I would argue mathematics isn't really a science, as it isn't outward looking to the world, rather it is a study of self-consistent rewriting rules. You can obtain any mathematical result without knowing anything about the universe (something also true for, lets say, analytic philosophy).
Sure, but then the distinction between science / not-science hasn't got that much value.
A not-science (math) is the most critical tool in the scientific toolbox, and hence contributed more to scientific progress than any isolated branch of science.
Sure, natural languages also contributed "more to science" than "any isolated branch" but learning a language (for its use rather than linguistics/anthropology) isnt a science. I think it is a useful distinction.
