Writing isn’t language at all, for reasons discussed in my comments below.
Which is part of what makes linguistics work on ancient languages so difficult - we’re having to use these imperfect symbols, which themselves aren’t language, to try to glean as many features about the actual grammars they’re intended to represent, which are language.
This is why we know much less about ancient languages than we do modern ones - because we have actual recordings of modern languages (the recordings themselves are also not language, of course; they just encode language much better than writing does), so we can get at many more features of the language in question.
Writing isn’t language at all, for reasons discussed in my comments below.
Which is part of what makes linguistics work on ancient languages so difficult - we’re having to use these imperfect symbols, which themselves aren’t language, to try to glean as many features about the actual grammars they’re intended to represent, which are language.
This is why we know much less about ancient languages than we do modern ones - because we have actual recordings of modern languages (the recordings themselves are also not language, of course; they just encode language much better than writing does), so we can get at many more features of the language in question.