Repository navigation
Should Doe, Jane Mrs read Mrs as a title rather than a middle name? #835
Description
The shape
After a family comma, a salutation written without a period at the end of the given part reads as a middle name. The same word with a period reads as a title, and so does the bare word when a suffix stands between it and the given name:
"Doe, Jane Mrs." → title="Mrs." given="Jane" family="Doe"
"Doe, Jane Mrs" → given="Jane" middle="Mrs" family="Doe"
"Doe, John Q Mr" → given="John" middle="Q Mr"
"Doe, John Jr Mr" → title="Mr" given="John" suffix="Jr"
"Doe, John Mr Jr" → given="John" middle="Mr" suffix="Jr"
"Doe, Mrs" → title="Mrs" family="Doe" (H1)
", Jane Mrs" → given="Jane" middle="Mrs"
This is common in real data. In the FEC's itemized-contributions file, where filers' software writes FAMILY, GIVEN MIDDLE SALUTATION in capitals (file dated 2026-10-18, measured 2026-10-22):
| Word | Reads as middle | Reads as title |
|---|---|---|
MR |
8,483 | 109,889 |
MRS |
1,902 | 24,619 |
MISS |
1,103 | 33 |
DR |
1,034 | 11,582 |
The title column is mostly the period-marked spelling. Across every title word, 13,243 names end with the bare word read as a middle name.
Proposed rule (H5 amendment, or a new H rule)
When: a family comma has fixed the family name, and the given part ends in a bare word whose title reading is a salutation, with at least one name word in front of it. Examples: Doe, Jane Mrs and DOE, JOHN Q MR.
Then: the word joins the title field, and the given part reads as it would without it.
Because: H5 (the trailing title run) keeps a bare trailing title word as a name word because "TITLES holding ordinary surnames" (Mary Jane King keeps family King). After a family comma that reason is gone, since the surname is already settled. Only the middle-name reading is left, and a salutation is not a middle name: it is a form of address, not a rank or office someone could be named after. The suffix path already reads it this way (Doe, John Jr Mr), so this would make the two agree.
Except:
- A title/suffix dual stays with the suffix run.
Doe, John Srkeeps suffixSr, andDoe, Jane Mskeeps suffixMs(44,920 FEC names). WhetherMsshould move is a separate question about the dual. - A non-salutation title stays a name word.
Doe, John King,Doe, John BishopandDoe, John Deankeep middleKing,BishopandDean. Office and rank words are borne as middle names. - A lone salutation keeps its H1 reading (
Doe, Mrs, title and family).
Accepted: a salutation that is someone's actual middle name, after a family comma, would read as a title.
Open questions
- Which salutations clear the bar? The
salutationkind holds about 120 words. Several needtools/name_usage.pycounts before they qualify, since they may be borne as given or middle names:master,madam,goodman,familie,herr,frau,signore,bwana,श्री. The kind compiles only intoTITLEStoday, so a predicate would need a property or set built from it. - Comma shapes not yet swept: an empty head (
, Jane Mrs, which reads as middle today), a nickname-only head, a particle surname (van Doe, Jane Mrs), a second comma (Doe, Jane Mrs, PhD), and all threename_ordervalues. The family-first order hasn't been measured at all yet. - Without a comma, the shape stays as it is:
Jane Doe Mrsreads familyMrs, which is H5's surname case, and nothing here changes it.
Where it was found
The fec-individuals local corpus, #834.
Activity
Address the PR review of #835: tests and comments
Tidy the #835 review fixes
Read a salutation ending the given part after a family comma as the t…
Release 2.6.0 (#880)