{Disarmed} Interpretation of occurrence in multivalued attributes

Hi,
In section MailScanner has detected a possible fraud attempt from “5.3.5.1” claiming to be MailScanner warning: numerical links are often malicious: 5.3.5.1 of Adl spec 1.5, draft 4, examples to both single and multi valued attributes are given within the context of subtype constraints.
It is explicitly stated in the doc that lack of occurrence in a single valued attribute where multiple options are given (with different node ids) this represents alternatives for that attribute.

Hi,
Is lack of occurrence interpreted in the same way for multivalued attributes (with cardinality)?

What is the meaning of this case: a multivalued attribute with cardinality where a member has occurrence and two others do not have? Two being alternatives and the one with occurrence being a regular member (if this is the correct term). An example to this would be:

counter_parties cardinality ∈ {*} ∈ {
PERSON[at0004] occurrences ∈{…} ∈ {
date_of_birth ∈ { … }
}
ORGANISATION[at0005] ∈ {
date_of_registration ∈ { … }
}
PARTY[at0006] ∈ { … }
}

The semantics of the below example is a little bit confusing for me, unless it is plain wrong. Introduction of occurrences constraints the number of instances of type person under counter_parties, but now we have no comments on number of instances of organization and party and more, if they are alternatives or not, and if they are, what is the constraint on the chosen one?

Should I conclude that use of occurrence in a member of a multivalued attribute necessitates its use on other members? Is this stated somewhere in the specs?

Ps: I’ve just realized that the term “multiply-valued” is used in the spec, so replace “multivalued” above with it please

Kind regards
Seref