18. Whether the author stated the rule is its own field

Context

attribution records how an entry's source was established — traced to an origin, attested in a standard reference, or traditional. It says nothing about whether anyone ever set the procedure down as a rule, and the two questions come apart.

Harsdörffer wrote instructions to his bookbinder and a numbered list of what each ring carries: the rule is his. The sonnet has no author who declared it; prosodists codified a practice already in use. Jean Paul titled a notebook Ideenwürfeln and left a structure in it, but never wrote the rule down — any procedure taken from it is a reconstruction.

Without a field for this, a reconstruction would enter the catalogue looking exactly like a rule its author published.

Decision

Attestation = Literal["author-stated", "codified", "reconstruction"]

author-stated is conservative: using a form is not stating it, so Perec writing La Disparition does not make the lipogram author-stated. A row claiming it must cite a source with a year, which a test enforces.

The expansion brief proposed two values. Three are needed, because most of this catalogue is inherited forms that no author ever stated and that are not reconstructions either — 110 of 148 rows are codified.

Consequences

The Jean Paul chapter can be honest about what it is doing. Every procedure taken from the excerpt books will carry reconstruction, and a reader will be able to tell it from Harsdörffer's rings at a glance rather than by reading the prose.

It also makes a claim the catalogue could not previously make: 38 rows record that their originator published the method, which is a different and stronger statement than recording where the form is documented.

Nothing carries reconstruction yet. A value with no members is usually a smell; here it is the point, since the chapter that needs it is the next one.