Reinventing Discovery - Michael Nielsen [125]
p 175: My account of the origins of open sharing of discoveries in science is based in part on Paul David’s article [49]. David points out that there is nothing logically inevitable about the emergence of openness in science, and that it was in large part a result of external forces acting on the scientific community, not merely forces within science. David’s analysis focuses on the earliest parts of modern science, and emphasizes how prestige seeking by monarchs and other patrons was a motivation for open disclosure of results. In my account I’ve also emphasized the motivation coming from the public benefit derived from open science. This motivation seems to have acquired more force in later times, as the power of the monarchs diminished.
p 176: The qwiki is online at http://qwiki.stanford.edu/wiki/Main_Page. In my description of the qwiki, I state that only a few pages are regularly updated. In fact, there is a part of the site that receives fairly regular attention: the “Complexity Zoo,” a resource for computer scientists that describes different types of computational problem. The Complexity Zoo needs separate consideration, however, for it is based on a project that was originally totally unconnected to the qwiki, and that later merged with the qwiki. As a result, for the purposes of this discussion, I’m treating the “Complexity Zoo” as a separate entity. It is, of course, interesting to ask why the Complexity Zoo succeeded when the rest of the qwiki failed. A full answer to this question is complex, but in brief, the Complexity Zoo has a much narrower scope than the qwiki, and because of this narrower scope a single dedicated person (Scott Aaronson, now of MIT) was able to build it out to the point where it became an extremely useful and well-known resource in the computer science community. The combination of its already high profile and its narrow scope has helped attract a few people to make occasional contributions to its upkeep.
p 176: The term “wiki-science” seems to have been introduced in an essay by Kevin Kelly [108]. Similar ideas were proposed independently (and, in some cases, earlier) by many people. An intntere discussion involving some early contributors to wikis may be found at the Meatball wiki: [137] and [138].
p 178: The job and graduation data for physics are based on the American Institute of Physics’ “Latest Employment Data for Physicists and Related Scientists,” available at http://www.aip.org/statistics/. I picked physics because reliable data are available. Anecdotal impressions from other fields confirm that the situation is similar.
p 178 Those science wikis that do succeed are usually in a supporting role for some more conventional project: a notable exception to this rule is the Gene Wiki, a successful wiki-based project to annotate genes. Part of what has helped the Gene Wiki succeed is that it is not an independent wiki, but rather a subproject of Wikipedia: if you’ve ever looked up a gene on Wikipedia then chances are that you’ve seen work done as part of the Gene Wiki project. The Gene Wiki benefits from the many people who already dedicate time to editing and improving Wikipedia, and from the high visibility Wikipedia pages often have in search engines.
p 179: For another perspective on user-contributed comment sites for science, see [148].
p 179: The final report on Nature’s trial of open peer review: [167].
p 180: Although the user-contributed comment sites for science are failing, scientists aren’t always unwilling to comment online about other scientists’ work. We saw an example along these lines starting on page 259, with science bloggers investigating the evidence for chiropractic offered by the British Chiropractic Association in their dispute with Simon Singh. Other examples include (1) a Polymath-style collaboration [173] in 2010, in which a group of mathematicians, computer scientists, and physicists worked together online to analyze a claimed solution to one of the biggest open problems in computer science; (2) a blog-based online discussion