![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Medium, Message, and Moderation
Oct. 21st, 2004 02:03 pmWell, I'm back. Maybe. For a little while, anyway.
This is an experiment, the same way that disappearing for a month was an experiment. The hypothesis I'm testing is: "It is possible to participate in valuable conversation on LiveJournal while minimising frustration and wasted time."
Experimenter bias is high. That's all I'm going to say about that.
Part the First: I'm Not a Human/Computer Interaction Expert, But I Play One On LiveJournal
Like most geeks who grew up in the Eighties, I cut my electronic-communications teeth on bulletin-board systems, primarily ones that participated in Fidonet. During that time, and later on Usenet, I got used to a certain model of communication: threaded, noncontiguous, monotonically increasing, asynchronous discussion. (That's a fancy way of saying "you can track who's saying what to whom, but they don't all talk at the same time, and related messages won't necessarily be next to each other; on the other hand, later messages will be either new threads or replies to earlier posts in a thread.") Because replies were often noncontiguous, quoting relevant bits of a message was common; because most people didn't check the board more than once or twice a day, messages tended to have a high signal-to-noise ratio. This was more the case on Fidonet than on Usenet, of course, and I see it as an artifact of the medium: Fidonet discussion groups were called "echoes" because every night, each participating board would receive a call from a distribution board, which would download copies of messages posted to the echo on that board and upload copies of messages posted to the echo on other boards. Thus, barring technical problems that prevented your sysop from getting that day's messages, every BBS carried a complete archive of every echo in which it participated, as far back as your sysop cared to archive it. On the other hand, until the echo updated every night, you could only read messages posted by people on your local system, so if you wanted people on other systems to reply to you, it was in your best interest to be as coherent and interesting as possible. Instant gratification was a lot easier on Usenet, even in the days of dialup-only; you still had noncontiguity and asynchronicity, but the degree of asynchronicity was a lot lower. It became easier for exchanges to be rapid and short.
Now, fast-forward a bit to the web-forum model. Threading becomes more obvious, reinforced by page layout, but noncontiguity disappears; except on really active forums (say, Fark, SomethingAwful, Unforum, or Slashdot), all the relevant context is right there on the page where you're replying. There's really no need to refresh your reader's memory of the discussion with a concise precis of The Story So Far if all he has to do is scroll up a few centimeters and see what's already been said. Asynchronicity drops off even more, prompted by self-reinforcing models that are, paradoxically, aimed at increasing discussion quality -- what's easier, getting a Slashdot post modded up +5 Insightful or +5 Funny? If all you want is ego-stroking, it's easier to pop off a quip and be visible on everyone's comments page inside of half an hour than it is to put serious thought and research into something that will, more likely than not, get lost in the shuffle.
I've tangented a bit, but it's with a purpose; I want to capture what I see as the prevailing communication model on the Web today, contrast it to what I grew up on, and compare it to LiveJournal.
LJ, obviously, hews far closer to the web-forum model. It's not as fast and furious as Slashdot or Fark (well, unless you're Cassie Claire, I suppose), but it looks like a web-forum and quacks like a web-forum, and the majority of users post to it as if they're posting to a web-forum. At the same time, however, the design of the Friends page brings with it a variation on a feature of the BBS model: out of sight, out of mind. A thread no one's replied to in a day or two is dead. Similarly, once an LJ post has scrolled off the Friends page of everyone who was likely to read it, comment traffic drops off dramatically (unless the post becomes a meme, e.g. the "no pity, no shame, no silence" meme); the post is, by and large, dead.
(As a side note, I should point out that I am mainly speaking from my own experience here, but I think this would be terribly interesting to analyse empirically. There's probably an HCI paper in it. Might be worth talking to![[livejournal.com profile]](https://www.dreamwidth.org/img/external/lj-userinfo.gif)
bradfitz about. Dunno.)
I value in-depth communication. I value a serious exchange of ideas. In the first few days after I went radio silent, the people who bothered me about it got an earful about how I'd become frustrated with drive-by discussion and "friendship by press release" -- and both those phrases have now been reduced to conversational macros in and of themselves, due to overuse (aka "having to tell people the same thing over and over again"). I'm not real keen on this either, but I'll expand on that in the next section.
(to be continued)
This is an experiment, the same way that disappearing for a month was an experiment. The hypothesis I'm testing is: "It is possible to participate in valuable conversation on LiveJournal while minimising frustration and wasted time."
Experimenter bias is high. That's all I'm going to say about that.
Part the First: I'm Not a Human/Computer Interaction Expert, But I Play One On LiveJournal
Like most geeks who grew up in the Eighties, I cut my electronic-communications teeth on bulletin-board systems, primarily ones that participated in Fidonet. During that time, and later on Usenet, I got used to a certain model of communication: threaded, noncontiguous, monotonically increasing, asynchronous discussion. (That's a fancy way of saying "you can track who's saying what to whom, but they don't all talk at the same time, and related messages won't necessarily be next to each other; on the other hand, later messages will be either new threads or replies to earlier posts in a thread.") Because replies were often noncontiguous, quoting relevant bits of a message was common; because most people didn't check the board more than once or twice a day, messages tended to have a high signal-to-noise ratio. This was more the case on Fidonet than on Usenet, of course, and I see it as an artifact of the medium: Fidonet discussion groups were called "echoes" because every night, each participating board would receive a call from a distribution board, which would download copies of messages posted to the echo on that board and upload copies of messages posted to the echo on other boards. Thus, barring technical problems that prevented your sysop from getting that day's messages, every BBS carried a complete archive of every echo in which it participated, as far back as your sysop cared to archive it. On the other hand, until the echo updated every night, you could only read messages posted by people on your local system, so if you wanted people on other systems to reply to you, it was in your best interest to be as coherent and interesting as possible. Instant gratification was a lot easier on Usenet, even in the days of dialup-only; you still had noncontiguity and asynchronicity, but the degree of asynchronicity was a lot lower. It became easier for exchanges to be rapid and short.
Now, fast-forward a bit to the web-forum model. Threading becomes more obvious, reinforced by page layout, but noncontiguity disappears; except on really active forums (say, Fark, SomethingAwful, Unforum, or Slashdot), all the relevant context is right there on the page where you're replying. There's really no need to refresh your reader's memory of the discussion with a concise precis of The Story So Far if all he has to do is scroll up a few centimeters and see what's already been said. Asynchronicity drops off even more, prompted by self-reinforcing models that are, paradoxically, aimed at increasing discussion quality -- what's easier, getting a Slashdot post modded up +5 Insightful or +5 Funny? If all you want is ego-stroking, it's easier to pop off a quip and be visible on everyone's comments page inside of half an hour than it is to put serious thought and research into something that will, more likely than not, get lost in the shuffle.
I've tangented a bit, but it's with a purpose; I want to capture what I see as the prevailing communication model on the Web today, contrast it to what I grew up on, and compare it to LiveJournal.
LJ, obviously, hews far closer to the web-forum model. It's not as fast and furious as Slashdot or Fark (well, unless you're Cassie Claire, I suppose), but it looks like a web-forum and quacks like a web-forum, and the majority of users post to it as if they're posting to a web-forum. At the same time, however, the design of the Friends page brings with it a variation on a feature of the BBS model: out of sight, out of mind. A thread no one's replied to in a day or two is dead. Similarly, once an LJ post has scrolled off the Friends page of everyone who was likely to read it, comment traffic drops off dramatically (unless the post becomes a meme, e.g. the "no pity, no shame, no silence" meme); the post is, by and large, dead.
(As a side note, I should point out that I am mainly speaking from my own experience here, but I think this would be terribly interesting to analyse empirically. There's probably an HCI paper in it. Might be worth talking to
bradfitz about. Dunno.)I value in-depth communication. I value a serious exchange of ideas. In the first few days after I went radio silent, the people who bothered me about it got an earful about how I'd become frustrated with drive-by discussion and "friendship by press release" -- and both those phrases have now been reduced to conversational macros in and of themselves, due to overuse (aka "having to tell people the same thing over and over again"). I'm not real keen on this either, but I'll expand on that in the next section.
(to be continued)