Indeed. To be precise, I should have written that the lack of X in a function's type assures you that the function does not cause X effects.