Ok, so I keep going about how I have this great idea that could improve the way scientists communicate. I also keep rambling on about what scientific communication is (or at least what I think it is), what it should be, what it will have to be. Finally, as promised, here is the overview of what my brand new system will be…once it’s actually coded up that is.
I’m probably a little stupid for writing this as it isn’t done yet, so I’m basically levelling the playing field between myself and all the people who could actually get this done but didn’t have the idea in the first place. Part of my naiveté is probably due to the fact that I believe
- nobody reads this blog (a bad assumption if I’m going to keep writing on it);
- nobody but me thinks this idea is going to work (also a bad assumption if I’m going to try and attract users);
- nobody but me thinks there’s money to be done with this (also a bad assumption as I’ve already said this is going to be ad-supported)
- somehow, having had this idea first will allow me to get it running faster (a bad idea since I’m not that good a programmer and I’m going to need someone to help me get it done)
- I’ll be able to keep the really juicy bits for myself so that I still have an edge (I’ll try).
Without further due, here it is.
Extending traditional peer-review
Currently, the current state of affairs is that before a paper is peer-reviewed it is deemed unsuitable. At least, that’s the official story. This isn’t entirely correct, as we’ve surely all referenced some internal reports that haven’t been peer-reviewed. Also, if we haven’t referenced them (as we should have) we’ve surely used some. Those of you in compressible fluid dynamics should know about the UTIAS reports, which are truly a gold mine. Absolutely not peer-reviewed, yet incredibly valuable. So why is it again we keep claiming that non-peer-reviewed material is bad? I think it’s time we extend traditional peer-review.
In this new system, everyone participates. As a scientist, you do research and then write a “paper”. (I’ll put quotes here because of the next point.) You then “publish” your “paper” by uploading it onto this system. Now this is where the differences with the traditional system start.
From this point on, your “paper” is available to everyone and anyone, including people outside of the system. (Maybe there’ll be restrictions here, so that if you wish, you can keep the report to yourself or only for your contacts, etc. Details, details.) Everyone who is registered on the system (and they should all be scientist) can “review” your paper (this could be a simple ‘good work’ comment to an extensive ‘you have miscalculated the probability of X’ type of review). You (or another user) can also go on the system and start a “journal”. If the “editor” of the “journal” wants to, he can decide to just add your “paper” to his “journal”, with or without traditional peer-review. You could also “submit” (read suggest to add) your paper to a “journal”. All the tools should be there for you, as an “editor”, to send a paper that’s been submitted to you, or that you just thought was really good and relevant to your “journal” through peer-review. You’ll be able to send a message to someone saying ‘Please review this article and send me your comments. This is for journal Y.’
In a nutshell, all “published” “papers” should be available whether they’ve gone through traditional peer-review or not. Traditionally peer-reviewed “journals” should coexist with the ones that don’t go through traditionally peer-review (or don’t care about peer-review at all).
Extending the definition of a paper
Today, in 2007, a scientific paper is still a text of variable length with static images. If you add a video or an animation of numerical results or something else which isn’t traditionally there, it becomes a different animal. Something that doesn’t really go through peer-review still today, really. I mean, seriously, some journals are claiming proudly that the paper has been enhanced on the web by including the original color figures. I think a “paper” should just be a collection of related scientific stuff. So I do this research, simulating forced coupled oscillators in a random lattice. Good. So I write it up, let’s say in a text. Because text is good. I also have these really good animations. I’ll include those too. Maybe my code is really clever, so I’ll include the source code too. So it looks something like this
On the funnyness of coupled randomly distributed forced oscillators:
- text with static figures (maybe in a pdf
- animations
- source code
All wrapped up in one. When you navigate the system, what you see is the title. Maybe you see an abstract with that. And if the author wants, maybe the abstract has animations with it. Maybe the abstract is only a video of me talking with animations from my results on top of that. So you like the intro. You click on the title. Voilà. Now you can review all the different bits. It’s all one package.
You want to download everything and look at it offline? Ok. We’ll wrap everything up in a zip file or a tar file or we’ll just download them as individual files to the place that you want.
In a nutshell, we need to go beyond text. Text and static images are a powerful way of expaining science, but it’s limited. Including moving objects expands our possibilities, but also brings about new pitfalls. Some are well-known, you need only open the powerpoint animations dialog to know what I’m talking about.
Expanding the concept of a journal
In the context of this new system, a journal is really broad. I’m just going to define a journal as a collection of articles with one or several editors, i.e. people who pick the articles that become part of the journal. The scope and selection process are open-ended questions.
So a traditional journal is still a “journal” under my definition. There’s one or more editors who pick articles on a given subject (the scope) and select them via traditional peer-review.
When you think about it, an individual scientist webpage is also a “journal” under my definition. There’s typically only one editor (you) with articles by a certain person (the scope) and the selection process is personal. If you wrote it and it’s good enough according to you (or representative or if you just wrote it) it makes it on.
A research group webpage is also a “journal” under my definition. It has several editors who pick articles on a given subject (the scope) and the selection process is not through peer-review. It’s based on how much the editors thought a particular piece of work is relevant and accurate, etc.
Care to think of other possibilities? You’ll have the choice in this new system.
Various unrelated but important tidbits
Open standards
This new system should embrace open standards (standards which everyone can use) or defacto standards (like pdf, jpg, etc). Whatever the case, we shouldn’t invent a new standard and forbid people from using it. If a new standard becomes necessary, let’s open it so that anyone who wants to use it and write applications that use that standard can. This also guarantees that if you want to take all your data because your computer crashed, you can. Also, if you decide not to use our system anymore, you have all your data readily at hand. It’s not in a useless proprietary format noboyd else uses. Interoperability is key.
Double as a professional social network, but don’t reveal all the user’s secrets
Let me give an example. I signed up for netflix. They’ve now added a social function. You can add friends to a list and share movie tastes and reviews, etc. The problem is that once you add them as a friend/contact/whatever name you want to use, you also both get access to ALL the activity on your account. I mean, all of it. I added friends of ours and the system tells me on what date they’ve returned what movie, what’s in their waiting list of movies to receive, how they’ve rated movies. EVERYTHING. This is too much. I’m uncomfortable with all that information.
So our new system should double as a publishing platform and a social network, but you should be able to limit how readily available information about you is. Information display to other users should be an opt-in not an opt-out. (I’ll discuss this in a later post still, this netflix experience has shaken me enough to talk about it again.)
Strong privacy poliy
This new system should only collect the minimum amount of information it needs about the users and it should take a strong stance towards protecting that information. It should be stated loud and clear the type of information which is gathered, how it is gathered, how long it is kept, in what form, what it is used for and what it will not be used for. Privacy and security of gathered data in social networks is important (and especially, what the limits of use of that data are) and the latest Facebook advertising campaign situation is a good example. People who used to take privacy and security of data for granted are taking a stance telling the business “We don’t like this.”
Closing word
Let me stop now. This is my brief, in a nutshell, overview of what this new system will be. It’s different from what I had envisioned when I started this blog. Writing about these different related subjects has made me think about the different issues and how to address them. Speaking with different people who have often raised valid criticism has made me review my positions and the above description is what came out.
So now I’m ready to do it. Coding time it is. Hopefully, there will be something in a rough beta stage in 3-6 months if I can convince my most trusted web developper to embark on the adventure. If I have to code it myself, we’re talking 6-12 months until there’s a usable platform, maybe more. Obviously, I’d like to have this out as soon as possible. Let’s see how it goes.
Recent Comments