218 ssp paper
- 1. What Strong Collaboration Means for Scholarly Publishing June 7, 2007 Larry Sanger When I was asked to speak to you, the Society for Scholarly Publishing, I have to admit that I found this puzzling, because I don’t know anything about scholarly publishing. Why should someone who knows so little about scholarly publishing be asked to give a speech to the Society for Scholarly Publishing? That’s a paradox. I found a similar paradox in an article by John Thompson in the Chronicle of Higher Education from 2005. Thompson wrote: “academic publishers can survive today only if they become something other than academic publishers” (June 17, 2005). The quote actually explains why I’m here. I’m here because I can tell you about a way to become something other than academic publishers. I suppose this is a little absurd, but as a philosopher, I am trained to take joy in life’s little absurdities. So I’m going to try to make the case that scholarly publishers should start expert Web 2.0 projects. Here’s my plan for the talk. • I’m going to begin by painting a picture, a vision of what information online could look like in ten or twenty years. In short, I’m going to build a castle in the air. But then I will try to put a foundation underneath it. • I’ll go over a number of examples of free encyclopedia projects from which we can learn. • Then I’ll draw out some general principles.
- 2. 2 2 • I’ll consider various business models for projects started by scholarly publishers. • Finally, I’ll give you some ideas for projects you might start. Here’s the question I want to answer first: what might the world of free vetted, reliable, edited information online look like in ten years? What sort of free resources might we see? Suppose it’s the year 2017, and we’re looking at the best-case scenario. In the best-case scenario, the Encyclopedia of Life would be an enormous success—it was recently announced, by the way, with a commitment of $100 million in grants. It has articles on the 1.8 million named and known species on Earth, with a detailed article, pictures, video where available, links to news articles, and various other resources. Basically, if you want to know about a species, you know where to go. Next, consider the Citizendium, which you can think of as Wikipedia with editors and real names. In the best-case scenario, it would have added millions of articles in hundreds of languages, but unlike Wikipedia itself, the articles have undergone a process of continual improvement, and there are now hundreds of thousands of expert- approved articles—and much other supporting information as well. In 2007, you can find some information about virtually any topic you like, on Wikipedia—but you’re not sure if you can trust it. In 2017, you can find information on those same topics, but information that you know has been checked by actual experts, on Citizendium. So much for general encyclopedias. What about other kinds of information? By 2017, the library digitization projects have gone brilliantly. The entire contents of major libraries—millions of volumes of both books and journals—have been digitized. Most 2 2
- 3. 3 3 copyrighted books still aren’t available for free viewing, except at some libraries, but they make research much easier. And it is possible for an individual to buy a subscription to services that give you full-text searching of nearly every book and periodical you could possibly want. The advantages of digitization have finally come home to archives, in 2017. It is a rare archive that has not digitized its entire stock, and made at least part of it available for free. So there are now enormous vetted and well-tagged and -organized sets of free photographs, video, and audio, which put Flickr and YouTube to shame. Meanwhile, scholarly publishers have spearheaded countless fascinating new scholarly projects, creating certain kinds of reference and academic work for the first time ever—made possible by the scale and dynamism of global collaboration. I could go on, but I suspect you’ve heard it all before. Sure, it’s exciting. But one gets tired of all the “vision.” I do, anyway. Let’s come back to Earth. I want to ask two practical questions. • First, how can humanity possibly get from here to there? • And second, what role might scholarly publishers play in getting us from here to there? It’s important, I think, to look at where you’ve been, if you want to know how to get where you want to go. So, let’s look at a series of free Internet encyclopedia projects, in roughly chronological order. • First, the Stanford Encyclopedia of Philosophy (SEP) is one of my favorite reference works. It lives online. Its articles are excellent, high-level introductions to all sorts of philosophical 3 3
- 4. 4 4 topics. They’re written and updated by experts and the whole production is edited by a veritable who’s who of contemporary philosophers. It’s a thing of beauty. The problem with SEP, however, is that it got started in 1995 and, after a dozen years, still numbers its articles in the hundreds, not the thousands. This is probably a function of the fact that there just aren’t that many top experts on the topics that SEP wants articles about. They won’t assign an article to just anybody. • Back in 2000, a peer-reviewed general encyclopedia project got started, called Nupedia. I was its editor-in-chief and organizer. Unlike SEP, Nupedia allowed anyone to volunteer to write an article, but articles still had to be assigned by an editor. We had a tiny budget and did manage to produce a few dozen articles within a year or so, and the articles were very high-quality. After a few years, particularly after I had to resign due to lack of funding, the project withered away. The problem was the same as with the Stanford Encyclopedia of Philosophy: slow pace of production. Here the problem was that not enough people wanted to go through Nupedia’s extremely rigorous, seven-step editorial process. • It was when I was trying to think of a way to improve upon Nupedia’s slow process that a friend told me about wikis— websites that allow anybody to edit any page instantly. Without even having seen a wiki, I saw that this could be the tool we were looking for, to make a simpler and more open method of 4 4
- 5. 5 5 content development. I guess many of you know what happened then. Wikipedia took off, and it never added any credible method of approving or certifying articles. In fact, without the influence of Nupedia, Wikipedia became actually contemptuous of expertise. Moreover, they didn’t require the use of real names, and they never developed any effective ways of reining in abusive behavior. As a result, while Wikipedia is an amazingly huge and useful resource, it remains of questionable reliability, and, as a community, it is off-putting to many people online who might be willing to contribute to a project like it. • Next in our catalog of projects is the Encyclopedia of Earth (EoE). This one got its start in the fall of 2005; I actually wrote some of the original project plans and policy documents for it. It’s a wiki encyclopedia devoted to everything concerning the Earth’s environment. But it differs from Wikipedia in several important ways. While articles are not assigned by editors, a byline is given to what authors happen to show up; a person has to be an expert on some aspect of environmental studies in order to contribute; and you can’t “watch the sausage being made,” that is, non- contributors can’t see page histories or the wiki-wide “recent changes” page. As a result, there isn’t much actual collabortion going on on the Encyclopedia of Earth. There is some, and they’re steadily growing, but largely because they’re aggregating content, by hand, from a number of different credible sources. 5 5
- 6. 6 6 • The Scholarpedia got its start in early 2006. I won’t say too much about this because it’s somewhat similar to SEP and EoE. It’s a specialized encyclopedia, concerning (at present) certain topics within neuroscience, mathematics, and computer science. It differs from EoE, and from Wikipedia, in that the articles are not open content; still, they are free to read. Articles are written by some really excellent experts, and reviewed by experts; as a result, however, like the philosophy articles in SEP, the articles are not really accessible to non- experts. Also, it uses a wiki, but there is very little actual collaboration going on. And as a result, there are only a few hundred articles developed, though I’m sure they’re quite excellent articles. • Next, imagine a free, specialized encyclopedia “strictly by the experts,” like SEP, EoE, and Scholarpedia. But imagine that it had $100 million to spend. Then you’d have the recently- announced Encyclopedia of Life—the encyclopedia project aiming to list 1.8 million species. It’s hard to say exactly how it will work, but their FAQ says, “Unlike conventional encyclopedias, where an editorial team sits down and writes the entries, the Encyclopedia will be developed by bringing together (‘mashing up’) content from a wide variety of sources. This material will then be authenticated by scientists, so that users will have authoritative information.” I have no critical remarks about the Encyclopedia of Life to make, because it doesn’t exist yet, and if you throw $100 million at a publishing problem, there’s a good chance you’ll solve it. It’s very exciting in any case. 6 6
- 7. 7 7 • The last example is one that started getting organized most recently: the Citizendium, a project I first announced last September, and which launched in a public beta version last March. Think of it as Wikipedia with editors and real names. As such, it occupies a unique niche. It’s a general encyclopedia, and makes full use of the wiki software and development model. Unlike several other examples given so far, articles really are developed collaboratively. But, like them, it makes a special place for experts. We call them “editors.” Our editors have two primary functions at present. First, they can review and approve articles; second, they can make decisions about questions of controversy, as necessary. But they can and do also play the role of author. A good part of our day-to-day authoring work on the wiki is done by editors. Nevertheless, we also invite contributions from the general public, who work as authors, shoulder-to-shoulder with the editors. You might think this would be a recipe for expert/amateur conflict, but so far we’ve seen little of that. The wiki has been under development for about seven months. In that time, we have added o about 1700 authors o about 240 editors o about 2000 articles If we continue to grow—I mean, to increase our rate of growth —as we have been, we should have hundreds of thousands of articles within a few years. We have a similar amount of content to what Wikipedia had after seven months—fewer 7 7
- 8. 8 8 actual articles, but our articles are longer, on average. We’ll also be expanding the number of approved articles we have, which right now is just over 20. What, then, can we learn from these projects? Before I draw a few lessons, I want to make a few stipulations. I’m going to stipulate, first, that an encyclopedia is better the larger it is and the more reliable it is: both quantity and quality. Second, I also want to stipulate that the community that creates the encyclopedia is also important; and the community is better if it is not constantly engaged in acrimonious controversy. Few of the encyclopedia projects we reviewed have grown very rapidly. Wikipedia and Citizendium have done pretty well so far, on that score. But the other projects grew slowly for various different reasons. The Stanford Encyclopedia of Philosophy and Scholarpedia have grown slowly, I think, mainly due to the fact that they are so exclusive. They insist that article authors be not merely competent and knowledgeable scholars, but actually distinguished in their fields. If they were to expand the set of possible contributors, they would of course have more contributors. My view is that it is only genuine good old-fashioned elitism that can justify the exclusion of competent scholars. I think this explains in part, by the way, why the Citizendium has done rather better, in terms of numbers of articles created, than the other expert-driven projects. So I advance this principle: expand your base of contributors as widely as is reasonable. Nupedia grew too slowly primarily due to a complex workflow. I’ve discovered time and time again that, when presented with a problem, 8 8
- 9. 9 9 scholars tend to want to create a new process, a new workflow, a new committee, to deal with it. This might solve the problem, but it also slows down production. The lesson here is: radically simplify your workflow. A wiki is an example, but only one example, of a tool that encourages a simple workflow. Next, consider those relatively recent projects that use wikis, but which don’t really operate as wikis—here I think especially of the Encyclopedia of Earth and Scholarpedia. They could, I think, grow much more quickly, if they really were collaborative—that is, if they had a relatively energized collaborative community. Well, how do you energize a collaborative community? Several ways. • First, don’t sign articles. Leave them unsigned. The reason is that, if you do sign articles, a few things happen that make collaboration difficult and unlikely. The author will discourage and resent input from others. And others will avoid collaborating on articles, because they don’t want to offend the author. • Second, for the same reason, actively discourage the idea of exclusive personal control over articles. Even if the project’s articles aren’t signed, some people will act as if an article he started really is his own, and discourage others from contributing. So you have to actually tell those people, “You may have written a draft of this article, but it isn’t yours.” • Third, positively encourage people to edit everyone else’s submissions. There are various ways to do this. You can put it in policy and help documents. You can have your most active and distinguished authors ask for help with articles they’ve 9 9
- 10. 10 10 started. Editors can ask one author to help another author. And so forth. Eventually, I think people will get the idea. • Fourth and finally, something that the Encyclopedia of Earth does, but which Scholarpedia does not do, is to use an open content license. This also helps to build a more dynamic community, because such a license is a guarantee to contributors that their collective work will always be free; it won’t disappear when the managing organization disappears. I think you might be surprised at how important this is to some people. It would be a huge mistake to think that experts and scholars are unable to collaborate, Wikipedia-style. The Citizendium project has demonstrated that the articles that result from such collaborations can be truly wonderful. Next, there is the problem of the lack of reliability, which I think is a problem mainly with Wikipedia. Here, the solution I recommend has made me a heretic in the Web 2.0 world, but it’s the obvious old- fashioned one for everyone else: if you want to be sure that some content is reliable, then you get experts to review your content. So, find a place for experts. The last problem I see is another problem of Wikipedia’s. Namely, it has an off-putting community, due to its immaturity and failure to enforce its own rules. Here, again, I recommend some heretical yet old-fashioned solutions: have contributors identify themselves with their real names, not pseudonyms, and, also, empower and require moderators to enforce rules consistently. So here is a run-down of the lessons learned from the free encyclopedia projects listed above: 1. Expand your base of contributors as widely as is reasonable. 2. Radically simplify your workflow. 10 10
- 11. 11 11 3. Don’t sign articles. 4. Moreover, actively discourage the idea of personal ownership of articles. 5. Positively encourage people to edit everyone else’s submissions. 6. Use an open content license. 7. Have contributors identify themselves with their real names, not pseudonyms. 8. Require moderators to enforce rules consistently. The only project that actually follows all of these principles is the Citizendium. I guess that’s not surprising since I’m the editor of the Citizendium and I wrote these principles. But I do practice what I preach, in this case. So, I know this is going to sound terribly immodest, but I guess what I’m recommending is that you start projects like the Citizendium. I can’t expect you to take this recommendation very seriously, partly because the Citizendium is free, and you, as publishers, are in business to make money. But, you know, people do make money by publishing free stuff online. I don’t, but other people do. Personally—and I know this must sound bizarre—but I’m really not in it for the money myself. The Citizendium is a non-profit, and I don’t expect to get rich, at least, not off of this. Still, other people do get rich by publishing free stuff online. Just think of the founders of Google, Yahoo, YouTube, and MySpace. Their business model, of course, is advertising. As you must know, online advertising is increasingly lucrative. As to the ethics of the thing, newspapers have been supported for years by advertising, and only radicals have complained about the ethics of their advertising. So, in principle, I personally don’t have a problem about supporting a project with advertising. 11 11
- 12. 12 12 Still, you might wonder why Wikipedia, Citizendium, and virtually all other of the encyclopedia projects I listed—except for Scholarpedia— don’t use advertising. I think the main reason is that their organizers and/or contributors hold the view that advertising equals corporate bias and corporate control. I don’t personally hold this view, but I respect it, and if it means I can’t have as many contributors, I will not run advertisements. Another business model, one that seems particularly viable in the world of scholarly publishing, is the “pay-to-play” model. The idea here is that if a university department wants to participate in some scholarly project to produce free information, organized by a publisher, then the department pays the publisher, and then the faculty and grad students can participate. This basically is the “open access” model, expropriated from journal publishing, and applied to collaborative content production. Another business model involves selling “premium content” to subscribers—for scholarly publishers, this again is not a stretch. I assume I don’t need to elaborate on this one, because it’s something you already do as a matter of course. Finally, a business model that is worth a try, though few people have actually tried it, is a sort of patronage program. The idea is that, as a publisher, you are hooked in to a large network of scholars. Suppose you were to invite people to pay for free content, created by your network? In other words, you, as publishers, solicit donations—from individuals and from institutions—and the donors can specify a few details about what they want. For example, suppose they want to support the creation of an anthology of important popular writings about global warming. Then they approach you with the money, and if it seems to be enough for the job, then you tap into your network, wrangle the content, and publish a collection of essays. The essays are free online; the funders get credit as patrons. 12 12
- 13. 13 13 I don’t know whether this is a viable model, but it seems like an interesting way to pay for free, expert-produced content. I think it’s worth a try. Eventually, perhaps very soon, the Citizendium will give it a try, by the way. I’ve written an essay developing the idea. It’s linked from larrysanger.org, and it’s called “The Role of Content Brokers in the Era of Free Content.” Finally, I promised you some ideas for projects. There are countless interestingly different ideas for expert-led collaborations; I’ll give you just two. So, you’re an academic publishing house. Let me assume that you think the collaboration train is leaving the station, and you want on. What can you do to develop your firm’s expertise in this area, and explore new business models? What sort of project should you start? Here an idea: a literature review. But not just any old literature review. A really thorough, comprehensive, and balanced review of every part of the literature of a field. In other words, a comprehensive account of the latest advances in the last, say, year. It is very labor-intensive for just one person to create this sort of overview of the literature even about some very narrow topic within a given field. But, of course, experts together know the literature of their fields far better than any one of them knows it individually. Furthermore, if they are writing summaries of the latest research, their summaries will be far better if they can correct each others’ mistakes. And bear in mind, also, that anyone who is a real expert in a field has to keep up with the latest advances—they all have to go over a lot of the same material, so they’re all doing the same thing. So a collaboratively-written, comprehensive summary of the literature of a field makes sense. 13 13
- 14. 14 14 I would absolutely love to see such reviews of the literature about my own interests of philosophy, the Internet, and Irish traditional music. I think it is only a matter of time before people start very regularly producing literature reviews collaboratively. It could be done using a wiki, and so wouldn’t be technically difficult to set up. And, perhaps for obvious reasons, scholarly publishers are the perfect organizations to spearhead (or help spearhead) such an effort. For the next idea, let me give a little background. There are many different ideas about collaboratively creating supplementary research material related to classic, public domain texts. Just think of the conjunction of three facts. Fact one: most of the important classic texts in all fields have already been digitized—and digitizing new versions of classic texts has been automated and made cost-effective. Fact two: access to these texts—and to a single version of a text— can be virtually universal among scholars, since they’re virtually all on the Internet. Fact three: in recent years, scholars in large numbers have finally begun to “get” the idea of strong, wiki-style collaboration. What follows from these three facts is that it is possible for scholars to work together in huge numbers on supplementary study materials for classic, public domain texts. In itself, digital study aids for classic and public domain texts are nothing new. The wonderful Perseus Project is one prominent example. But the Perseus Project isn’t really very strongly collaborative. So here’s an idea that I have developed on the textop.org domain. 14 14
- 15. 15 15 Imagine a scholar working through a digital copy of Thomas Hobbes' Leviathan, dividing it into chunks of approximately one paragraph in length. Imagine the scholar labelling the chunks by function, summarizing them, and placing them into a single outline, reordered, beginning with the most abstract topics, like Metaphysics, and working down to applied topics, such as Political Philosophy and Philosophy of Law. Imagine the scholar repeating this process, summarizing and as it were filing away paragraphs of text for, say, the fifty most influential works of philosophy, with all the chunks of texts summarized and collated into the same outline. This would require a scholar on the order of 5-10 years, depending on diligence. But if many scholars were to work on the project together, the amount of time required would be a fraction. Next imagine doing this sort of text collation for other fields beyond philosophy, and texts in languages other than English. The result would be an unprecedented, highly interesting, and fantastically useful resource. But, obviously, this would require a massive collaborative effort. It seems to me that this is the sort of collaborative project that scholarly publishers could take on. Now, if you want to use this particular idea, I want in, because I absolutely love this idea, and eventually I plan to do it myself. I’ve actually prototyped it. I hope you have found this useful. I’ve covered a lot of topics very quickly and roughly. I briefly painted a picture of what scholarly collaboration might produce by the year 2017. Then I reported on a series of free online encyclopedia projects, and from them I drew a series of principles about how to organize a successful collaborative project. I offered a few business model ideas that a publisher might use to fund collaborative projects. And, finally, I gave you a couple of what I think are fairly intriguing ideas about what scholars can do together. 15 15
- 16. 16 16 Maybe they are just waiting for you to bring them together to collaborate online. 16 16