Well noticed :-). Both of those are discussed at length in the book.
First, "ways of talking" have domains of applicability. GR's domain of applicability doesn't include the very small scale, so it doesn't make any predictions there to contradict qm.
The book also talks about value judgements, and is very explicit that different people's core values may differ from each other. So I guess poetic naturalism doesn't apply to them? That bit wasn't clear, but I would only apply these axioms to "is", not "ought" statements.
First, "ways of talking" have domains of applicability. GR's domain of applicability doesn't include the very small scale, so it doesn't make any predictions there to contradict qm.
The book also talks about value judgements, and is very explicit that different people's core values may differ from each other. So I guess poetic naturalism doesn't apply to them? That bit wasn't clear, but I would only apply these axioms to "is", not "ought" statements.