| Out of curiosity, have you spent time in academia? > The rationale behind publishing in long-form journal articles is not really valid in our century. No, it's perfectly valid. Some ideas can't be expressed in short form. A 5-page long proof requires a 10-20 page article to properly motivate and provide context/explanation. An evolutionary or piece-by-piece doesn't make any sense; you'll end up with fragmented dead ends. Waste of everyone's time. > and findings are typically summarized in their 3-4 figures anyway I'm not really sure which fields you're referring to, but I assure you that you're over-generalizing. This isn't even remotely true even for very empirical sciences. > There are proposals for new approaches Usually these are appropriate only for a single subdomain or methodology and only provide one particular and extremely opinionated view on the results. Such as the proposal in the article you posted. The overhead isn't worth it and there's a huge risk the relevant field(s) move quickly enough that the presentation method becomes obsolete before it becomes useful. > It should be possible for each finding to cite other findings it relied on ...I don't know what to say to this... > thus creating a graph of influence/significance Yeah, we have this. It's called bibliometrics. UNIVERSALLY HATED by anyone who's not a bean counter. Good people ignore them. Bad people optimize for the metrics and it becomes a stupid game that has nothing to do with the thing you're trying to measure. > The current situation with citations often involves friends citing friends , so i don't consider it reliable This is just one reason among many that bibliometrics (a.k.a. any necessarily poor attempt at "a graph of influence/significance") are a poor mechanism by which to judge science... > It should also include open questions , thus giving research directions. In many fields literally every paper includes this. In most it's rather obvious to anyone who comprehends the paper what the next steps are. |