Wiki science/Wikiresearch

Wikipedia was started beginning 2001 and can be considered highly successful. It is however not the proper place for original research. Hence the idea of Wikiresearch, a project to do wiki-style scientific research: collaborative and under a free license.

Wikipedia
Wikipedia is a free encyclopedia being written collaboratively by voluntary contributors from all over the world. The site is a wiki, meaning that anyone with an internet connection can edit any article except for a few protected pages Wikipedia:About. Since 2001 over 260.000 English language articles have been written and over 410.000 articles in other languages. Wikipedia's parent organization is the Wikimedia Foundation Inc., a non-profit corporation organized under the laws of Florida.

All the articles are licensed under the GFDL and the MediaWiki software that the project runs on is released under the GPL. This makes sure the encyclopedia remains free. Anyone could start a mirror or a fork.

Serious and less serious information about the sociology of Wikipedia can be found on the Wikimedia Meta-Wiki.

MediaWiki
Mediawiki has quite some advantages over other Wiki systems . It makes no use of UpperCaseLinks or little icons to indicate the existence of a link; the colors of the links indicate whether it's an external link or an existing or still non-existing article. The edit history has an adequate diff function, which facilitates maintenance and makes eradicating wiki vandalism a snap.

The modular WikiTeX system adds the capability to incorporate rendered TeX objects inside MediaWiki output. GetWiki, which is used by Recyclopedia, is a fork of MediaWiki and adds XML import functionality. The problem is that the modified parts are licensed under a Creative Commons license and cannot be reintegrated into MediaWiki, since the project people want to keep it GPL only.

Some more free collaborative projects
The Mediawiki software is also used for several other collaborative projects with the goal to create free and open information:
 * Wikimedia projects, all using the GFDL:
 * Wikibooks, a collection of textbooks
 * Wiktionary, a multi-lingual dictionary
 * Wikisource, a collection of source texts and their translations
 * Wikiquote, a collection of proverbs and quotes of famous persons
 * Wikivoyage, a travel guide
 * projects not from Wikimedia:
 * Wikitravel, a travel guide.
 * SourceWatch, a collection of information about "public relations firms, think tanks, industry-funded organizations and industry-friendly experts that work to influence public opinion and public policy on behalf of corporations, governments and special interests." Using the GFDL.
 * Consumerium, a project that develops free software and infrastructure for "storage, transport and display of product information to consumers and feedback to the producers to enhance the consumer experience, advance product development and to keep this planet as hospitable as possible in the future". This project decided to do most of its research directly on Wikipedia.
 * CorpKnowPedia, a project that charts the corporate landscape.
 * designbreak.org, a wiki for creating and collaborating on research and engineering projects
 * MetaCollab.net, an open research, meta collaboration (a collaboration on collaboration) with the aim to explore the similarities and differences in the nature, methods and motivations of collaboration across any and every field of human endeavour. Hosted by Wikia.com.
 * Wikiresearch, a synthesis of knowledge through collaboration
 * Citizendium, an encyclopedia that seeks to create credible articles

No original research
Wikipedia is not the place for original research such as "new" theories.

Wikipedia is not a primary source. Specific factual content is not the question. Wikipedia is a secondary source (one that analyzes, assimilates, evaluates, interprets, and/or synthesizes primary sources) or tertiary source (one that generalizes existing research or secondary sources of a specific subject under consideration). A Wikipedia entry is a report not an essay.

Free software and free information
In comparison to free software the threshold to join a free information project is much lower. There is not a lot of unnatural syntax to learn. Typos and spelling errors won't crash anyone's computer and won't cause loss of valuable data.

It's not important to be an expert to add information to a wiki. Almost any part of human knowledge are welcome in the Wikipedia project - as long as it's written down in an encyclopedic manner. Information about travelling will feel more at home in the Wikitravel project. Course books are warmly welcomed at Wikibooks. Nearly anyone with good will, the ability to type, and Internet access can add some of his or her knowledge.

Projects like Wikipedia share the spirit of free software yet have the advantage of being much more accessible. There are many more people that can correct a spelling error than there are programmers that can fix a bug in even the tiniest software project.

Generally most wiki projects evolve like stone soup. At first there is hardly anything interesting to be seen. Then people come by and think "Ah, I can add some lines here". The Wikipedia project attracted thousands of people thinking like this in only three years' time. Many of these people became regular contributors. Contributors can roughly be divided into three groups: correctors of spelling mistakes and typos, people who add some lines now and then, and content adders.

Already Wikipedia is the biggest freely available encyclopedia on the internet. Considering the exponential growth of the project, it should surpass all other existing encyclopedias within the next decade - both in quantity (100 million articles in 100 languages?) and in quality - already Wikipedia is linked to by many news web sites in the world.

Scientific institutes and the market
Currently there is a trend to move financing of science from the government towards the marketplace. According to the accompanying neoliberal view scientific institutes such as universities or more specialized research institutes are to find more of their financing in the market place, which according to the neoliberal theory would lead to more useful knowledge and products.

Private funding may bias the resulting findings, since members of the institute might lose their income when the results aren't to the funder's liking. Another consequence is that more and more results will only be available to the funders, or to parties willing to pay and use the results under restricted conditions. Results can be in the form of:
 * software
 * patents, for example on software,
 * file formats (such as MP3)
 * knowledge about organisms, medicine

Since science is built on existing results, institutions that act on the marketplace are often less eager to quickly share results, which thus results in less cooperation. In this regard there are huge differences between different scientific cultures. At one end of the spectrum one can find abstract mathematics, that initially have very few direct applications and at the other end there are applied sciences such as robotics and genetics, where results are of utmost commercial interest.

In mathematics it is normal to publish results with the greatest care for detail. In applied sciences, where institutes tend to be much more competitive, less room is left for cooperation. This often leads to publications that reveal certain aspects of the results, but not a detailed implementation.

Free software
Developers in the 1970s already shared their software in conformity with the principles of free software. In the 1980s companies started imposing restrictions on users with the use of license agreements. In 1983 Richard Stallman started working on the GNU project and introduced the concepts of "free software" and "copyleft", which are specifically devised to give users freedom and to restrain the possibilities for privatization. 

These days nearly every scientific institute makes use of free software of some kind. Although typically a free/propriety software combination is used instead of all free software, such as running MATLAB on a GNU/Linux OS. Often there is no urge to move away from the proprietary model to a more collaborative model.

A lot of free software is created by scientists. Sometimes this is paid for the institute they work for, but more often this is done in their spare time. 

Publications
For a scientist it is important to publish results. Most renowned journals are owned by large publishers.

Many scientific articles are by publishers who place restrictions on their use. Some articles can be freely accessed online (like on the U.S. National Library of Medicine's PubMed Central); in many physics disciplines, especially in astrophysics, cosmology and high energy, theoretical physics, nearly all articles are made freely available on the arXiv preprint server. But often online access is restricted to paid users. Redistribution is hardly ever possible. Modification or reuse beyond simple quoting is out of the question. 

Often publications discuss results obtained with software that is not freely available. In many fields this causes problems with falsifiability of the results. If the software is only roughly described in an article, it is very hard for a reader to repeat the obtained results.

Even inside institutions there is sometimes limited cooperation and mostly competition.

Peer-reviewed publishing in wiki format
There is now a wikicity for peer-reviewed wiki journals.

Wikiresearch - practical
A first experiment of Wikiresearch has already been started in the form of Wiki science. This is a study of the way Wikis are growing, changing and adapting.

Putting finished articles in Wikiresearch is to be encouraged. Published as well as unfinished articles. Also articles rejected by journals and conferences are interesting; they might contain research slightly too original and if the quality isn't yet what is expected from scientific articles they can be improved at Wikiresearch.

Maintainership and responsibility
The MediaWiki software allows to restore previous versions and see which user (or IP address) changed what. Until now the various Wikipedia projects have experienced relatively little vandalism. Regular users with a login often check what has been changed and by whom, where anonymous edits are regarded with a bit more suspicion.

Certain Wikipedia pages are attractive to vandals, but these pages are also checked very often by regular contributors. So on a busy Wikipedia such as the English version vandalised text tends to stay for at most a couple of minutes and the perpetrator will see that his or her action had only very little impact. This is also the reason it can be a good idea not to protect sensitive pages like Bush at the Wikipedia (see also this article's changelog for an indication of how much vandalism occurs at articles about sensitive topics).

For original research things could be a bit more tricky. Only time and a functional system can tell how much of a problem vandalism will be on a research wiki. The Wiki science resource, though still a struggling project, is an interesting test case.

Modifiability, authority
One way to do Wikiresearch is to have several main authors or rather authorities. This however goes against the wiki principle and it probably wouldn't lead to the amount of participation that can be attained with a more anarchistic model, where anyone is allowed to edit and add.

The license
The GNU Free Documentation License (GFDL) would be a good choice since it ensures license compatibility with the existing GFDL text corpus. It would be possible to have one wiki with pages that have a different license than the encouraged license, though this should possibly be discouraged since it comes in the way of this compatibility.

If somehow another license is chosen, at least the attribution of the writers should be ensured. This is good for tracing back the original author and is a way to avoid legal plagiarism. For a Creative Commons license this would mean that the Attribution option needs to be picked. Extra options, like ShareAlike or NonCommercial would mean text can't be incorporated in GFDL projects like Wikipedia or Wikibooks. NonDerivative goes against the entire idea of working in wikistyle.

Note that the Debian project considers both the GFDL as well as the Creative Commons Attribution license to be non-free. There are some people (such as Sloyment) who prefer not to use a copylefted license in order to provide the most freedom that is possible. More discussion about licenses to use for wikis is going on on this wiki (FIXME: Host wikilegal.wiki.taoriver.net not found).

Extensions of WikiMedia software
Though the ratio of the feature list and the time spent on MediaWiki by developers is extremely high there are still some possible extensions that would be of great use for a scientific research wiki.

Being able to import and export Dia and OpenOffice.org diagrams or even a more direct way of drawing sketches would be a great way to communicate new ideas.

If there is a need for accompanying source code this code should definitely not be under the GFDL. It is better to use the GPL or any other free software license.

Sciences
Mathematics is the science most adequate for Wikiresearch. Except for conjectures and axioms it is entirely based on strict proofs, that leave little room for opposing viewpoints. Mathematical Wikiresearch could after a specific period simply be copied to Wikipedia, that is, when enough eyes have dimished the possibility of mistakes and errors.

Though experimental sciences shouldn't be excluded. A bit of biological research has already taken place at the article about color blindness (see also the talk page).

For sciences that require experiments with test persons it can be very interesting to use Wikiresearch. Questions, test pictures and test sounds can be placed on the wiki and there will be enough test persons if there is less need for a .. sample of the general population.

Sciences such as history that are usually about defending a specific point of view are less adequate. But for these Wikipedia can be a proper place if there is enough will to keep a neutral point of view.

Another interesting field for Wikiresearch is VLF-reception with the PC.

Language
Since the late 19th century English has become by far the most important language for both scientific and political affairs Latin. This gives people whose mother tongue is English or a related language (Germanic, Indo-European) a big headstart in their higher education.

The Wikipedia project has encyclopedias in over 80 languages Wikipedia:Multilingual coordination. Polyglots are free to translate information from one language into another and this way even very small language communities can work on extending a knowledge base in their own language. Many languages for which no publisher would ever see an opportunity to create an encyclopedia now have growing encyclopedias.

This freedom of translating texts can be fruitful for the extension of human knowledge. Research can be translated from English, or any smaller pivotal language, into other languages and research can be undertaken in any other language and translated into a pivotal language later.

People
Traditionally science and scientific texts state the names of the authors more or less prominently. Wiki projects tend to anonymize texts; on MediaWiki wikis authors are usually only mentioned in the page history. Anyone can make little improvements, although there are usually one or more core authors of a text or article.

Most established scientists probably won't be attracted to working on Wikiresearch. There are however people who will be attracted to doing Wikiresearch: people who haven't finished their studies (yet) - or haven't even started them; people who chose a different career path; or scientists that want to work on fields completely different from what they are known for.

If a research wiki can attract a sufficient number of qualified people the wiki can probably qualify as a system equal to peer review. The difference is that peers aren't chosen among a small group of highly respected people. This can lead to both more original research as well as more charlatanesque works. The expectation is that especially the starting period will be of utmost importance.

Related

 * A lot of work can be done on improving Wikipedia articles, that can also be used as a clear definition for general concepts.


 * Results could take the form of a Wikibook.


 * Wikiversity is a Wikimedia project which is geared towards learning. Wikiversity policies to support research and researchers is being developed at the multilingual Wikiversity hub.


 * Presentation given at Oekonux 3 conference


 * Research is also being done at the German Wikipedia. See Wikipedistik


 * Research on Meta-Wiki should serve as a master page for collecting all research projects.

Examples

 * The TrustLet project