Two major sides: 1. Reputation. Journals are a provider, mainly, of a note saying "this paper is worth reading". This reputation is calculated largely by how m

Two major sides: 1. Reputation. Journals are a provider, mainly, of a note saying "this paper is worth reading". This reputation is calculated largely by how much the papers they publish get cited later on. Due to the speed of iterations of science (how long it takes for a published paper to be used in more work, and then that new work to be published) it takes a long time to get past this. 2. Organised peer review. Despite the reviewers working for free, it's still not a free process. There's a great twitter thread about this from someone at peerJ (founder?) but I can't find a link now. Getting more and more common is to pay the publisher and have the work made openly available, or at least to deposit the work in an institutional repository. [disclaimer, I work in the science & publishing area for Digital Science, but on analysis of data not the publishing itself]