"It must show, for example, that the history of grammar is not the projection into the field of language and its problems of history that is generally that of reason or of a particular mentality, a history in any case that is shares with medicine, mechanical sciences, or theology; but that it involves a type of history - a form of dispersion in time, a mode of succession, of stability, and of reactivation, a speed of deployment or rotation - that belongs to it along, even if it is not entirely unrelated to other types of history." The statement from Foucault's The Historical A Priori and the Archive reminds me of how languages are learned by native, second, and third learners. Foucault has a really great and accurate opinion on grammar of languages. In modern time, the language of English is in another stage. That said, English has come along way and evolved into the model that fit with the current timeline. From the past, English had different grammars in different formats for each specific time of period. In each distinguish time of period, English fitted their language learners' style of learning and how the entire nation thought about knowledge. Of course, it creates gaps for the latter ones to learn the previous ones (e.g. modern English learners to learn the grammar in Romantic period), that also creates another gap of difficulty for current language learners to communicate with the ones after them, since the current ones always judging the latter ones' knowledge. 

That lead to a good point of view, which is an archive should be emphasize on quantity or quality? There is no guarantee an archive could be everyone wished for. It is like a society, a society itself is an archive, an archive with many categories; but if we have to narrow down the topics, the common answer will be humanity. In humanity, there are good and bad, legal and ilegal. But how do we build this archive? It was the same in WWI, there were some soldiers who did not want to go to war, and some were in contrast.