You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Should a parsed title say what kind it is, and a word its gender? (Dr. Mr. Sir John Smith reads three titles the same) #1461
The question. The vocabulary records a title's kind (salutation, academic, religious, nobility, military, office, occupation, kinship) and, where a word gives a definitive clue, a gender (bint, Mrs, doña). Neither reaches a caller. Every title token reads the same:
>>> [t.reading.kindfortinparse("Dr. Mr. Sir John Smith").tokens[:3]]
['title', 'title', 'title']
A suffix token, by contrast, already splits four ways (generation, credential, honorific, post-nominal).
What it would allow. A caller could pick the salutation out of a title run to write Dear Mr. Smith, or drop the academic title and keep the courtesy one. CLDR's single title field does not model this.
The shape proposed.Reading gains one or two optional fields, filled only where the reading came from the vocabulary. A title read by its shape (Insp.) or added by a caller with the generic kind="title" has none, so an empty field means unknown, not "has none".
Open:
Is there a caller for this? No report has asked for either. This issue exists so the RFC's open question has a home; it is not scheduled.
One field or two: the vocabulary kind and the gender are separate facts and could land separately.
One kind or a set: none of the 816 title words carried two title kinds on 2026-10-28 (the RFC's count, not re-taken).
Whether a view should name the salutation alone, as generation and credentials split the suffix.
Gender is recorded only on words that state it (decision 9). It must never be inferred from a given name, and the docs would have to say so.
Each field is an ADDED_IN field on a pickled type.
Activity