CaGEL* explains the concept of "indirect complements" on page 443 as follows:
If it's the complement of a noun, be it direct or indirect, it's part of a noun phrase (NP) headed by the noun, right? So, all three underlined "indirect complements" must be part of the respective NPs.
Now, regarding [8ii], CaGEL adds this:
Here, we have this year added between so great a loss and that we're likely to go bankrupt, thereby rendering the latter (previously an indirect complement in [8ii]) not within the NP, as duly noted in the second quote above.
Now I wonder whether the "indirect complement" of a noun should be treated as part of the NP in the first place. The sole reason for construing it as part of the NP, I think, is because "the licensor is another dependent of the head (or part of one)."
But the licensor of the quoted example sentence including the intervening this year is so (i.e., part of another dependent of the head), and yet in this example sentence, the content clause (that we're likely to go bankrupt) is considered outside the NP, which means it's not a complement anymore, direct or indirect.
So my question is whether there is such a thing as the "internal complement" of a noun in the first place.
Where the indirect complement is outside the NP due to an intervening adjunct like "this year", the content clause is still labelled as a complement (an indirect one) since it is licensed by the head of the earlier NP.
Here's how CaGEL defines a complement of a noun (page 439):
That is, both complements and modifiers are internal dependents, which are by definition "contained within nominal constituent".
Given this, I wonder how @BillJ can call the content clause "a complement" when the content clause is not "contained within nominal constituent".
*The Cambridge Grammar of the English Language by Huddleston & Pullum