Hamann (“Metacritique”) says that “words as undetermined objects of empirical concepts are entitled critical appearances, specters, non-words or unwords, and become determinate objects for the understanding only through their institution and meaning in usage. This meaning and its determination arises, as everyone knows, from the combination of a word-sign, which is a priori arbitrary and indifferent and a posteriori necessary and indispensable, with the intuition of the word itself; through this reiterated bond the concept is communicated to, imprinted on, and incorporated in the understanding, by means of the word-sign as by the intuition itself.”

But not everyone knows this. It’s not clear, for instance, that Saussure knows it.

Saussure grasps the first part of Hamann’s description - that the sign is arbitrary and indifferent; but he loses track of the a posteriori necessity of the association of word-sign with the intuition of the word itself. Hamann is surely right; word-signs are arbitrarily affixed to concepts or things only in their origins (and not entirely then even - neologisms borrow from the existing stock of words), but once the signs are affixed, they become necessary.

Hamann is also superior to Saussure in his emphasis on the role of custom and repetition in the process of “imprinting” concepts on the understanding. Binding a word-sign to an “intuition” doesn’t by itself constitute language. If I call the crevasses in my knuckles “garvels,” and no one else does, the association of word-sign and concept/thing doesn’t communicate knowledge or count as a part of the language. Others need to adopt ths same usage, and use it again and again.