Transcendental Limits

"The English language is like a broad river on whose bank a few patient anglers are sitting, while, higher up, the stream is being polluted by a string of refuse-barges tipping out their muck." (Cyril Connolly)

This was the year I lost my taste for social media. I left Twitter and even stopped blogging for a while. It turns out, however, that I love blogging and I missed writing in this medium. Also, I think the ease of blogging sets an important standard for knowledge communication. There is, in an important sense, no longer any excuse for not being able to read a straightforward statement of the opinion of a prominent figure, whether a scientist or a politician. It's not like such a statement is difficult to publish.

If you've been accused of research fraud or plagiarism, your statement of guilt or innocence should be on your blog immediately. If some new discovery has been made that bears significantly on your research, your interpretation should be forthcoming. If you have made a discovery, a statement of your methods and findings should be neither behind a paywall nor "under review". Just blog it. Anything else only exhibits your ulterior motives.

The good thing about blogging is that there is nothing necessarily anonymous about it. You can correctly identify yourself and keep your blogging otherwise entirely impersonal, i.e., within your area of publicly asserted expertise. There is no particular reason to hide, nor, like I say, to keep your opinions on matters you are (often publicly funded to be) qualified to hold opinions about to yourself. If everyone has access to clear and coherent statements of what you know and how you know we're all better able to do our work.

The best example of a blog that is using the power of the Internet to the fullest is Andrew Gelman's Statistical Modeling, Causal Inference, and Social Science. Andrew's posts are written informally and with a clear point of view, usually from within his own expertise or with a clear awareness of the limitations of its application to a particular case. Andrew's commenters are widely regarded as among the best on the planet. The comments are [very lightly]* moderated. I don't know how much work that demands of Andrew, but his judgment (or his bot) seems both pretty open-minded about the range of opinion that is allowed and sure-eyed when it comes to filtering out garbage.

But what about the "trolls"? Well, first of all, we have to qualify that term. One pivotal moment in my distaste for social media came a while back, when a critical discussion of some published research results was shut down by the moderators of OrgTheory, a blog I had been reading and commenting on for years. I had been participating, and many of us thought the discussion was on point and interesting. Still, the explanation for shutting down the comments was the catch-all excuse these day: "the trolls were moving in". That was a very inaccurate way of describing what was going on. Weaknesses were being exposed in a piece of published research, that is really all.

Now, it is absolutely true that trolls do, often, move in and make conversation difficult in less moderated forums. But most platforms have simple, effective and relatively unobtrusive ways of dealing with them. Trolls can often simply be ignored. They can be hidden. And they can be blocked. If you can't makes sense of someone's statement as anything other than an insult, just leave it alone. Assume that no one else is taking it seriously either. If someone you respect brings it up and says, "Why didn't you address [the troll's] point?" you can ask them to restate it in language you can engage with. These aren't sticks and stones. Simple advice like that.

That said, there do seem to be many un-moderated comment sections that are best described as "cesspools". This doesn't mean the posts themselves are worthless, just that getting anything useful out of the comments is too much work to be worth the bother. And contributing something to them is likewise a waste of time. That's why personal blogs with a few hundred readers, all working within a relative narrow range of expertise and writing about subjects from a point of view that is technical enough to discourage attention whores and trouble makers are preferable to the so-called blogs of major media organizations.

I recently had what might be called a "liminal" experience, something that showed me what is wrong with today's media, or what we sometimes still call the Internet. Through channels I'll coyly leave on one side for now, someone told me about a website called Omegle. I'm pretty sure it's entirely safe to use the site as a long as you don't give out any personal information. But proceed at your own risk, I guess.

It is endearingly simple. You have no account or profile, and are therefore fully anonymous. When you go on the site, you get paired with another equally anonymous user for a one-on-one chat. (There is a completely random mode, and a function that lets you declare an interest in a particular subject to be paired with accordingly.) You (or they) can end the chat at any time and instantly start another one. The site claims that there are tens of thousands of people online at any given time.

In principle, one would think that such a site would serve a purpose. Why shouldn't a brief conversation with a total stranger anywhere in the world (someone who has nothing to lose except the time involved) be intermittently interesting? Well, you can try it. I spent almost fifteen minutes trying to talk to someone interesting, constantly running into, not trolls but (I think) bots. The conversations went like this:

[CHAT BEGINS]
Me: Hello
Stranger: girl here. im really bored. i'm from Illinois.
Me: I'm bored too. At work.
Stranger: i got pics, u should kik me if u wanna see ;)
Me [if I type fast enough]: No thanks.
Stranger: [a Kik user name]
[END OF CHAT]

You can have literally hundreds of these exchanges, I imagine, and never run into someone who is actually a bored girl (or boy or man or dog) in any real place. Obviously, there is something about a completely free platform for communication that immediately gets descended on, not by trolls looking to infuriate people, but advertisers hawking their wares. (For obvious reasons, the site is seen as a market for pornography.)

And yet, I would think a filter could be made to keep bots and merchants out. It would, however, probably require that at some end of the site anonymity doesn't exist. That is, you would need to have an account that you couldn't really just "burn" to break the rules. Unless something like a spam filter or anti-virus software could be designed to essentially "Turing test" all the interactions and block the relevant IP addresses, you'd have to put your real self "at stake" somehow. Otherwise, it seems, there is just no reason for advertisers not to abuse the medium and crowd out the very attention that supposedly makes it a market. As far I can tell, Omegle at this point must just be thousands of bots telling each other they're "bored in Illinois" and promoting their Kik accounts.

I'll leave it there. Internet communication appears to depend on some sort of transcendental limit that is able to filter spam without incurring too great a cost to the communicator. There are many models for this. A blog like Andrew's is one of them; a place like Wikipedia is another. It is sad (but entirely understandable) that a place like Omegle can't actually exist. Somehow it shows us the practical limitation of entirely free speech. It's a sort of worst case scenario after the collapse of decorum.

__________
*Andrew has clarified the way comments are managed: "the comments on our blog are not moderated. There's a spam filter which occasionally catches non-spam by accident (and then I can never find them unless someone specifically emails me that this has happened) and some comments are held in a cache which I check once or more per day and sort as spam or real. But I think that in all these years there have only been three or four non-spam comments that I've actually deleted because of offensive comment.