Initially, I was going to write this as a response to this discussion. It seems, however, a larger issue, which I think merits a separate thread, perhaps a sequel to this one. For people who have not read the threads, the context is that two comments of Pete L. Clark on the thread about sponsored speakers were removed without notification by Jeff Atwood.
Apparently this is the new criterion for removing a comment:
if the moderator agrees that the comment does not add to the post in a constructive way
I find Jeff Atwood's standards for removing comments puzzling and disturbing. I myself have removed comments, on very rare occasions, where they were genuinely noise, offensive, or spam. I am also surprised that Jeff Atwood is enforcing these standards (and, more generally, assuming the duties of a moderator) on the present website, of which he is not an active contributor (that is, mathematically), when there are probably bigger issues (i.e., on meta.SO) to take care of, where his role might be more useful. The president of the U.S. does not participate in a local board meeting.
I think that, if these heavy-handed interventions continue, we will have actual problems on our hands. Let us consider the "optics" of the present one. Pete L. Clark, professional mathematician and high rep user, known for his frequently entertaining, witty, and intellectual remarks over the internet (for those who have followed MO, at least) says something on a thread where his contributions might be most intently followed -- on a thread on math.SE's potential role in the mathematical community. Jeff Atwood, who is not a member of this community, unilaterally deletes the comments, without even consulting the moderators first.
Here is the text of Dr. Clark's comments:
@everyone else Not to rub it in, but the people who go to math conferences tend to be research mathematicians. For those people, if you go to a conference stumping for math.SE, they'll probably look at you a little confused and ask "Do you mean MathOverflow?" When you say no, you mean a more recent general purpose math site that is owned and operated by someone with a self-professed ineptitude for mathematics and documented exasperation with mathematicians, many will probably be disappointed.
@all: I scratched my head for a good while about this, and then I googled a little and saw that Jeff Atwood has left basically the same message at many other SE sites. Perhaps that information will be relevant to you if you're wondering how well thought out this is relative to the math.SE community in particular.
And here's what Dr. Clark had to say about those comments in the discussion following their removal:
I'm sure that Jeff Atwood knows which comments he deleted. I do not save such comments for posterity, but the gist of the first comment was that I didn't see that the original post made a lot of sense or applied to math.SE in any clear way, and this confused me until I saw that almost the same [or in fact maybe identical; I didn't check that carefully] posts have been made on many other [maybe all?] SE sites. So I took as an answer to my question that this initiative is not particular to mathematics in any way.
In the second comment I mentioned my opinion that the professional math community would probably receive such sponsorship rather awkwardly, since MathOverflow is currently the flagship math Q&A site among professional mathematicians. I included a remark about how, unlike MO, the founder of math.SE has gone on record as being hopelessly inept at mathematics and has also expressed particular exasperation at the way mathematicians behave...the implication being that he has not exactly acted as an ambassador for mathematics.
It is very hard for me even to see how these are nonconstructive. It seems that the only "nonconstructiveness" of these deleted comments lies in their being, er, not uncharacteristically praising. Perhaps the "self-professed ineptitude" upset Mr. Atwood, but it is hard to see how quoting Mr. Atwood's own words to argue that he is not an ambassador for mathematics could reasonably do so. Dr. Clark later clarified that he realized that the SE team had made similar posts across the network (which answers the first question), but his point that this proposal has not been completely thought out, and that advertising a general math site like math.SE would be received a bit funnily at a research conference, is completely reasonable.
But whether it is reasonable isn't even the issue. Mr. Atwood apparently (according to an email he sent us) thinks that this is somehow critical of the community; if it is, then he simply could have responded in kind, rather than effacing it from the thread. (And what is wrong with being critical of the community?)
So I, as a moderator, have no interest in removing comments as "nonconstructive." Although math.SE is not formally a professional forum, its most active participants are either professional mathematicians or experts in related fields. I always saw math.SE as a (non-research-level ) sister to MathOverflow. Which is to say that this is supposed to be a gathering of adults interacting as equals (and potentially arguing, on meta), who do not (entirely reasonably) wish to be treated like prattling kindergarteners.
As far as I know, the three standard moderators (I'm going to say that Jeff is ex officio a moderator), do not delete posts for these expansive reasons. So I find it upsetting that Jeff would walk over all community norms to enforce these guidelines. For instance, suppose by some accident I were given moderator powers on StackOverflow. The extent of my programming knowledge is that I can script a few lines in bash or get Python to compute the Fibonacci numbers; I wouldn't be able to distinguish between Ruby and PHP. So if I started deleting comments about the programming community that dared to be critical, it'd be surprising if I didn't get called out for it.
(For the record, my brief perusal of SO and meta.SO suggests that those communities tend to be much more acerbic and harsh in their comments than we do over here. As far as I can tell, we are a relatively gentle community, and should be even gentler in moderation; there have been, to my knowledge, only two high rep users whom we have suspended.)
I don't know StackOverflow. I do know it is a company that wants to make a profit (and I realize maintaining this place costs money, and would understand the possible introduction of mild and tasteful advertising or something like that), and perhaps it is thought best to have a sanitized community where insufficiently deferential comments are deleted. But that's just so totally antithetical to what mathematics is about. On the main site, a high-schooler can correct a college professor if she has made a mistake; on the meta site, a regular user should be able to correct Jeff Atwood.
To wrap this up, I am asking Mr. Atwood to think very carefully before making such unilateral interventions in our community. There are three moderators, and all of us check the website relatively frequently. There are already automated features to remove genuinely offensive comments (six flags deletes a comment). Mr. Atwood's presence as a day-to-day moderator on this website should not, thus, be necessary.
I have made this request before, and for a while, it seemed to work. But Mr. Atwood has continued to play a role as a moderator behind the scenes, usually (fortunately) restricting himself to obvious spam. It seems that now, he has chosen to ignore his own theory of moderation and unilaterally remove thoughtful comments. These sorts of actions alienate not only Dr. Clark, but also the rest of us. Many of us, frankly, find your actions, not such comments, to be unnecessarily abrasive. Please stop.