Academia is conservative. It's got a central tendency which results in a highly-coupled federated system with a rather high entrance barrier. One of the requirements has so far been that you must follow a scientific supervisor... until it's time to lead. The new communication capabilities make us question the position of this and many other established attributes of Academia. We ask why it's so and conclude that the scientific community needs to learn a decentralized way from software developers
Sat, Feb 16 2018, visiting AG Baskakov I finally get to try and discuss these problems as well as the perspectives of research in and outside Voronezh.
Here I will just outline some propositions of AG Baskakov addressed to a student standing just in the begining of his path. I believe these propositions pretty much describe the way Academia is today. The quoted are the translated pieces of our dialogue and the rest are my comments.
- "You won't achieve anything without a supervisor", i.e. you must begin with a
- "Try and find any recognized mathematician who made it his own way. You won't find one! There was Gallois but he grew up in France", i.e. an appeal to history.
- "You also don't need anyone besides a supervisor", it suffices to have a supervisor.
- "You don't need people of your generation" --- you could think you need a competitive environment, someone to share and exchange ideas with, an inspiration for something completely new, or solving some applied problems in "background", but you really don't. These things are optional.
- It doesn't matter if there aren't in the city many other people doing real research or if the most researchers at the university aren't exactly in tune with the rest of the world. Apparently, you're supposed to abide.
How it came to this
"If I have seen further it is by standing on the shoulders of Giants"
It is known that Science isn't built just by a man --- man's life's too short. Science isn't even built by men of a generation --- we only further the knowledge acquired by the predecessors as far our abilities allow and our needs require us to. Hence 'tis natural that humankind's knowledge was commonly being developed by groups of people following the most prominent of us, then by the ones who followed the followers, and so till us. It's an ever evolving system with many branches and clusters. But its key feature to note is that one's first struggle always is finding an entrance. An entrance is someone who'd become their advisor, their teacher, and mentor. Someone whose line they would follow and further.
Is there no fallacy? Do ones have to find themselves what's basically a second father? Do they have to choose a single line and can't one pursue many?
It just wasn't feasible before but it may be now with the new technology. The many pieces that are the Scientific Knowledge today we could build together into a connected system available to everyone. The current state of research, the latest challenges and achievements could be tracked online in a consistent way. We could provide a guidance for newcomers in the form of manuals&tutorials, as well as lists of open issues, ranked and classified. That's what engineers and software developers are used to do. That could be the entrance. That could smoothen this steep learning curve and lower the threshold.
Yet not only could it be an entrance it could also define the workflow. We already have open review systems. We've even got services resembling social networks for scientists where you could track one's path and interact in some ways. We've got git for cooperation. We saw outstanding systems evolve with many independent contributions of many unrelated people in an almost random manner1. We know now that this randomness also provides some validation of the result and increases the robustness. Academia oughts to adopt and improve these decentralized methods.
An appeal to history doesn't seem to be a convicing argument. I believe supervisors as they are now at least can be optional though it'll always be easier to be guided by someone experienced. The immediate concern should be making research contribution-friendly and adopting the existing task-tracking and version control tools for common research. We shall keep the image of FOSS community in mind. It's only an initial stage and we shall learn our needs in progress before we can build more elaborate tools.
See the next writing in this series.
Raymond, Eric. "The cathedral and the bazaar." Knowledge, Technology & Policy 12.3 (1999): 23-49. URL: www.catb.org/esr/writings/cathedral-bazaar/ ↩