53 entries
The transfer records that survived the gate. One structure per entry: how the failure works, the apparatus that resolved it elsewhere, the near-twin that would invalidate it, and the test that separates them.
Every candidate we take seriously is written into a transfer record before anyone acquires data: the mechanism, the apparatus that resolved it in its own field, the symptoms it would produce here, the false friend that would invalidate the transfer, and the observation that separates the two.
The false friend is the load-bearing field. Every attractive idea has a near-twin that looks the same at first glance and means something else. If the author cannot name that near-twin before the result is known, the work is still an analogy rather than a hypothesis.
These are the records that survived. They are published in the open because a method that cannot show its own working is asking for trust it has not earned — and because the counts in Blueprint 01 are read off this corpus.
Butterfield named the habit of writing history from its ending in 1931, and it is still the default shape of a textbook. Brinton built the template that produces it, and made it useful by recording where his own cases refused to fit.
famous problems analytic number theory
Goldbach's conjecture has been confirmed for every even number up to 4×10¹⁸ and is still a conjecture. Number theory has an unusually exact account of what a checked range is evidence about, and it is not the length of the range.
political movements revolutions
Chartism failed on every attempt it made and five of its six demands are law. The outcome arrived up to eighty years late and was credited to other campaigns, which is what happens to a claim with no scheduled resolution date.
enumerative combinatorics probability theory
Any five distinct numbers contain three that rise in order or three that fall in order. Erdős and Szekeres computed the exact length that forces it, which turns a run into a measurement of the window rather than of the series.
The hypothesis test taught in every introductory course is two rival frameworks spliced together by textbook writers after their authors stopped negotiating. The seam is visible in ordinary use, and the thing it hides is who holds each error.
A number that follows from a definition, a number settled by measurement, and a number that depends on an unstated choice get printed to the same decimal place. Asking what more effort buys separates all three.
Joan of Arc has one of the densest records of any medieval life, and her prosecutors produced it. Four properties get collapsed into the phrase well documented, and they trade off against each other rather than moving together.
Plimpton 322 is dated to a thirty-eight-year window and its purpose has been contested for a century. The field prints both facts at their own confidence, which is an apparatus, and the reason it can is that the loss of context has a date.
probability distributions experimental design
A distributional estimate can be mathematically tidy while the design that generated its observations left whole groups outside it. Distribution parameters and experimental allocation are two halves of the same claim.
genealogy historical documents
A well-sourced chain can establish that a record was made without establishing the underlying biological or economic relationship. Genealogy's source hierarchy is a discipline for naming that gap.
A function can be continued beyond an observed region only because it belongs to a class that constrains every extension. Complex analysis makes the constraint exact; a fitted curve should disclose when it does not have one.
mathematical logic computability
Finite examples can make a rule engine look complete even when no procedure can decide every case it will meet. Logic separates consistency, completeness and decidability; computability supplies the hard boundary.
A normalization rule can make two records look identical by removing the marks that kept them apart. Scripts and language records make the loss visible because they retain the original form alongside a reading.
A claim may be widely repeated before its first recoverable source, and a source may be old without supporting the modern version. Pseudohistory and folklore supply separate tests for the story, its attestation and its transmission.
Agreement on the text does not produce agreement on the rule for reading it. Textual canons and theology show that disputes often begin after the primary record has been fixed.
A calendar can contain an observed feast day, a legal date and a settlement convention that move differently. Calendar rules and festival records make the authority behind each date part of the datum.
combinatorial designs graph algorithms
A shortest or cheapest route can look diversified because it has many hops while repeatedly relying on one untested pairing. Combinatorial designs supply the coverage constraint that graph optimization does not.
Names are evidence of identity, not identity itself. Naming systems preserve crossovers, spelling variation and publication thresholds that make a clean join look more certain than it is.
historical geography historical places
A name can survive a border change, a conquest and a map revision while the thing it denotes changes. Historical geography makes the mapping explicit; a time series should do the same before it compares locations across years.
explorations historical geography
A line between two points is not a journey until it carries mode, season, authority and the source that drew it. Explorations and historical geography keep the route and the later map from becoming one fact.