On 17/02/2011 13:19, Carcharoth wrote: Unfortunately the magazine article completely ducks the issue of what the conjecture is. Even on a charitable view, it confuses a necessary with a sufficient condition, which would be the *whole point*. This kind of this is actually why this one has not been solved yet on WP: we (rightly) don't allow people to waffle around the facts in order to claim they are explaining. (If you think we do badly, have a look at a standard mathematical encyclopedia: http://eom.springer.de/p/p073000.htm.) Charles

On 17 February 2011 17:09, Carcharoth <carcharothwp(a)googlemail.com> wrote: Something technical but right is 100% better than something highly readable and clearly wrong. That said, the trouble with obsessive nerds who want things 100% right is that articles become hideous unreadable thickets of subclauses. But then, research appears to be a more widely available skill than good writing. - d.

On 17/02/2011 17:09, Carcharoth wrote: Actually I wouldn't in that case; but I might in the case of a more "Scientific American"-style treatment. By the way, I'm not saying that the exposition of mathematical articles, and in particular the lead sections, cannot be improved, because in most cases it can. There is the issue of finding some "middle ground" between an accurate factual treatment (with wikilinks of technical terms) which is what a mathematician from another field would want, and a more popular treatment. Compare for example http://en.wikipedia.org/wiki/E8_%28mathematics%29, and in particular the fourth paragraph of the fourth section on "Representation theory", with http://www.aimath.org/E8/. The latter treatment is a decent example of the media coverage that a certain computation received not that long ago: but you can't extract from it exactly what was done (just that people thought it was exciting, and some general context). Anyway, I think this issue is going to remain with us. My experience with expository writing is that, no matter how much effort you put into the basics, there will always be someone who thinks you should do more. Charles

Actually, I think the point of the mathematics articles, is that many of them (especially the more advanced ones) are written and used by practising mathematicians. See the comment here: http://en.wikipedia.org/wiki/w/index.php?title=Wikipedia_talk%253AWikiProje… So some might object to your use of the term sophomore, but the rest I agree with. You need people who have experience explaining things to make things like that accessible, but I would suggest technical writers and those who are good at popularising and explaining science and maths topics. Carcharoth On Thu, Feb 17, 2011 at 1:57 PM, Fred Bauder <fredbaud(a)fairpoint.net> wrote:

On Thu, Feb 17, 2011 at 3:15 PM, David Gerard <dgerard(a)gmail.com> wrote: However, one of the arguments being put forward is that too much explanation breaches the provisions against "not a textbook" and "original research" (i.e. providing your own opinions instead of sourcing it to others). I have some sympathy with that viewpoint, and the view that there is a need to balance these issues that are in tension with each other. Carcharoth

Here is my attempt at a historical explanation for the way things are at the moment. First, mathematicians in general are often reluctant to say things that are mostly right but formally incorrect. It's part of the culture of the field, which was reinforced by a certain writing style that became popular in advanced mathematics in the later 20th century. Second, for a few years there was a lot of pressure on Wikipedia to tighten up the referencing on math articles. Writing techniques that were commonly accepted in the early days, like inventing examples or making up informal analogies, were suddenly deemed "original research". Edits like this [1] are not rare today, where someone thought that a section that seemed easy and informal must actually be OR. Fortunately the examples in that section are actually covered in many textbooks, so I could just add a citation. But if the example was written just for Wikipedia, it would be very hard to maintain if someone seriously challenged its inclusion. The current state of many math articles reflects a combination of these trends. When we were asked (not always nicely) to make math articles stick to the sources, which are usually written in a dry, technical way, math editors mostly agreed. After all, we can read the sources, so we can read articles that resemble them. Recently, there has been talk of making articles more accessible. But many of the tools that we would use in other writing aren't available. * We can't just leave out the technical bits, like most popularizations do. * We can't invent examples and explain them in detail, because of the original research policy and because Wikipedia isn't a textbook. * We can't freely use analogies and informal explanations, for the same reasons (see [1] again). Many math editors care about accessibility, of course. But the confines that we are asked to write in are very tight, which makes it a particular challenge. - Carl 1: http://en.wikipedia.org/w/index.php?title=Kleene%27s_recursion_theorem&…