Should capitalized() write the Dutch IJ digraph under a Dutch choice? (PIETER IJZERMAN gives Pieter Ijzerman) #1242

str(parse("PIETER IJZERMAN").capitalized()) gives Pieter Ijzerman. Dutch capitalizes the digraph ij as a unit at the start of a word: IJzerman, IJsselmuiden, van IJzendoorn. Mixed-case input is left alone (IJsbrand IJzerman keeps its case), so only case repair is affected: one-case input, and capitalized(force=True).

Measured (2026-11-08): 8 Tweede Kamer members' surnames contain an IJ word. ORCID NL and BE writers spell IJ 38 times, Ij 7 and ij 2.

A default rule would be wrong elsewhere: Igbo Ijeoma and Urdu Ijaz are written Ij.

Questions:

• Make it a case rule that applies only under a Dutch choice (parser_for(..., languages=["nl"])), the way R4 (case repair reads the unjoined mark) keys particle case on vocabulary?
• Or leave it, since the writers themselves get it wrong 7 times in 45?