Hacker News new | past | comments | ask | show | jobs | submit login

Yeah, this is a big problem with less rigorously covered subjects. If you think of Wikipedia as a summary of existing sources on a subject (which is what it aspires to be), and the existing sources suck, then the summary, by transitivity, also sucks. But there's not much that can be done about that within the scope of the mission, defined in that way. If you instead think of Wikipedia as a compendium of true information on all subjects, the problem gets much harder: then it would aspire not only to summarize all information ever produced, but also to vet all that information for accuracy, correct anything incorrect published in any academic field or in the popular press, and fill in gaps where the third-party sources are lacking entirely [1] ... plus convince people that these corrections, despite no citations, are true. I think that is better tackled in separate projects: one project (Wikipedia) to summarize the existing state of writing, and different projects to improve it in specific fields, e.g. a project to improve documentation of 20th-century punk rock, or to document the history of open-source software.

Wikipedia has, though, tried to cautiously make a few exceptions to what counts as a citation to address some of the more specific problems relating to individuals. Personal blogs are not generally considered published sources, but are acceptable sources for the specific case of summarizing what the person who writes the blog themselves thinks about a subject, in cases where that's relevant [2]. This is used most commonly to cover the "subject's side", e.g. if there is an article about someone that includes negative information, and that person has responded on their own blog, but hasn't managed to get a newspaper to publish their rebuttal, Wikipedia will still cite the rebuttal. I guess that's along the lines of asking your sisters to make an unambiguous statement somewhere about your existence.

The questionable reliability of news articles also means that they tend only to be treated as acceptable sources for newer things. If you're writing about WW1, it's frowned on to directly cite New York Times articles on the conflict, because some of them were wrong in hindsight, and we now have much better books and journal articles written about it, which have done all the legwork of reading through the newspaper archives and assessing their reliability. At this point, making a new argument about the conflict based solely on previously-unnoticed news articles is original research that ought to be submitted to a journal, and only to Wikipedia if it's accepted by the historical community first (this is not purely hypothetical with WW2, where people really do try to come up with novel interpretations of the Holocaust based on a new reading of old newspaper archives). Alas, for newer stuff there's often no such alternative, short of just not covering the subject at all, so newspaper articles are accepted on something like the Syrian civil war, because they haven't yet been superseded by anything better.

One thing that's interesting to me is that some of this source-evaluation difficulty would be simplified if they had applied some of the same rules as Wikipedia. Sometimes I will cite a New York Times obituary for an article, because it contains a bunch of facts about a person conveniently collated. But then I wonder: where did the NYT obituary author get this information? Is it based on solid first-hand research? Summarizing old NYT articles? Cribbing info from Encyclopedia Britannica? It would be nice to know!

[1] I wrote a bit about the gap problem elsewhere, http://www.kmjn.org/notes/wikipedia_notability_verifiability...

[2] http://en.wikipedia.org/wiki/Wikipedia:SELFPUB




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: