On 01/06/2014 10:53, Ting Chen wrote:
Nowaday Wikipedia articles (across all major languages) are highly
biased in style and in content to academic thesis.
There is good reason for this: 'anyone can edit'. In an encyclopedia produced using the 'one best way' approach, there is sparse use of references and citations. Take this article on the syllogism http://plato.stanford.edu/entries/medieval-syllogism written by Henrik Lagerlund. I don't spot any references, and generally SEP is sparing in their use. Henrik doesn't need to supply references, because he is an expert in his field, and because there is a traditional peer review process supporting SEP.
In Wikipedia by contrast, 'anyone can edit', and there is no equivalent peer review process, and so the only control is insistence on citations.
This is part of what makes it difficult for newcomers. I remember well the period 2006-7. The growth of Wikipedia was tremendous. Before that, it was possible to manage the occasional 'idiosyncratic' contributors. Towards the beginning of 2007 it became impossible. Then two things happened. (i) It became much easier to get the 'idiosyncratic contributors' blocked. Before that, you had to make a very strong case to a non-involved admin. After that, it progressively became more like shoot on sight. (ii) The policies on citation became increasingly established and enforced. This made it much easier to gain control of an article. 'Idiosyncratic' contributors found it difficult to find reliable sources for whatever version of flat earth theory they were promoting, and got discouraged. There was also (iii) an easy way to control the quality of an article was to impose a sort of change freeze on any contribution, good or bad. I still maintain contact with the few editors left on the Philosophy and NLP articles, and they tell me this is how they achieve it.
Of course, all this will have the effect of deterring contributors. But the underlying reason is the trade-off between quality and participation. If you have a large user base under the 'anyone can edit' policy, then you are going to have quality control problems. If you address the quality problem by any of the three methods above, then you will have to limit participation in some way. No brainer.
I would advise anyone with an interest in this to read Aaron Halfaker's seminal paper on this. The links are in his post here http://lists.wikimedia.org/pipermail/wikimedia-l/2014-May/072267.html .