Corroboration · An audit of the record

The Classification Record

The paper trail behind the Classification of Finite Simple Groups, censused with respect

A register of registers, third artifact

The Classification Record

The proof of the Classification of Finite Simple Groups runs to tens of thousands of pages across half a century and a hundred authors. It is accepted. What nobody keeps is a living account of the record it rests on: what was cited, whether those citations were ever published, and whether the account is still current.

507

references censused

across two foundational bibliographies, machine-readable

8.1%

not published literature

a currency statement, never a correctness one

7

fixable defects in the record

all repairable from public records, none touching the maths

1

hedge that dies under compression

candour survives where experts address experts

Why this one is different

The other two registers on this site record claims to open problems. This one is the opposite. Here the theorem is settled and the mathematicians are right. The subject is the record, and the finding is quiet: a proof is only ever as current as the last time someone updated its paper trail by hand.

Philosophers of mathematics have argued that whether the Classification is really understood by the community is an empirical question, not a matter of declaration. A census of what the proof actually cites, and of how faithfully that record has been carried forward, is one instrument for answering it. This is that instrument, pointed at the proof's own bibliographies.

The headline, in one picture

Published versus non-published references, by bibliography

Surveys 172 (2011)Surveys 172 (2011): 334 published of 352Surveys 172 (2011): 18 not published of 352334 published18 not (5.1%)Gorenstein 1986 surveyGorenstein 1986 survey: 132 published of 155Gorenstein 1986 survey: 23 not published of 155132 published23 not (14.8%)
Published literatureNot published literature

Non-published is a currency statement, not a correctness one. A thesis or preprint is not an error; it is a reference the record stopped tracking.

Show the numbers
BibliographyPublishedNot publishedTotalNot published %
Surveys 172 (2011)334183525.1%
Gorenstein 1986 survey1322315514.8%

Where to go next