Why Should Publishers Organize Science?

I found this post titled "What scientists really want from digital publishing", and, after reading it, I'm pretty certain this scientist doesn't want what's offered. Before I get to the details, here's what computer scientst Philip Bourne offers:

"as a scientist I want an interaction with a publisher that does not begin when the scientific process ends but begins at the beginning of the scientific process itself"

I actually want to do away with publishers--I see them as a necessary evil, including the Open Access ones. What I really want is to be able to communicate to other scientists (and share data) without an intermediary. But I'm getting ahead of myself. I do agree with this description of the problem:

He thinks scientists need help with management of data in general, and specifically:

  • · Project management. They use e.g. basecamp for project management but email folders are primary - this in an unhealthy 'hub and spoke' situation
  • Content management. It's a mess with content stored all over such as on slides, posters, lab notebooks etc.
  • Manage negative data. They generate way more negative data than positive - Negative data is important. But you can't find it - it stays hidden. This needs to change.
  • Software. All the software they create is open source but when the grad student that wrote it leaves, it's lost.

I don't disagree with any of that. These are problems. In fact, where I work, we have people who specifically deal with these sorts of issues (as I've mentioned before, genomics is much, much more than putting prepared DNA in a magic box). But the solution is daft:

The current situation is:

* 1. Ideas
* 2. Experiments
* 3. Data gathering
* 4. Conclusions - it's at this stage that the publisher comes in

But why couldn't the publisher come in at the data stage? They could help store it for our group. Or even earlier, at the ideation stage: The moment I jot down a few ideas, the publisher could control access to that information and then at some point down the line when the access is opened up - that's when it becomes 'published'.

It's hard enough to work with publishers when you have a reasonably finished product. The idea of inviting Reed Elsevier to be intimately involved with the entire process is ludicrous:

Parenting Fail
see more pwn and owned pictures

There is a larger problem here: as more and more disciplines of science become data-rich and informatics-based, scientific success will require infrastructure to successfully manage and integrate all of these processes. But publishers shouldn't be the lynchpin of those efforts: scientists should be. At Major Genome Center, internally (and with collaborators), we use wikis to organize and communicate (and some wiki architectures are better than others). Externally, scientists should built out sites that display the data and offer the interactive features Bourne wants (here's an example, but obviously much more could be done). Publications and other presentations can be added to such sites to, and offer interactive features--in fact, if done well, publications could be a stripped down version of the site.

So we do need to rethink project and data management, communication networks, and the like. But I'll be damned if I let a publisher be involved from the outset.

More like this

Nature has been busy diving into the "web 2.0". Now it looks as if the folks at Natiure have two new projects for the science masses: - Nature Precedings, a website where people can dump all their spare data, unpublished manuscripts, powerpoint slides, posters, images and where readers can leave…
This is the third in a series of test essays I'm doing to prepare for my comprehensive exams.  The questions for these essays come from 3 places: ones I've made up based on my readings, ones assigned to previous doctoral students, and ones my advisor makes up based on my readings.  I'm assuming the…
During my summer blogging break, I thought I'd repost of few of my "greatest hits" from my old blog, just so you all wouldn't miss me so much. This one is from September 3, 2008. There was some nice discussion on Friendfeed that's worth checking out. ===== Some recent posts that got me thinking…
Since early days indeed, it's been possible to bypass journal publishers and libraries in a quest for a particular article by going directly to the author. Some publishers have even facilitated this limited variety of samizdat by offering authors a few ready-made offprints. I've even had…

The data management problem is serious, but figuring out to organize the data is often a big chunk of the science. Mendeleev's periodic table of elements was serious science, and the kind of thing a publisher could do. Biology is swamped with data. They had to develop the entire field of bioinformatics just to think about the problem.

On the other hand, if someone is proposing some kind of safe locker / lab notebook approach where the system actually helps scientists get their work done I'm all for it. Again, this isn't the kind of thing publishers should be involved with.

I agree, data management is science discipline for itself. Take look at today's genetics, biology, economy, medicine... All of them are data-richer and richer each day and data management involved in those disciplines are more and more complex - science for itself.

I completely agree -- we are already far too dependent on publishers to act as both our pimps (putting our results out there) and drug dealers (getting rewarded with citation metrics).

By Geraldine (not verified) on 13 Apr 2011 #permalink