I said academia was conversational, but these are not normal conversations. First because the standard intervention is a 10,000 word long monologue. Second because journals curate what is good enough to be allowed into the conversation using peer-review. These journals are another layer in the academic status economy. They try to publish those monologues of most interest to most people and thus most likely to increase the status of their journal within those academic conversations. The higher the status of a journal the more able it is to bestow prestige upon those who publish in it, and so the more academics will send it their most exciting ideas. A virtuous circle of prestige appears around the journal, which translates into outsized profits for its multinational corporate owner - since every university needs to subscribe to it or risk missing out on the most exciting part of the conversation.
If we were building it from scratch, I don't think we would set up a system like this in which publicly funded research is privatised and commoditised, resulting in enormous profits for a handful of companies while systematically excluding ordinary citizens, journalists, and poorer country universities from access. There are many attempts to change things, from protests against price-gouging behaviours, to boycotts of the most loathed company, Elsevier, to outright defiance (pirate sites like scihub and libgen), to the creation of free Open Access alternatives. I wish them well.* Here I want to open up a new front that is not about access but how the product gets made: the treatment of peer reviewers