9
$\begingroup$

Generally in my answers I like to add sources to, in some sense, back up what I claim in said answer. I often use wikipedia, but recently have seen a few comments saying "don't trust wikipedia entirely for advanced mathematics". I personally haven't seen any fallacies from wikipedia, but then again I'm not exactly in high level math (the math I'm studying right now I can't exactly find on wikipedia ;)). The general question here is as in the title:

Is wikipedia a valid source to use on MSE?

Recently (meaning today) I have been using wolfram pages as my backup. I do not believe the general public can edit that so I feel it may be safer. Is there some sort of regulation for this?

$\endgroup$
  • 11
    $\begingroup$ How is it "safer" to use as backup something over which you (or, for that matter, anybody other than a corporation) have no control whatsoever, something that the owners can hide and charge you for at whim? $\endgroup$ – Andrés E. Caicedo May 23 '14 at 2:04
  • 3
    $\begingroup$ @AndresCaicedo generally, I would trust wolfram to publish valid mathematics articles, this is what I mean by "safer". From your comment it seems one should balance fearing capitalism and false information, is this what you meant? $\endgroup$ – DanZimm May 23 '14 at 2:13
  • $\begingroup$ @user147263 ah I didn't catch that, only searched in the meta-MSE. This still doesn't answer whether it's an acceptable source to use on MSE however. $\endgroup$ – DanZimm May 23 '14 at 2:14
  • $\begingroup$ @user147263 I did not see that statistic on the post, albeit it is a good one! $\endgroup$ – DanZimm May 23 '14 at 3:12
  • 1
    $\begingroup$ Maybe related: Are mathematical articles on Wikipedia reliable? and Pete L. Clark's answer in another thread on meta. $\endgroup$ – Martin Sleziak May 29 '14 at 12:12
22
$\begingroup$

Yes, Wikipedia is usually valid. Link to it freely, and trust it at least as much as you trust a textbook or a site like Wolfram. This has been discussed at length here.

If you have time to actually read or glance through what you're linking to (which is ideal), then you can make sure to link to good articles, not just to good sites. When I link a page, I often use Google to find several pages and then pick the one I find the most relevant or useful in this particular case. I have found that sometimes the Wikipedia article is best, sometimes the Wolfram article is best, and sometimes some other site is best.

$\endgroup$
  • 10
    $\begingroup$ I would also add that when I link to a wiki or any reference for SE, I am implicitly declaring that I agree with the accuracy of the content in the link (it is valid), and I want the reader to take heed. $\endgroup$ – J. W. Perry May 23 '14 at 6:23
12
$\begingroup$

If you're worried about the volatility of a wikipedia article, if you check the sidebar, in the "tools" section, there is the "permanent link" link, which links to the specific version of the page you're looking at.

So, for example, if someone follows this link which was obtained as I describe above, they will be taken to the version of the ZFC page that wikipedia was displaying at the time of this posting, whether or not it has been revised in the meanwhile.

$\endgroup$
5
$\begingroup$

Try not to make links a major part of your answers, then you have nothing to worry about!

For example, if you want to link to Terry Tao's blog, then link too it and include a nice quote from it or a brief summary which gets your point across. For example, do this rather than this.

I think wikipedia links, and indeed internet links in general, should be only used for one of the following*.

  • "For more information" links. That is, you explain something and then you say "for more information, see here".

  • Defining something they they really should know using a link. The reader could look up "normal subgroup" in a book, you are just being helpful and saving them time. Plus, they really shouldn't need to look it up.

  • A link to a, for example, quotation you are using (as with the Terry Tao blog example, above).

This is for two reasons.

  1. If an answer is heavily reliant on links then it is susceptible to "link rot", where the website linked to stops functioning and so the answer is invalidated. (Note the use of the "for more information" link!) EDIT: A genuine example of industrial-scale link rot can be found here, when Springer encyclopedia of mathematics changed its website.
  2. It means that your answer is self-contained, which is pleasing. Moreover, I don't want to have to click on a link to work out what on earth you are going on about!

*These are the only reasons I could think of at the time of writing. I am sure there are more, so feel free to suggest them. Nor do I believe that there are no exceptions.

$\endgroup$
  • $\begingroup$ Why should a question poser "reall [not] need to look up" the definition of a Normal subgroup. Some people will only be learning, and may not have come accross it before. Obviously this depends on the context, but I think in general this is bad advice. $\endgroup$ – Flint72 May 29 '14 at 18:27
  • 1
    $\begingroup$ @Flint72 That wasn't meant as a general statement, it was more meant as an internal monologue for an imaginary specific situation: "They question is from the start of a second course on groups, so they should know what normal subgroup is, but because it is so fundamental to the question, and because they might have forgotten it, I'll give them a link." $\endgroup$ – user1729 May 30 '14 at 8:14
4
$\begingroup$

I use Wikipedia links frequently for reference definitions of terms and occasionally named theorems, and only sparingly for authority or proofs of the truth of propositions (although I do often link to other sources for such material).

It occasionally happens that someone finds a mathematical claim in a Wikipedia article doubtful enough to pose a Question here about it. In this way MSE can act as a backstop/truth-checker on Wikipedia, and I think this is a natural advantage of the Creative Commons licensing used by both sites.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .