Dissonance of purpose: What kind of site should MU be?

Question

It appears that there are two factions on MU. Those of us who want MU to function as something like a lower-level MO, and some of us who want a less-focused general math forum.

The main difference here is how picky MU should be with its questions. I summarize them as follows:

Lower-level MO: Questions should contain mathematical content at any level, but should be focused explicitly on mathematics. This appeals to students enrolled in math classes (or classes in allied fields(physics, chemistry, economics, etc.) or in an undergraduate math program.

General Math Forum: All questions are welcome if they are math-related, but not necessarily math-focused. This appeals to the general public, but appeals less so to students.

These two factions have different enough ideas that it may warrant a fork of math.SE.

Here's a list of people to read on meta for a good idea of how each faction would like math.SE to run:

Lower-level MO:

Tom Stephens
Katie Banks
Noah Snyder
Akhil Mathew
Charles Matthews
Harry Gindi

General Math Forum:

Justin L.
Noldorin
John Gietzen
Vivi
Edan Maor

If I've gotten your position wrong, please feel free to edit this post and fix it. If you'd like to elaborate on your own position, please feel free to leave a comment or an answer.

Discuss.

Answer 1

Several people here have addressed things I wanted to say more eloquently than I've been able to, but let me add a few things.

I don't want us to be like MO in the sense of "exclusionary" that several people have brought up (I would take issue with that, but...). MO has gone a long way towards providing a useful service and community that was not on the Internet before MO. In the same way, I think there are a number of people, represented by some here, who strongly desire a place to ask and answer questions of interest to undergraduate and early graduate students of mathematics, professionals in related fields, and anyone who uses substantial mathematics in their everyday work or play and has had difficulty finding answers to some of the questions that come up. (Note that this covers most of what was included in the Area 51 proposal for math.SE, even though it is not all-inclusive in the way some "SO partisans" are promoting).

MO has remained a big part of the discussion, I think, not so much because it's crucial that we decide our position and function in relation to MO specifically, but because several people have looked at MO at some point and said "wow, that's the kind of thing I was looking for--but it's a little bit too specialized for me." MO is a natural thing to look at because it's a StackExchange-based site about mathematics that has answered a lot of the questions we're asking and is doing very well for what it does. SO is a natural thing to look at because it's a StackExchange site that has done well, though it's not about mathematics. But we're not just looking to decide that we're a certain percent MO-junior and a certain percent SO. (This is one reason the partisan language and the listed division into camps at the top of this question bothers me a little bit, though it's a minor point).

One of the reasons I have been advocating so strongly for a big undergrad/grad presence is that I think it is not inevitable, whereas the arrival of a large number of users who would like math.SE to be more of an SO for math is inevitable. I think that if we don't set up a structure that encourages "higher-level" math questions from the beginning, we'll get hardly any: they'll be overrun by sheer volume, and people won't come here to ask those questions. I suspect that's not as dire as some are painting it--in particular, that there will remain enough people interested in answering "beginning questions" to keep the site going. But I think math.SE would lose something--the ability to fill some of the void between MO and something like "Ask a Topologist," Art of Problem Solving, or the cesspool that is something like Yahoo Answers. I have nothing against a broad mathematics site, but I don't see what we are adding to the resources and communities available online if we don't position ourselves as a group particularly welcoming of higher-level questions (or at least some kind of more narrowly-defined question) where they would go oft unanswered on other math question sites.

I hope this communicates something non-stupid; please edit and comment, etc. But I'd like to emphasize that I wouldn't describe myself as an "MO partisan," as much as I have learned from and am influenced by MO. I think this is absolutely a conversation we need to have, but not in the language of rifts, cracks, or taking sides.

Answer 2

Excellent question, I agree that it's the underlying issue of a lot of disagreements. I kind of disagree with your characterization of the different positions, though.

When I think of what type of site MU should be, I imagine it to be a place, much like Stack Overflow is for programming, where you can go and ask any question about Mathematics. That's the entire definition, as far as I'm concerned.

As far as I'm concerned, every other decision flows from the idea that MU is a place to ask any Maths question you want. The questions we should be closing are only those which aren't asked properly, e.g., questions that no one can answer because of not enough info, etc. Any question that is answerable, should be answered on MU. That includes even the basics, because a lot of people with questions about Maths will have basic questions.

As I see it, pretty soon, we're going to have many users stumbling in here, who don't really know that much about Maths, but are still interested in asking questions. Chances are, they won't have much mathematical sophistication, nor knowledge of how these SE sites work. These are the kinds of users I expect to be using this site.

Note: This is my view of things. There is one big problem: is this kind of site going to be interesting enough for people to stick around and answer questions? Stack Overflow has managed to still be interesting for good programmers to answer questions on it (and most of the really basic questions have already been answered, so there's nobody asking them again). Will this work with the crowd here? I don't know.

Answer 3

I think that you have my position correct. However, let me elaborate:

The way Stack Overflow was built was on the grounds that any programming question that had an answer could be answered there.

This lead to an environment of openness that was attractive to expert programmers.

If you look at the state of Stack Overflow now, you will see that it is growing steadily, and most of the questions are real, difficult questions. If we allow the easy questions to be asked, yes we will have a couple of boring questions every day, but we will be able to close duplicates, and leave the "most recent" questions view to be mostly populated by pure math questions.

Answer 4

One of the reasons for my stance (which you've characterized correctly) is that fundamentally, StackExchange sites are not intended for discussion. The format is explicitly not that of discussion: CW posts don't get reputation bonuses, comments are posted in order of votes and not in chronological order, and so on.

There are already math discussion boards. By contrast, math.SE has the potential to be a place for people to ask and answer specific questions about mathematics. MO has been very successful with this model, and I think the niche for general math is still unfilled.

I'd say that we ought to be more tolerant of discussion-y questions than MO, since we're not trying to run a website aimed at professionals. However, I'd vote to close a question like this, even though it's fine on another site.

Answer 5

Charles Stewart brings up an interesting point with

"[he/she]...wants to ask real mathematics questions ... but doesn't want to be obliged to internalise a specialist mathematician's way of looking at and talking about mathematics."

As one continues with their study of mathematics, the feel of the subject changes and one's understanding of how to "do" mathematics changes. Initially, math is an exercise is pushing symbols and computing a solution to a mathematical question or statement. Later on, new mathematics is learned through arguments (proofs) that describe a new concept or a theorem solely in terms of previously internalized pieces of mathematical knowledge. At this point, to "do" math means to learn how previously internalized pieces come together to reveal new mathematical truths. (This is grossly oversimplified, but I think it will work in order to make this point)

To those of us who have come to this point in our development (and it is not a pinnacle), questions such as "What is an inner product space" and "Intuitive reasoning behind the Chain Rule in multiple variables?" can only have answers like this and this.

Now, this poses a problem for our site. The askers of those questions were probably not looking for the answers I have highlighted. But - to me, a member of the first camp in Harry's classification, answering those questions with any less rigor would be a disservice to this community (whatever level you are), and any answers that are hand-wavy I would down vote and criticize without hesitation.

I don't know what kind of answer the asker may be expecting. I have read and have enjoyed books like A tour of the Calculus, but MU cannot be forum for the artful exposition of popular mathematics. There are shows like Numb3rs (which has just been cancelled, thank god), and other blogs for that.

So, how can we address the divide?

Answer 6

I've been pondering this issue from just after my first few hours on this site and I have a few thoughts I'd like to share.

First, I think the divide between the two groups is not much of a divide at all, for the majority of people in each group. I think that there have been some remarks here and there deriding people for being stupid or asking stupid questions that have served to create the appearance of a greater divide and/or served to polarize people. The more that I think about it, the more that I suspect the two groups are not diametrically-opposed in views, even at the extreme.

Second, it seems to me that there are four likely scenarios:

1. We don't hash this out before the public beta and it turns out there isn't a problem at all--things work as I and others feel they have on SO, where the questions span a wide range and that's okay.
2. We don't hash this out before the public beta and the "lower-level MO" people are right. In this case, I suspect that many of them will go on to form the MathUnderflow site that they wanted this to be.
3. We hash this out now and the "general math forum" people prevail. This will likely lead to either 1 or 2, though possibly with reduced participation from the "lower-level MO" crowd, depending on whether they stick around to see which case falls out.
4. We hash this out now and the "lower-level MO" people prevail. In this instance, I strongly suspect that those of us who did not envision such a site will go on to try form the more-general math SE site that we'd envisioned.

Given the directions of my leanings, my preference would be to leave this unresolved and see what happens.

Finally, I feel like a prevalent argument for the "lower-level MO" direction is that open math forums have failed in the past, so it doesn't matter if the SE model has worked for compsci/programming on SO, because that's not math. Having experienced numerous math and compsci forums over the years, including usenet, I don't think the compsci forums that predated SO were any better than the math forums, so I don't see any reason to think that math.SE can't follow the model of SO for success.

Answer 7

I'd like math.stackexchange to be exactly what was set forth in the charter:

A Q&A site for people studying math at any level & professionals in related fields

What kind of site do I think we should be building? Well, eventually, it would be a place full of people with interested in math, coming from all backgrounds and all levels of ability. Anyone can post a question, from the electrical engineer who forgot a few topics from the complex analysis class he took back in 1963 to the middle-schooler asking about a question he saw in Mathcounts that afternoon. Simple questions posted in accordance with the rules will be answered almost immediately. Hard ones may sit for a while, but if they are interesting, will draw upvotes and partial answers, but they'll get answered too, even if it takes 2 weeks and a 500-rep point bounty.
People looking for an answer to some math question will search for it on google, skipping the top result (which is a wikipedia artcle, and will be until the end of time) and find out that someone else already asked the same thing on math.stackexchange, receiving 3 outstanding answers. After this happens 4 or 5 times, that person realizes that this is a useful site for getting answers to math questions, and becomes a regular user.

Incorrect or misleading answers get downvoted. Poorly formed questions get comments that give guidelines for improvement. Questions in violation of the rules get downvoted and closed. Duplicate questions will be rare because past questions are clearly labeled and have comprehensive answers.

Questions will be correctly tagged: the majority will be labeled correctly by the asker because the tag system is sensible. Those that are not will be quickly corrected by high-rep users or moderators. People interested in any topic at any level of difficulty can easily find questions they care about and filter out ones that would be annoying noise.

That's the ideal, anyway. That's how it should be. The problem is, how do we actually accomplish that...

Answer 8

I could be wrong, but I doubt that SO has a lot of questions like "Could someone give me an intuitive understanding of how to associate a value with a subclass in c#."

I think people should try to avoid unnecessary jargon and do their best to explain something, but I also think askers need to commit to the fact that math questions are answered with math, and you need to work to understand them.

Answer 9

Incidentally, there is a similar divide on MO, between "leftists" (who prefer to close fewer questions) and "rightists" (who prefer to stick to the original FAQ in a strict-constructionist manner). I myself am something of a jaded leftist who eventually saw the virtues of running a tight ship; the consensus seems to have shifted towards conservatism. Nevertheless, the divide still continues there, as you will find if you read meta threads; there are respected contributors who disapprove of current policy.

Closing threads has always been somewhat contentious -- more so than it needs to be -- which is the main source of the MO divide, and it's possible it will be the source of many future debates on meta.math.SE in the future. But then again, perhaps this is only betraying my own bias as math.SE is an MO-esque site.

However, I suspect that the "divide" is smaller than it seems, and in particular less so than on MO. Both sides seem to agree on the main points (that the website should be relatively open to answerable general math questions, and overly vague questions should be closed).

In any case, I don't think it is necessary to resolve our disagreements right now; we don't have sufficient data about what the culture of this website is going to be like, and most likely both visions will evolve slowly. There is plenty of time to decide what this site should be. The fact that I've personally been on one side of this so far is likely no more than a reflection of my having spent a fair bit of time hanging around MO.

Answer 10

One thing that we have to take into account is that a large number of StackExchange sites are forming. We should try to ensure that each question goes to the most appropriate site/sites. Of course, there will and should be some overlap and when there is, hopefully the answers should differ depending on the site. For example, the MathOverflow answers might be much more advanced than those on here, or questions like this one would be much more focused on the computational complexity on StackOverflow. I think that we should try to have as wide a scope as possible, only limiting the scope to improve the quality of questions on this site.

Answer 11

I have been thinking about the various ways to address the differing desires on this site and I have come to the opinion that perhaps there is a way to merge both, but to do so we will all have to admit that there will come a time when we are no longer interested in this site. When there is some new shiny tech that let's us make better math websites.

I think if we want this site to have the largest appeal we need to make it as exclusive as possible, initially. This is the reverse of what I was initially thinking, but it makes the most sense when I think about it objectively.

Ideally we will reach a sweet spot with this site, where math professors will be coming to answer questions, (even though they might be annoyed by many of them), college kids will be answering questions, and getting their tough questions answered, and everyone will also be able to ask and answer questions.

I think that is ideal, but I think it will be temporary. Eventually as the site grows the more advanced members will look elsewhere to congregate. And eventually all the question will become noise...which is kinda of where SO is going.

And then we start over. Hopefully with better tech.

But we are not going to attract top talent as easily by saying this a math website for everyone. If we could initially limit users to only ivy leagues schools, that would really help with the brand of the site. I know that sounds awful, but bear with me.

Which sounds better. "Its a math site where anyone can answer your questions" of "Its a math site where the brightest math stars from Harvard answer your hardest math questions."

Don't be confused, I don't even think you need to attend college to be intelligent to answer difficult math question. That's not my point. I'm talking about branding. Branding is all about status.

I want this site to be for everyone, but I think making it exclusive at first will get us there the fastest. If we can't limit the user base, then who we advertise to initially, add what questions we have on here will help set the tone.

Answer 12

I think you've misread Vivi's stance. I apologise for speaking on her behalf, but it may be worth saying this quickly, if there is a crack that is spreading: I understand that she wants to ask real mathematics questions, of the sort relevant to an economist, but she doesn't want to be obliged to internalise a specialist mathematician's way of looking at and talking about mathematics.