"Semantics, like mathematics, is not an 'evolved discovery' of some independent external thing or "law" but the outcome of rough cognitive patterns noticing rough cognitive patterns. This limits what we can say about the existence of formal semantics to the concrete; it really exists only in the brains that know of it, in other words. The aforementioned rough patterns are progressively refined and given further layers of abstraction - by these same brains."

I think the part about maths is a fallacy. Semantics are going to be highly culturally linked. Maths will be lightly linked.