Vanilla 1.1.9 is a product of Lussumo. More Information: Documentation, Community Support.
Jared Keller from TheAtlantic.com wrote to me to ask some questions, and I thought I'd copy and paste his questions and my answers here, in case people want to disagree, argue, or expand on them! Also, I think Ben Webster spoke on the phone with Jared this afternoon, and perhaps will tell us how that went. It seems I have to split up this post into several sections, sorry about this.
Hi Jared,
I've written answers below each question.
On Sep 20, 2010, at 13:48, Jared Keller <...snip...> wrote:
Hi Scott,
Thanks for getting back to me so quickly, I've got a couple questions with regards to MathOverlow; feel free to answer them at your leisure.
1) What was the thought process behind creating MathOverflow? Was the idea of collaborative research a long-time interest of yours? I know variations existed before, namely Gowers' Polymath Project. Just want to get a sense of your focus and intent leading up to the creation of the MathOverflow.
I wasnt directly involved in the prehistory of MathOverflow: that was Anton Geraschenko and Dave Brown. You can find a description of their thoughts, and how they got started with Stack Exhange in an early post on meta.mathoverflow.net. I'm offline at the moment so can't look it up for you, but you may find looking at meta interesting: all of the politics and daily management happens there. Inevitably much of the planning happens offline, however, since many of the moderators are or until recently were at Berkeley. This is actually an important point: it's no surprise that MathOverflow got started at Berkeley, as a certain critical mass of enthusiastic and forward thinking grad students engendered the environment in which this sort of new approach to doing mathematics can be discovered.
I got involved, along with Dan Erman and shortly afterwards Ben Webster, during the first week the site was running, but before any public announcement. Essentially, Anton was looking around for Internet-savvy friends, both to help seed the site with good questions before going public, and to ensure there was an initial group of invested people. The first public announcement was on the Secret Blogging Seminar, a joint blog of several former Berkeley math grad students, including 3 of the current MO moderators (me, Ben and Scott Carnahan). (Interestingly, MO seems to have seriously affected both the volume of posting and the readership on many research level math blogs, SBS particularly. No one much seems to mind however, as it seems natural that everyone involved wants to spend their time on MO now!)
I've certainly been interested in using the Internet to improve the way mathematics gets done, but most of my interest has always been in how publishing works: the particular "questions and answers" model is definitely Anton's.
(I should double check the dates on the polymath project: my vague memory was that MO predates it, but I'm probably wrong.) ED: yup, totally wrong.
2) Was there a particular architecture you were looking for in creating the forum? What made you decide to go with StackExchange/StackOverflow? What's your relationship with the team at StackExchange?
See above for the first two parts. It's interesting: before Stack Exchange 2.0 came out, we'd felt that we had a great relationship with the SE team. Anton would regularly correspond with team members, and they were pretty responsive to our requests. In fact, shortly before 2.0 Joel Spolsky met Antimony for dinner to talk about how MO and SE were going. This all reflects the fact that MO was pretty much from the beginning amongst the most active of the SE 1.0 sites, and certainly it had the most cohesive and "professional" community.
Since the advent of SE 2.0, well, they've been busy. There's meant to eventually be a process for existing active SE 1.0 sites to transition, but we haven't heard much about it. I'm a little worried --- in particular there are a few things in the transition that we're going to insist on, and they might not see things our way --- but Anton I think remains confident it will pan out okay.
Anton and I had lunch with one of the founders of quora a month or two ago, to talk about MO. He'd previously suggested that we consider migrating and offered help, but prior to lunch we'd already told him it seemed very unlikely. It was really interesting to see his perspective: he took notes, and I wish I had too.
3) My Stanford friend describes MathOverflow as a "social media revolution" for the math world. Do you think this is an accurate claim? Do you see a role for social media portals like this (based on Digg, Reddit, etc) in the mathematical research in the future?
The really important point, which your questions don't exactly address so I'm going to stick it in here anyway, is that we've tried really hard (and I think largely succeeded) to differentiate MO from "the Internet" and social media, in the minds of our target audience, research mathematicians.
The Stack Overflow team comes with similar biases: they've consistently turned down requests for "social networking" features in the software (at least until the chat servers came along). We've taken it further, however.
Mathematicians as a whole are surprisingly skeptical of many aspects of the modern Internet, despite having been early adopters of email, etc. In particular, things like facebook, twitter, etc. are viewed as enormous wastes of time. Indeed, Greg Kuperberg (a mathematician from Davis, heavily involved in the arxiv and well respected regarding 'maths on the internet') initially told us that MO was "too twitter-like" and that thus he didn't like it. Fortunately he changed his mind and became one of the most active users.
Given this, we've tried to make the forum as "professional" as possible. In the early days, we worked hard (emailing new users, etc) to encourage everyone to use their real names, which has paid off handsomely. (Many of the exceptions amongst the most active users are obvious initialisms, simply to ensure that googling for the full name still returns a home page rather than MO content!) We've been pretty strict about good behavior, too, absurdly beyond what is common on the Internet. If it wouldn't be appropriate at department tea or particularly during a seminar, it's not on at MO, either. (You can find occasional disgruntlement over this at meta.)
So while I agree with your friend's claim that MO really is having an impact on mathematics research, and it surely does count as "social media", I don't think this has any significant implications for mainstream social media and scientific research of any variety.
4) What distinguishes MathOverflow from other math forums?
Well, what is there?
Historically, sci.math.research, but it pretty much died, in the same ways that everything good on usenet died.
After that, blogs (especially "the week's finds in mathematical physics", possibly the oldest blog on any subject, along with its companion "the n-category cafe" and Tim Gower's and Terry Tao's blogs), russian livejournal pages, polymath, ...?
In some sense, there just wasn't much out there that admitted general contributions. Still, I think the Stack Exchange model was particularly well suited to mathematics. Mathematics is pretty much the canonically perfect field for throwing up precise, technical questions. There are plenty of questions with no room for interpretation or discussion, and which admit a definitive, identifiably correct answer, and far more of the questions mathematicians care about fit this description than in any other field. These are the questions that work best on SE, and we've unabashedly tried to set up the social norms on MO to prefer them.
Moreover, mathematicians love answering other people's precise, technical and above all esoteric questions, so there's scope for a very very long tail. You wouldn't believe how frequent it is to hear people talking at tea "Is there an X that Ys but doesn't have a Z?", purely for the sake of understanding all the consequences of a definition and for the pleasure of thinking about something so baroque. (Algebraic geometers, who are a dime a dozen on MO, are particular prone to this, I think.) Now mathematicians don't have to go to tea anymore...
5) Your guidelines for asking questions are pretty stringent. Is there a guiding principle shaping them?
Not so secretly, Anton has said: "I want MO to be useful to me!" Thus he's carefully training a huge group of mathematicians to promptly and thoroughly answer any questions that arise in his research outside his usual expertise. More seriously, you should read the "How to ask a good question" page on the main site, and many of the early threads on meta. (It's unfortunate that there's no good index to the interesting stuff on meta.) A standard guideline is "questions should be of interest to research mathematicians". This immediately sets a rather high lower bound on the technical level: anything that might be a routine question from a student in office hours for any undergraduate (indeed, many graduate) course is out. That guideline leaves open the interpretation as "of interest to anyone who happens to be a research mathematician" or as "of interest when they have their research mathematician hat on", and there's been plenty of back and forth here. One of the big successes of meta has been the active discussion of this sort of issue, and you can see the "rules" have evolved somewhat over time, as have individual participants opinions.
But again, Anton's secret answer really gets at what I personally want mathoverflow to be: restricted to technical questions, as few "general interest" questions as are felt to be required for "community building", and such that people think of it primarily as a tool for connecting them with other experts on questions that arise during research, but that are obviously best answered by someone else.
6) How successful has MathOverflow been as a node for mathematical collaboration? For example, were there complicated problems posted that might've puzzled mathematicians that produced elegant answers in a matter of days after posting?
Read the mathoverflow success stories thread on meta. I have a paper out of a mathoverflow question, for example.
7) Could the model you've developed for MathOverflow potentially work for other disciplines?
Very likely, and we're about to see with all of the new SE 2.0 sites. For the reasons mentioned above, I think mathematics occupies a special place relative to what this sort of software can do, and we might not see many SE sites where such a high proportion of the users are experts. (Theoretical CS being an obvious candidate.)
This is what I've got for now. I may send you more as I learn more about the program.
Thanks so much!
You're welcome, and I'm looking forward to seeing what you end up writing.
btw -- do you mind if I post these questions and my answers on meta? You'd likely get better answers from others as a side effect!
Thanks for posting, Scott.
I'd like to react to one part of Jared Keller's first question:
Was the idea of collaborative research a long-time interest of yours? I know variations existed before, namely Gowers' Polymath Project.
I think there is a misapprehension out there, which Keller may or may not hold, that collaborative research itself is a new idea in mathematics. Polymath and MO are definitely something new, but collaboration and department tea are not. A great deal of mathematical research has been collaborative for a long time, and has become much more so in the last few decades. (Older technological innovations like email have presumably played a major role in that development.)
I don't have a ton to add about the phone interview I did. I think it went well, and I at least got the sense the feeling was mutual. The questions (and answers) are not too different from what Scott said above, though I think we talked a little more about the general future of math and the web.
1 to 9 of 9