now underwrites work on reproducibility and methods reporting that scientists themselves endorse.", "barrier": "None to the description. The inference from it to constructivism is the open problem above.", "belief": "Established as description. The strong inference is contested.", "review": { "gate": "J\u00fcrge", "date": "2026-07-09", "verdict": "HOLDS, enriched", "evidence": "Latour & Woolgar, Laboratory Life (Sage, 1979). Enrichment: the 1986 second edition dropped \"Social\" from the subtitle \u2014 the authors themselves performing the separability of description from constructivist inference that this entry claims." } }, { "id": "lis_bibliometric_laws", "dom": "lis", "domLabel": "Library & Information Science", "cluster": "cs", "kind": "result", "form": "theorem", "status": "established", "mat": "settled", "name": "Bibliometric regularities: Bradford, Lotka, Zipf", "stmt": "Scholarly literature concentrates predictably: a small number of journals carry most articles on a topic, author productivity is heavy-tailed, and word frequency follows a power law. These regularities hold across fields and centuries.", "prog": "Descriptively robust. The generating mechanisms \u2014 preferential attachment, and others \u2014 remain partly open, which is why this is a result with an open question attached; exact exponents vary by corpus and strict fit tests sometimes reject the precise form \u2014 the heavy tail, not the exponent, is what replicates.", "barrier": "None to the regularity; the mechanism is not uniquely identified by the distribution.", "belief": "Established as pattern. Mechanism underdetermined.", "review": { "gate": "J\u00fcrge + Caliber", "date": "2026-07-09", "verdict": "HOLDS, one refinement", "evidence": "Bradford (1934) scattering; Lotka (1926) inverse-square productivity from Chemical Abstracts 1907\u20131916; Zipf word-frequency. Refinement applied: exact exponents vary by corpus and formal fit tests sometimes reject the precise law; the heavy-tail shape is the robust part." } }, { "id": "trans_indeterminacy_result", "dom": "trans", "domLabel": "Translation & Interpreting Studies", "cluster": "hum", "kind": "result", "form": "analysis", "status": "established", "mat": "settled", "name": "The argument from radical translation", "stmt": "Quine established that behavioural evidence cannot uniquely fix a translation manual: rival manuals compatible with all dispositions to assent are constructible. The argument stands as an argument; its metaphysical conclusion is what remains disputed.", "prog": "The formal point is widely granted, though Evans- and Fodor-style objections press against even the formal step. Whether indeterminacy of reference entails indeterminacy of meaning, and whether that is a serious cost, is the live dispute.", "barrier": "None to the argument's validity. The dispute is over its soundness and scope.", "belief": "Argument established. Conclusion contested.", "review": { "gate": "Caliber Deep", "date": "2026-07-09", "verdict": "STANDS after one softening", "evidence": "CORRECTED (one-sidedness class, third instance): \"granted by nearly all parties\" softened. Chomsky grants constructibility while reducing it to ordinary underdetermination; but Evans- and Fodor-style objections press against the formal step itself (Wright assesses their cogency). Now reads \"widely granted\" with the objectors named." } }, { "id": "dh_stylometry", "dom": "dh", "domLabel": "Digital Humanities & Computational Text", "cluster": "hum", "kind": "result", "form": "mechanism", "status": "established", "mat": "settled", "name": "Authorship attribution by stylometry", "stmt": "Function-word frequencies discriminate authorship reliably enough to settle disputed cases. The Federalist Papers attribution is the canonical demonstration, since confirmed by multiple independent methods.", "prog": "One of the humanities' genuine verification successes: a contested question of fact, resolved by a method whose error rate can be estimated on known cases.", "barrier": "Adversarial authors can imitate. The method assumes the author was not trying to defeat it \u2014 which AI-generated text now routinely does.", "belief": "Established for non-adversarial cases. Undermined by generation at scale, which is the whole problem.", "review": { "gate": "J\u00fcrge", "date": "2026-07-09", "verdict": "CITATIONS RESOLVE", "evidence": "Mosteller & Wallace (1964) Inference and Disputed Authorship confirmed real, function-word Bayesian attribution of disputed papers to Madison; independent later methods concur." } }, { "id": "demog_malthus_refuted", "dom": "demog", "domLabel": "Demography & Population Studies", "cluster": "soc", "kind": "result", "form": "analysis", "status": "established", "mat": "settled", "name": "The Malthusian trap, escaped", "stmt": "Sustained per-capita income growth alongside population growth refuted the claim that population necessarily outruns subsistence. The mechanism of escape \u2014 technological change outpacing population response \u2014 is agreed; its timing is not.", "prog": "The pre-industrial regime fits the Malthusian model well; the post-1800 divergence does not. The model was right about its era and wrong as a law.", "barrier": "None. Filed as a refuted general claim that remains a good local model.", "belief": "Refuted as a law. Retained as a description of the pre-industrial regime.", "review": { "gate": "J\u00fcrge", "date": "2026-07-09", "verdict": "HOLDS", "evidence": "Clark (A Farewell to Alms, 2007) confirms the frame: no systematic living-standard gain before ~1800, break dated variously 1600/1800/1860 \u2014 the entry\u2019s \"timing is not [agreed]\" is exactly right." } }, { "id": "polad_campbells_law", "dom": "polad", "domLabel": "Public Policy & Administration", "cluster": "soc", "kind": "result", "form": "mechanism", "status": "established", "mat": "settled", "name": "Campbell's law", "stmt": "The more a quantitative indicator is used for social decision-making, the more it will distort and corrupt the processes it was meant to monitor. Documented across education testing, policing statistics, hospital targets, and research metrics.", "prog": "Repeatedly confirmed. It is also the reason a verification regime must not become the metric it optimises \u2014 this Atlas included.", "barrier": "None. The law constrains the design of every evaluation system, including ours.", "belief": "Established, and directly self-applicable.", "review": { "gate": "J\u00fcrge", "date": "2026-07-09", "verdict": "CITATION RESOLVES", "evidence": "Campbell, \"Assessing the Impact of Planned Social Change,\" Evaluation and Program Planning (1979); wording confirmed verbatim against multiple sources; formulated in the 1976 Dartmouth paper." } }, { "id": "arch_jacobs_streets", "dom": "arch", "domLabel": "Architecture, Design & the Built Environment", "cluster": "hum", "kind": "result", "form": "framework", "status": "established", "mat": "contested", "name": "Mixed-use streets and informal surveillance", "stmt": "Jacobs's claim that mixed-use, densely trafficked streets support informal social control has substantial observational support and remains the working consensus in urban design. Causal identification is weak, and the adjacent broken-windows programme does not replicate cleanly.", "prog": "Widely adopted in practice; the evidential base is largely observational, and the policy applications diverged sharply from the original claim; empirical tests split \u2014 Philadelphia mixed-use blocks show less crime while an LA zoning study found purely residential blocks safer than either, so even the sign is contested.", "barrier": "Self-selection into neighbourhoods; the treatment is not assignable.", "belief": "Working consensus, weak identification. Filed as established-with-caveat, not settled.", "review": { "gate": "Caliber Deep", "date": "2026-07-09", "verdict": "STANDS, enriched", "evidence": "Adversarial pass: entry\u2019s two-sidedness vindicated by the literature itself \u2014 Philadelphia analysis finds mixed-use blocks with less crime while an LA zoning study finds purely residential blocks safer than mixed-use (\"had it backwards\"), and reviews report no agreement on the sign. Enrichment applied naming the split." } } ] }