In his trio of must-read essays on open science in 2006 (parts 1, 2, 3), cancer researcher Bill Hooker identifies open notebook science as his “personal favorite” expression of a broader movement that includes everything from publishing open access journals to developing common standards for data sharing. The reason? He believes it can help cultivate new cultural norms of openness, transparency and immediacy in the daily life of scientists.
So what exactly is open notebook science? Following in the tradition of preprint servers like arXiv.org but digging even more deeply into the roots of the research process, open notebook science is an effort to document and share science-as-it-happens: exploratory ideas, raw data, lab protocols and more. As the enthusiastic response to author Michael Nielson‘s recent post on sharing more “useful” knowledge shows, the practice is generating an enormous amount of excitement, as well as sparking questions about what’s next. How do we ensure that what we post online is truly “useful”? Can we publish in a way that optimally leverages the network and contributes more signal than noise?
To explore these questions and others for our series of posts on the future of science, I talked with Lorrie LeJeune of the OpenWetWare project. Launched in 2005 by members of the Endy and Knight labs here at MIT (and originally dubbed “Endypedia”), OpenWetWare provides resources to make research easier and encourage researchers to take a more open approach to science. Over the past three years, its wiki-powered site has become a hub for open research, and now serves a community of 4,100 users from research labs all over the world. LeJeune, whose remarkably diverse background includes work both as a molecular biologist and a marketing manager and editor at O’Reilly, joined the project in February as its new Managing Director.
First, can you tell me how you ended up at OpenWetWare? What drew you to the project?
The short answer is because I’m interested in science, open source, disruptive technology and publishing, and OpenWetWare allows me to explore all four areas simultaneously.
The longer answer is that I’ve spent my career following my interests. I have an undergraduate degree in Animal and Veterinary Sciences (specialty in dairy science and microbiology). I worked in biotech in the 1980s doing protein chemistry and yeast genetics. I spent the next fifteen years in publishing, working on digital publishing technologies at the MIT and University of Michigan presses, and as a marketing manager and editor atO’Reilly Media. After a four-year detour through the world of mobile telecommunications at Orange, I learned about OWW. And here I am.
OpenWetWare is generating a lot of interest in open research and especially in replicating the model across many different disciplines in science. Of course, in many cases, there are factors that discourage sharing pre-publication information. Would you say part of what OWW does is help demonstrate what can be shared, as well as the benefits of “baking in” sharing?
Yes, exactly. Many researchers are not opposed to sharing their work, and they do so regularly in the privacy of their labs and their departments. They generally draw the line at sharing with the world until the article is published, and they can get the attribution for the work. Attribution is the key to being successful in academia: getting tenure, getting the grant money, getting the grad students, etc. Most scientists have been taught that to talk about one’s work prematurely opens the door to “scooping,” or having your ideas stolen by another researcher, who will do the work and publish the results before you do and get what should have been your credit.
How does OWW combat that perception?
We make the argument that if you share your information publicly, it is in fact attributed to you, and if someone at some point in the future tries to present your work as theirs, you can point to a trail of content you posted. Granted, there are complexities–especially in the world of wikis where people can edit other people’s stuff. But if you post it, there’s a record that can be traced.
To paraphrase what we state on the OWW wiki, some users of OpenWetWare think that the best thing to happen would be if somebody “stole” your idea and finished the work before you. Then you could go work on another idea. Good researchers usually have more ideas than they have time to explore, and having more people exploring those ideas will in the end benefit your research. OpenWetWare’s mission is to make research better and lower the barriers to sharing. If we can facilitate exploration of more good ideas so that everybody is doing the best work possible, as quickly as possible, everyone wins.
Another reason to be open–especially if you’re working in a nascent field of study–is to bring attention to your work. One fact of life in “open science” discussions is that there are going to be people who think it doesn’t make sense in their (“very competitive”) field to be open about their work. However, there are many, many scientists whose principle problem isn’t being scooped–it’s that no one notices their work. This is especially true among younger scientists still making a name for themselves or folks in smaller fields.
I could argue that there is already significant incentive for young scientists to publicize what they are doing as openly and early as possible. This open group will either be scooped out of existence, or will be more successful thanks to all the unintended benefits of making its work accessible early. We really won’t know which it is, however, until we run the experiment.
All that said, we’ve had a lot of discussions here at OWW about sharing systems that take advantage of the collaborative power of wikis, but also bring in time-stamping, DOIs and other proofing systems that traditional publishing uses. I think the bottom line is if you can be credited with the idea or the experiment or the conclusion, you’d be willing to share it with more people, if not the world. And there are important advantages to sharing early-stage research more widely.
Once a member of your community has decided to publish information, does OWW encourage people to mark it with attribution and usage rights?
We’re still working out the best way to deal with that. The OWW site itself is under Creative Commons and GNU FDL licenses, but we are an early-stage project. Right now, technically speaking, we’re a wiki–with all the power and frailty that wikis have. We are looking at ways to incorporate an official form of time-stamping, using an outside validation body–sort of an online notarization process. And as I said, we are also looking at how to implement DOIs for documents.
Do you plan to explore a “best practices”-type approach–where you offer your community policy guidelines for publishing, as well as dealing with the legal issues surrounding sharing?
Yes, we will have to explore policy guidelines, especially to reassure people that they will be credited with what they publish. The legal issues also have the potential to loom large. Some industry folks may not use OWW because they’re concerned about patenting. On the other hand, many people don’t think about licensing or copyright at all before they publish. If we want to allow sharing of data, we need to make sure we’re using and communicating the right intellectual property and licensing information. Fortunately, our friends at Science Commons are willing to help us in that department!
In terms of developing guidelines for the community, I think we’ve gotten to the point of opening the door, but most researchers are still milling around at the threshold taking an occasional glance out. Not many people have actually stepped through. We (meaning the open research community) need to figure out the rules of engagement. Once we have even a few guidelines for sharing, I think the rate of openness in science will accelerate.
What can we do to spur more revolutions in the way we practice science? Or to put it another way, what can we learn from OWW? Did the project come about in part because the people behind it are relatively young and acclimated to sharing? Or is it that they practice in an area of science well-suited to sharing–a younger science?
Why is OWW happening? Two factors come to mind:
1.) Because it can. Tools for information sharing and social networking have evolved sufficiently to allow it to happen.
2.) The generation of graduate student researchers launching their professional lives now may be the first to have been raised “on the Web,” using Web-based tools. They’re very comfortable sharing their ideas and information in a public, online environment.
I think this is essentially true because it seems to be the more established researchers who worry more about sharing information on the Web. Science has always had a secretive side, and probably will continue to until the funding and tenure systems change. So some of the hesitancy could stem from having cut their scientific teeth in an era where having conversations with colleagues, giving presentations at conferences and publishing in print journals were the only “safe” ways to share ideas. In addition, many senior researchers simply don’t have the time to fool around with learning to edit wiki content. Graduate students have the time, the early training and the motivation to use whatever resources they can find to get their work done.
I agree that sharing is likely easier in some branches of science than others. In the newest branches, like synthetic biology, everyone is on the fringe. More people are struggling just to get their work noticed. And newer fields are smaller. People are more likely to know one another, which in turn makes sharing easier. If you know someone, even slightly, you’re much more inclined to enter into a collaboration with them, so you can move both your own and their work ahead faster.
Of course, even wiki- and Facebook-savvy scientists don’t share everything. They seem more willing to put protocols and other “safe” materials out in front of the public eye than they are unpublished research data.
So I think the upshot is that people are very interested in sharing, but exactly what and how much is shared is still very variable.
What are the projects at OWW that you’re most excited by?
There are lots of examples, like our new one-click open lab notebook, but I’m especially excited about a project Julius Lucks is pursuing. He’s a post-doc at Berkeley who’s interested in publishing, and has written and published on OWW a paper on using Python for science. When I asked Julius if he’d be willing to write a full-length “Scientific Programming with Python” book on OWW as an experiment in open-access writing, he jumped at the chance. He now has an outline and two co-authors, and we’ve launched the OWW Open Writing Initiative. Wikis are designed for collaborative writing and editing, and I look forward to learning how the process will work on OWW.
What’s next for OWW?
We’re exploring ways that OWW can help make research easier, faster and more open. To that end, we’re looking very closely at where the problems and opportunities are within the traditional cycle of research. What tools and services can OWW develop or foment to help researchers move, for example, more swiftly from literature search to experimentation at the bench? Or from collecting data to writing a paper about what that data signifies? We also hope to grease the research pathways by providing tutorials on tagging data so it’s easier to search, or on choosing a research project, or when to stop working on a project. By using Web-based resources to help research move faster and more smoothly, we believe that it can’t help but become more open.
Previous posts in this series: