> Should P88 be defined inversely (P88 forms part of (consists of) and then be subp of P89, > or should that be in an RDF recommendation
I think it should be made a subproperty in the standard, and then in RDF. It would be better not to invert it (to avoid the need for data migration), but if you cannot have "forward is sup-property of backward" in the standard; or it would be too confusing: invert it.
