But when in the first setting out he takes it for granted without proof that distinctions found in the structure of all languages have no foundation i
