User:Linas/Original research, peer review and reputation on Wikipedia
The creation of a new, on-line, peer-reviewed journal is proposed. The journal would be novel in that it would combine successful aspects of various online technologies and social mechanisms. In particular:
- Preprint servers, such as ArXiv, have been an enormously successful mechanism for distributing pre-prints.
- Blogs have attracted readership not only because a given blogger is interesting, but also because the readers can comment, discuss and "feed back" about what they've read.
- Wikipedia has proven to be an enormously successful venue for the authoring and publication of encyclopedic information, and has evolved a number of interesting and sophisticated social mechanisms that help maintain article quality.
- Peer review by circles of editors with the highest reputations enables the publication of journals with impeccable reputations, such as Nature or Science, and there is no reason that the concept of reputation cannot be transfered online.
An appropriate combination of the best features from each of these technologies could potentially result in a new way of presenting human knowledge. In particular, the entire spectrum of content quality, from rough drafts and diaries of original research, to finished articles undergoing publication, to the creation of encyclopedic review articles, can be accommodated.
There are two major challenges to creating a new scientific, peer-reviewed journal. First is the creation of the social network of reputable editors who are the life-blood of the publication. Second is the creation of a "name-brand" with a wide readership. This is critical for attracting distinguished authors and new editors. Insofar as Wikipedia is already engaged in the creation of review articles, and already has a name-brand, this proposal suggests that the new journal be run as an adjunct of Wikipedia, as that will be a critical mechanism for attracting and retaining the interest of top-quality authors and editors.
The road to publication
The introduction summarized the core ideas. Rather than jumping to an abstract discussion of these ideas, let us first review, in concrete terms, how the actual publication process might work. A core assumption in the following is that publication will occur within the framework of the current Wikipedia. This assumption is made not only because WP posseses interesting technological and social mechanisms, but also for business reasons: Wikipedia is the name-brand to be leveraged in the creation of the journal. However, this discussion also assumes some alterations of (longstanding) WP policy, and so is "non-trivial".
We start by not only allowing, but encouraging the creation of original research on WP user pages. Thus, for example, User:Linas/Wacky thoughts is an example of "original research". It may not be very good research, and in fact may be little more than a collection of notes, or a maybe a diary. In a sense, it could be construed to be a "blog". Based on the reputation of the author (and maybe with the help of advertising), it may attract readers, and those readers might comment on the talk page.
To respect author priveledges and rights, it would be reasonable to expect that these user pages are to be edited only by the original author. That author may invite friends to help edit as well, but editing by anyone else constitutes vandalism or worse. In particular, uninvited edits by powerful Wikipedians, such as Jimbo Wales, might be construed to be vandalism. If someone wants to be a co-author, they would approach the author in the usual way: e.g. via talk pages. To help enforce these rights, it might be reasonable to have an access control list mechanism to physically bar interlopers from editing these pages.
To maintain a collegial, collaborative atmosphere, I thus invite my trusted circle of WP friends and acquantainces to co-author and edit this page. Specifically, I invite User:Oleg Alexandrov, User:Jitse Niesen, User:Lethe, User:Charles Matthews, User:Chalst, User:Dmharvey, User:Silly rabbit, and User:Hillman to edit this page. I may have forgetten to list someone I hold in high regard, and thus, this would be socially embarrasing for me. Oops. However, by inviting so many "friends", some of whom I do not know so well, I illustrate that this is not only a collaborative process, but also a partially political process.
Currently, user pages in WP are covered by the GFDL. In the modified system, it may be worth reviewing whether this is still a desired aspect. In particular, true original research in the biological sciences, computer sciences and engineering may lead to actual patent applications. Thus, the appropriate policy to be applied to user pages would need to be carefully stated and reviewed.
If an author desires to publish an article, the next stage would be the "preprint stage". Once an article is sufficiently well developed, the author would request a move of the article to a new namespace, for example, to [[Preprint:John Smith/My Theory]]. This move would be under the contol of a minimal peer-review mechanism: namely, the move would not be made until it is explicitly approved by two or three editors who have such approval authority. The actual criteria for acceptance would be minimal: the goal here is to keep out link-spam and not protoscience. In principle, if a "crackpot author" got together with two or three "crackpot friends" who had approval authority, then "crackpot articles" could enter the preprint stream. And at this stage, there is nothing wrong with that, per se. As long as there are some minimal standards, its hard to distinguish between protoscience and pseudoscience, and this is not the correct stage to do it.
There will be user pages that will never become preprints, and this is fine. The idea of a preprint is that the article is "almost done being baked": it contains important ideas, it is a structured presentation of those ideas, and it is intended for mass readership. By contrast, many or most user pages might not be for "mass consumption"; by contrast, the preprint is an inherent declaration that "I want and am ready for a broader readership".
Thus, for example, this page, that is, [[User:Linas/Original research, peer review and reputation on Wikipedia]] is (almost) ready for the preprint stage. In this context, I will ask that it be copied to the namespace Wikipedia:Original research, peer review and reputation on Wikipedia, as this is vaguely analogous to the preprint namespace.
Some preprints may stop at this stage. Others will move on to the next stage, which is publication.
Under construction -- what do we mean by "publish"?
- A journal is an editorial board. The respect for a journal derives entirely from the quality of the work done by the editorial board: did they allw crap to get published, or did they uphold the very higest standards of peer review and publication?
- Thus, within WP, we want to establish "editorial boards": highly exclusive collections of editors who place thier mark or seal of approval on certain given (versions of) articles. Since personal reputations and, potentially, promotions and salary are involved, all the usual modes of conduct of of an editorial board come into play, including politics.
- A "WP editorial board" is synonymous to a "journal". Here, "publication in a journal" is synonymus wih getting an editorial board to place its seal of approval on an article. The quality of an article is judged by the journal it appeared in, and, conversly, the journal is judged by the qality of the articles it publishes.
- Publication by multiple "WP journals" is allowed.
- Compare and contrast meta:Article validation and meta:Article validation proposals (which lists 32 distinct proposals!!!) and Wikipedia:Stable versions, Wikipedia:Good articles and Category:Editorial validation and the WP:FA and meta:Referees
- User:DavidLevinson/Future proposes version certification process, whereby an article be certified by and individual, a "club" and/or a "league".(An analogy to sports clubs/leagues). Individuals may band together into "clubs", and clubs into "league". Indiiduals who certify crappy articles or abuse the certification process can be kicked out. Similarly, clubs can be kicked out of leagues.
- Wikipedia:Semi-protection policy is really a vandalism panacea, and not really relevant to editorial control. It would prevent anonymous editors from vandalising certain pages.
Critique of meta:Article validation
- give an important tool by which trusted editors can be mechanically assisted in repelling vandalism
Part of the problem is vandalism, but a (more) significant problem is mediocre edits or untrusted edits.
Now lets look at this process again, but from a more abstract and ciritical point of view.
A question remains: "why do this?". I'll answer personally: after a year of editing at WP, I've decided that the mechanism for editing and content creation is quite interesting. I like the abilty to wikilink concepts and words: this keeps me from wasting effort to explain basic principles. I would very much like to experiment with this mechanism for the creation of original research. The mediawiki technology is superior to what I can find in any blog I know of. I also like the blog-like social aspects here, namely, that there are talk pages hwere discussions can occur, without polluting the article itself. Lets critique the alternatives:
- Personal Websites. Hard to set up, hard to operate. Hard to fnd readership. This is why cnetralized blogs are so popular: someone else manages the infrastructure.
- Blogs. Blogs don't have the support for TeX or inline mathematics. This makes presentation of science results harder than it could be.
- Planetmath. Planetmath does have this vision, but has not yet built up the momentum that Wikipedia has. Planetmath is also limited to math, whereas this proposal is broad enough to cover not just the sciences, but also the publication of original research in history, sociology, etc. Also, visually, mediawiki is prettier than Planetmath. However, planetmath is a viable contender.
- Arxiv.org. ArXiv lacks a place for discussion. It lacks online editing, and it doesn't have a way of presenting the change history of modified submissions. The upload process of entering revised pdates is slightly cumbersome.
This aint done yet.
- problems with current WP social mechanisms including the Hillman list