Why does De La Cruz, Ed lose its given name? #575

After a comma, C1 decides between the listing form (Family, Given) and trailing credentials (Name, PhD). Part of that decision is counting the words before the comma: two or more name words means the name is complete, so what follows is a credential. The count treats each particle as a name word. A particle surname like De La Cruz or van der Berg is a single family name (P2 joins it), but it counts as two or three words. So the comma reads as a credential comma, and either the given name disappears or the parse splits the particle chain.

• The ambiguous-class rows (Ed, Ma, MA) are a regression against 2.3.0 that hasn't been released. This cycle's name-word count (Use letter case to disambiguate suffix vs. surname: "Jack MA" vs "Jack Ma" #289, Should a credential run after a comma keep its comma when the last word is an ambiguous acronym? John Smith, PhD MEng reads first PhD, last John Smith #544) brought back 1.4.0's readings, and for particle surnames those are wrong. The release-log bullet that says John Smith, MA gives a suffix "which is what 1.4.0 read" is right for full names and wrong for these.
• The unambiguous row (van der Berg, PhD) is long-standing: it comes from v1's "more than one word before the comma" guard, and every release splits inside the chain.
• For De La Cruz the parse is wrong under either reading of the comma: De La Cruz is all surname even standing alone. The van rows are different. Standing alone, van der Berg and Van Buren read given van, family der Berg/Buren, because a leading van can be a first name (P1's fork). So the split is the standalone reading plus a credential, not a reading that is wrong in itself. The fix lets the comma settle that fork, since the listing form puts a surname before it (decided below).

Proposed rule

Rationale: a particle joins the name word after it (P2), so a particle chain and its head are one family name, not several name words.

Statement: wherever C1 counts the words before the comma, a particle and the name word it attaches to are one word. The particle reaches that one word and no further, as P1's fold counts. A connective join and a bound given name are not counted as one word.

Examples (target readings):

"De La Cruz, Ed"         →  given="Ed"  family="De La Cruz"        (2.3.0's reading)
"van der Berg, MA"       →  family="van der Berg"  suffix="MA"     (as "Smith, MA" reads)
"Van Buren, Ed"          →  given="Ed"  family="Van Buren"
"van der Berg, PhD"      →  family="van der Berg"  suffix="PhD"    (as "Smith, PhD" reads)
"De La Cruz, M.J. K.L."  →  given="M.J."  suffix="K.L."            (decided, see below; as "Cruz, M.J. K.L." reads)
"John van Buren, Ed"     →  suffix="Ed"                            · unchanged: two name words
"Lloyd Webber, Ed"       →  suffix="Ed"                            · unchanged: no particle; #563's accepted hazard
"Ortega y Gasset, Ed"    →  suffix="Ed"                            · unchanged: a connective join is not one word
"abdul Salam, Ed"        →  suffix="Ed"                            · unchanged: a bound given name is not one word

Decided

• De La Cruz, M.J. K.L. moves to given M.J., suffix K.L. (Derek, 2026-10-01), reading as Cruz, M.J. K.L. does. It is a rules.md#C1 example today, where two dotted pairs after De La Cruz flip to a credential and report. With De La Cruz counting as one word the count no longer reaches it. That is acceptable. C1's sentence that two dotted groups are "a credential however little else speaks for them" is amended to sit behind the count, and the example moves with it.
• The comma settles P1's fork for a leading van (Derek, 2026-10-01). Van Buren, Ed reads given Ed, family Van Buren, and Van Johnson, Dr. reads family Van Johnson, although each part standing alone reads given Van. The listing form puts a surname before the comma, which is evidence the standalone parse does not have.
• Particle chains only (2026-10-01). The first version of this proposal also counted connective joins (P3) and bound given names (P5) as one word. Both are left out. P3 already keeps a one-letter connective in a three-word name as its own word, so Ortega y Gasset doesn't join even standing alone. A bound given name builds a given name, and P5 gives up a family word when the name has no other (abdul Salam alone reads given abdul, family Salam).
• Name orders: the particle reaches only the word it attaches to, so de Mesnil Jean, Dr. under FAMILY_FIRST stays family de Mesnil, given Jean, two words.

Blocks #564: that rule's comma-position caps reading would inherit this count, so Van Buren, MARTIN and De La Cruz, MARIA move with it.