Sheik abdul salam splits the bound given name that abdul salam ahmed joins #369

A bound given-name prefix joins the following word into one given name — abdul salam ahmed gives given abdul salam. After a title, that join does not happen, and only one word becomes the given name:

Sheik abdul salam        title='Sheik'  given='abdul'      family='salam'
Sheik abdul salam ahmed  title='Sheik'  given='abdul salam' family='ahmed'

The join needs three pieces that are neither title nor suffix (BoundJoin.STRICT), so with a title and two names it declines.

Sheik Abu Bakar looks like a counter-example but is not:

Sheik Abu Bakar     title='Sheik'  given='Abu Bakar'  family=''

That works only because abu is also a particle, so the particle chain fires — nothing to do with the bound given-name vocabulary. abdul is bound-given and not a particle, and it does not chain. Confirmed by the piece structure:

Sheik Abu Bakar     [Sheik][Abu Bakar]        head tags: particle, vocab:bound-given, vocab:particle-ambiguous
Sheik abdul salam   [Sheik][abdul][salam]     head tags: vocab:bound-given

Why it matters now

#367 makes titles transparent to the leading-particle exception, which removes the particle chain in Sheik Abu Bakar and regresses it to given Abu, family Bakar. The reading most people want — given Abu Bakar — would then have to come from the bound given-name rule, which is where it should have come from all along.

The question

Should a bound given-name join be allowed after a title with only two following pieces? That would give:

Sheik Abu Bakar      given='Abu Bakar'    family=''      (restored, for the right reason)
Sheik abdul salam    given='abdul salam'  family=''      (changed from given='abdul', family='salam')

The second row is the real decision: is Sheik abdul salam a two-word given name with no surname, or given abdul plus family salam? The same question the BoundJoin.STRICT threshold answers "reserve a piece for the family" everywhere else.

Note the interaction with GIVEN_NAME_TITLES: Sheik is a given-name title, so it already asserts that what follows is a given name. That may be the signal that licenses the join past the threshold — in which case the rule would be narrower than "after any title".

Scope

Default-order change; needs a tools/differential pass with ledger entries.

Split out of #367 so that change stays one mechanism.