So I went to this talk on Thursday about open science by Michael Nielsen. He quit studying quantum computing, in which he was apparently doing quite well, so he could “work on open science” so he must know what he’s talking about. It was one of those talks where I was constantly thinking of problems with what the speaker was saying and then congratulating myself on how clever I was. Then most of my comments got answered throughout the talk and others were answered during the questions. I did raise my hand a few times, but never with enough conviction to get called on… No matter, that’s what blogs are for, right? Anyway, these are my thoughts, they are likely neither novel nor comprehensive.
Two years ago, I wrote an article for our departmental magazine called something vague like “Science and the Internet,” in which I talked about twitter, RSS feeds, Tim Gower’s Polymath Project and some other things. Needless to say, I felt pretty gratified when this guy used the same example (Tim Gower, that is) as the crux of his ‘open science is better science’ conjecture. He also mentioned Terence Tao’s blog, and his own personal use of sites like delicious and friendfeed. Then at some point, I realized that these examples are really more like the modern equivalent of when you end up rabidly discussing your work after a few pints in the pub or a few cups of coffee in the cafe. In other words, so far the internet is used as a very informal way of communicating science (friendfeed, really?), which while invaluable no doubt, fits more into this category:
Really, for open science to be effective it needs to be nudged over to the work-work side of this Venn Diagram. Is this possible? I’m going to address this via another talk-fueled realization: it makes the most sense to separate ‘open science’ into two categories: (1) open data and (2) open ideas.
(1) The first one, open data, is something that, according to Nielsen, is actually being pushed for among funding bodies. But it seems odd to me that funding bodies can just say “oh, if we give you this money, you need to share your data.” Okay, cool, but HOW are we supposed to share our data? For most fields this infrastructure is not in place. To do this properly would require a very large initiative on the part of governments, funding bodies, and/or libraries to set up a stable and reliable data storage infrastructure, that currently hardly exists on a local (lab/department) scale let alone a global scale.
(2) As hard as that sounds, the second category, open ideas, is probably the more controversial and more difficult to implement. The Tim Gower and Terence Tao examples are of this sort. To implement this kind of experience, using many minds scattered throughout the world to solve one problem, on a large scale would require what many have tried to create and failed: a facebook for scientists. But do we even want that? Do I want my entire scientific identity to be based on a profile page? I suppose my social identity kind of already is, but somehow that’s not the same. What really bothers me about this, though, is that the people who talk about open science a lot are the scientists who already have a blog, a twitter, an internet know-how that is above average. Even most younger scientists, adept at social media for their own affairs, rarely interact with an online scientific network and wouldn’t know where to start. With this in mind, it will be difficult to develop a model of ‘open science’ that does not automatically alienate a large portion of the scientific population just due to lack of social media literacy.
For both of these branches scooping is a worry, yes, but the reason scooping is so feared is because jobs depend on publications, which really this whole ‘open science’ thing started because people are getting angry at how backwards the scientific publishing industry is. Another interesting Nielsen mentioned in his talk was Henry Oldenburg’s unorthodox methods of encouraging scientists to publish in first peer-reviewed journal, Philosophical Transactions of the Royal Society, established in the 1600′s. Apparently Oldenburg had to trick scientists into thinking their competitors were ahead of them, so that they would be willing to publish what they had so far, and thus the whole community could benefit, despite the scientists maybe being a little pissed at Oldenburg afterward. Now somehow the original usefulness of journals has been corrupted, and in fact they have been attacked as doing exactly the opposite of what they were created for in the first place.
In the days when the first scientific journals were created, science was different, there were no biotechs, no multi-billion dollar government health agencies, and no modern patent laws. Galileo requested patronage directly from the Medicis by offering to name Jupiter’s moons after them (also a fact from Nielsen). Today we still have to appeal to the wealthy for funding, but the system within which we have to appeal is very different from those first days of peer-reviewed science when scientists were in one way or another self-funded. Which was possible then for a variety of reasons that are just plain not the norm now. Maybe, though… Maybe we should try to revert back to that golden era when the first journal was created. If we moved outside of the current system and all suddenly reverted to the old days of Galileo! When scientists were an elite class and nobody published at all except by cryptic letters to their friends (which we could replace with twitter!) and then appealed directly to the aristocracy for funding! Then, with a new system of gonzo science where each of us is our own J. Craig Venter and nobody is beholden to any of these crazy journal things, we could all be free to do the science we should have been doing all along!!!
Anyway, the point I’m trying to make with my gratuitous use of exclamation points, is that nobody is really talking about the same thing when they talk about ‘open science’, and the variety of problems we want it to fix is kind of unreasonable. Open science is not going to stop our careers from being decided by departmental politics, it’s not going to prevent us from having to get actually interesting results, and it’s not going to get rid of reviewers with ridiculous demands. But most importantly this thing that we dream of and can taste so vividly is not just going to happen overnight. Sure, public opinion will help governments/foundations/etc realize ‘open science’ is something worth investing in. But talking about how awesome it’s going to be isn’t going to get it done, whatever ‘it’ may be.