A word-initial apostrophe reports unbalanced-delimiter (Pieter 't Hart gives family 't Hart and reports it) #1241

parse("Pieter 't Hart") reads correctly, given 'Pieter' and family 't Hart, but it reports unbalanced-delimiter. The same happens with Pieter van 't Hof, Hart, Pieter 't and Jan 's Jacob (the last also reads middle 's). 't is listed (nl.toml, an article, never-given), and the reading never treats the apostrophe as a quote. The report is the only thing wrong.

N2 (same-character quotes) reads an apostrophe inside or at the end of a word as a letter (O'Neil, Aube'). A word-initial apostrophe opens a candidate quote that never closes, so the report fires. The curly ’t doesn't report.

Measured (2026-11-08): in the Tweede Kamer's open data, every member whose name holds 't or 's- is reported (van 't, 't, Storm van 's-Gravesande). In ORCID NL, 12 family boxes open with 't, 48 with van 't and 10 with in 't. A report on a name nobody finds ambiguous teaches callers to ignore the field (AGENTS.md, "a kind is worth adding only if a reader would hesitate").

Question: Should a listed word that starts with an apostrophe ('t, and 's if it's listed) be read as a word before the quote scan, or should N2's carve-out cover an apostrophe followed by a single letter and a space?