I’ve been taking a break from Twitter and reflecting on why I needed to take a break from Twitter. The short version is that dealing with other people is a fundamentally difficult problem and I’m not sure that Twitter is helping. I have hopes for Twitter, but I have several very critical things to say about it too (and when I say them, I will be doing that thing where I argue in favor of my position and just trust that there are valid points that people could make against my position and not volunteer to make those points for them—it’s my rant, dangit).

What is Twitter good at?

Much of what is good about Twitter is what’s good about email: it makes practical a wider spectrum of expression than before. At the point where email and Twitter came along, we already had lots of ways to express ourselves - good ones, too! The written letter, the telegram, the telephone, the fax, the postcard, the magazine, the newspaper article, the letter to the editor. These were all good ways to express oneself: however, they were not, collectively, good at expressing all of the things that a person might want to express. Change over time made some of them less good (farewell telegrams) and overloaded some (the fax machine changed how people dealt with letters).

What email and Twitter have in common is that they made it practical to express things that were before possible to express, but impractical. They are communications innovations not because they do something better that older forms did well, but rather, because they do something well that older forms did poorly. This is why Twitter is supremely unlikely to replace IRC: they are good at different things. Conversely, this is why email often replaces a letter that is written on paper, sealed in an envelope, marked with a stamp, and put in the mailbox: most of what you want a letter to do, an email does better. Not all, but most.

So I don’t harsh on Twitter for brevity and incompleteness: it’s an ephemeral medium, by design, and it’s to the credit of Williams, Stone, and their team that they cultivated technical limitations into an interesting way to express oneself. “Twitter is not good at sustained and complex conversation” is merely facts on the ground: this is part of why Branch and Medium exist. On the other hand, I’m perfectly willing to kvetch at individual Twitter users for using the service for things that it is not good at and never will be good at—more on that later.

The question of what is Twitter for? is perpetually up for debate; I won’t attempt it here. What I can authoritatively answer, though, is the question of what do I want Twitter for? I want a stream of answers to the question “what’s interesting lately?” (with a conscious caveat that people of course provide the same variety of answers to that as they do in offline life), a river of idea-seeds, and a conversation at the speed of a fencing match. Twitter should be a great place to start things: to find or express a new idea, to ask a question and get concise answers, to point people towards something that matters to you.

Often, Twitter falls short of this. This is not necessarily Twitter’s fault, but it’s certainly a good enough reason to leave.

Noise without signal

Too often, Twitter is like a noisy party: even if it’s full of people you like, you can’t be paying attention to everyone at once, and a high enough population means that it becomes difficult to actually talk even with people who are trying to talk to you in return. That’s an annoyingly easy situation to get into even when you’re trying to avoid it. By contrast, Twitter turned into the flying, destructive, multi-generational party from Life, the Universe, and Everything some time in 2010 and has not improved since (including the part about “one of the problems with a party that never stops is that all the things that only seem like a good idea at parties continue to seem like good ideas”).

The noisy-party thing is quite annoying to me: I’m picky to start with, I can’t stand following more than about 100 people. My Twitter stream stops being useful and enjoyable when its population gets significantly bigger than my monkeysphere. It’s similar to my thoughts on how many physical objects you should own: if you’re ignoring an object in your home or a voice you’re following on Twitter, why is it there in the first place? You are a human with a finite attention span that you don’t actually have a choice about respecting. If you’re following 1,000 people on Twitter, it is almost certain that you’re not really listening to what most of those people have to say—which means in my opinion that you should just not be following them.

People are used by spammers

On top of that, people don’t always speak for themselves, but regularly have their voices appropriated (consensually or not as the case may be) by the most tediously screeching and incoherent kind of commercial and political interests. Reading Twitter is less worth my time when people push every gas-station fill-up, every FourSquare check-in, and every whale-shaving petition they sign, to their Twitter stream—part of why Tweetbot and TTYtter were my favorite clients was that they had robust filtering abilities to silence such bullshit.

That’s just the things that people click on themselves, mind you: there’s also a thriving community of accounts that are spammers from top to bottom, ways to get people to tweet on your behalf without noticing beforehand, and attempts to hijack Twitter accounts for the purpose of posting malicious links of one kind or another. As busy as that community is, fortunately it is not very successful. So the main problem here is less that malicious actors may jump in, than it is that people choose to permit the forces of idiocy to speak on their half. Ugh.

Private accounts are bullshit

This annoyance is multiplied when people make it harder to listen to them speaking in the first place by the expedient of private accounts—which I consider to be a wart on Twitter’s design. There are a small number of good reasons to have a locked Twitter account: if you have to ask, yours probably doesn’t make the grade. So I’m talking about the majority of locked accounts here, not 100% of them. That said: locked accounts degrade the Twitter experience and show poor judgment, and I wouldn’t be the least bit sad to see Twitter disable them entirely.

One of the big virtues of Twitter is serendipity: the property that things you say might be picked up by unknown others, that you may have unplanned interactions, that someone may say “ahah, but what about this related piece of information?” and you’ll have something new and interesting in your day. Private accounts are free-riders there, partaking of the benefits of that community dynamic without giving back.

Locked accounts also make the Twitter experience worse by giving the middle finger to the normal “follow or not?” decision. Normally on Twitter, you decide whether or not to follow someone by looking at what they’ve said and asking, “do I want to see that in my stream of new messages?” But private accounts are like “nope, trust me or fuck off.” Your plan for evaluating whether a private account is worth following rather involves being on the other side of this airtight hatchway. That’s bad, that breaks the virtuous cycle of “if you want people to follow you, post things worth reading.”

Creating a private account also usually shows poor judgment: one of the normal failure modes of private-account-holders is thinking that Twitter is a substitute for LiveJournal. In the LiveJournal paradigm, it’s totally reasonable to have an account where your statements are private by default and once in a while you say something to the world at large, or to have your primary account and your “after dark” account. That works somewhat on LiveJournal, but it’s a ruinously bad fit for Twitter.

On LJ, your replies to other people’s statements (comments on their entries) are generally not private, and you can selectively step out of private mode to say something to the world. LiveJournal provides an array of technical tools that you can use to modulate your expression between public and private. Twitter does not give you such tools: if you reply to someone’s tweet, they can’t see you unless they’re already on the other side of the airtight hatchway, and you can’t say “this tweet will be public, but just this one time.” So you’re not a full participant the way you are on LiveJournal.

The other way that private accounts show poor judgment is summed up in “After Dark” accounts. I think that those are almost always a bad idea: maintaining multiple Twitter accounts is a major headache, but more importantly, Twitter is a broadcast medium. You are not excused from discretion! Even if Twitter is “oh that’s where my friends are” to you, there are things that it’s a bad idea to say in public—and Twitter is public, like the Web in general. Communicating discreetly with your friends about your kinks is definitely not one of the things that Twitter is good at. I also think that if you are the kind of person who is comfortable talking about your sex life (or your mental health, or similarly mainstream-culture-proscribed subjects) in your public voice, then people who can’t handle hearing about that from time to time should not be following you. Your voice should be yours first, when it’s a personal voice like Twitter: it is fine and necessary to modulate one’s professional voice according to audience demands, but a personal voice should be personal and heartfelt and true-to-self above other things.

False urgency and false equivalency

Then there’s a somewhat more abstract concern: the “flattening” of expression. Because of the size limit on tweets and the way that Twitter strives for simplicity of presentation, all tweets have roughly equal visual weight. Unfortunately, this means that drastically different content looks and reads the same. It bothers me that “my hopes for a good Transformers movie died,” “my dog died at the age of 20,” and “my mother died at the age of 48” all have the same look. It’s extraordinarily challenging to modulate this with regards to user-generated content, but I think that some level of it is necessary because inviting us to consider every statement on Twitter to be equivalent in importance, is not a good thing.

This is worse when it comes to arguments, because “I don’t think the evidence supports that” looks like “you smell like a goat’s asshole” in your timeline. Or more germanely: the sponsored tweet looks very much like your friend’s tweet about how the President is a godless Commie Muslim bent on destroying the gold standard, which in turn looks very much like your other friend’s tweet about how the President is a simpering and ineffectual prat who’s set back his cause by decades.

I am not entirely satisfied with the metaphor that Twitter “flattens” these expressions, but it’s important to me because it’s symptomatic of a design that is not clearly enough communicating to people what Twitter is good at and why they should use it for that—and not for other things. Of course people are going to use Twitter for things that it’s bad at as well as for things that it’s good at. We do that with email, letters, and video, too. It still bothers me.

One aspect which especially bothers me is that this flattening can lend a false urgency to things. One feels a need to reify alarming and attention-grabbing tweets before they disappear by replying to or acting on them. This is a dangerous road to walk. We already as humans have problems judging urgency: misjudging urgency reliably drives us to do stupid things. Twitter’s format encourages this because the ephemeral nature of tweets creates a sense of time running out, of opportunity in the middle of disappearing. It is no accident that that exact urgency is also used as a sales technique: it drives us to make hasty and poorly-considered decisions.

I suspect that this problem would be less annoying if I were less enmeshed in the furry community, which has an ongoing issue with people who are already grievously prone to inaccurately judging urgency and to thereby creating massive social problems. The urgency problem is a human universal, though, so I’d probably see the same thing to greater or lesser degree among medievalists or knitters.

What now?

I’m not sure how badly all of these things bother me about Twitter—but it’s definitely enough to make me take a break for a while. I’m also consciously cultivating other forms of my own expression—like this blog!—to see whether I’m thinking things that are best expressed in ways other than Twitter. I’m enjoying that so far, and I expect that it’ll be a fruitful exercise.

If you decide to stay with Twitter, I wish you the best, and I hope that you either find these problems less serious than I do or have better ways of coping with them (if you do, please mention it to me). For my part, I’ve noticed that in addition to these problems existing, they are problems either with the design of Twitter as a service and social mileu, or with how other people use Twitter. Neither is within my power to unilaterally change, which makes turning my face away from Twitter more appealing since I can change that quite easily.

postscript: A correspondant points out that Penny Arcade also has a critique of certain styles of Twitter usage.