« Inequality |
Main
| Anil on the loss of accidental social intermedaries »
January 14, 2004
Adolescence goes public
Posted by Seb Paquet
The New York Times Magazine has a
good article on teenagers who keep online journals. The author observes how the uptake of journaling among teens opens up new windows enabling everyone to peer into the experience of adolescence:
A result of all this self-chronicling is that the private experience of adolescence -- a period traditionally marked by seizures of self-consciousness and personal confessions wrapped in layers and hidden in a sock drawer -- has been made public. Peer into an online journal, and you find the operatic texture of teenage life with its fits of romantic misery, quick-change moods and sardonic inside jokes. Gossip spreads like poison. Diary writers compete for attention, then fret when they get it. And everything parents fear is true. (For one thing, their children view them as stupid and insane, with terrible musical taste.) But the linked journals also form a community, an intriguing, unchecked experiment in silent group therapy -- a hive mind in which everyone commiserates about how it feels to be an outsider, in perfect choral unison.
Rather than forming a single, interconnected network, journals form a multitude of
relatively closed worlds:
Blogging is a replication of real life: each pool of blogs is its own ecosystem, with only occasional links to other worlds. As I surfed from site to site, it became apparent that as much as journals can break stereotypes, some patterns are crushingly predictable: the cheerleaders post screen grabs of the Fox TV show ''The O.C.''; kids who identify with ''ghetto'' culture use hip-hop slang; the geeks gush over Japanese anime. And while there are exceptions, many journal writers exhibit a surprising lack of curiosity about the journals of true strangers. They're too busy writing posts to browse.
I wish I could learn more about those journal writers who are indeed curious towards strangers and the role that
communities of interest play in satisfying that curiosity.
Comments (4)
+ TrackBacks (0) | Category: social software
- RELATED ENTRIES
- › Spolsky on Blog Comments: Scale matters
- › "The internet's output is data, but its product is freedom"
- › Andrew Keen: Rescuing 'Luddite' from the Luddites
- › knowledge access as a public good
- › viewing American class divisions through Facebook and MySpace
- › Gorman, redux: The Siren Song of the Internet
- › Mis-understanding Fred Wilson's 'Age and Entrepreneurship' argument
- › The Future Belongs to Those Who Take The Present For Granted: A return to Fred Wilson's "age question"
1. Clay Shirky on January 14, 2004 10:31 AM writes...
"Rather than forming a single, interconnected network, journals form a multitude of relatively closed worlds:"
I'm fairly certain that Nussbaum is wrong about this. The original research that got me started on link distribution was in the winter of 2000, when Scott Hieferman turned me on to LiveJournal. At the time, it had something like 250,000 users, a small enough number to spider relatively easily, which I did.
Because of the "You have to know someone to get a journal/small median number of friends" characteristics, I was expecting to see many discontinuous pools of individuals. I had just finished reading Duncan Watts' Small Worlds, and was looking for counter-examples.
And LiveJournal turned out not to be a counter-example. Though the median number of friends on the system was 5, the average was 12, indicating a high degree of connectivity. The surprise, which Nussbaum misses, is that there is a huge, world-changing difference between 'disconnected' and 'sparsely connected' -- sparse connection is all you need to make "a single, interconnected network."
Nussbaum says it, and then misses its importance: "Blogging is a replication of real life: each pool of blogs is its own ecosystem, with only occasional links to other worlds." The occasional links -- Gladwell's "Connectors" -- are all that is needed to create a six degrees of separation pattern.
Permalink to Comment2. sneJ on January 14, 2004 11:20 AM writes...
Then play anthropologist -- get a LiveJournal account (they're freely available now, no more "invite codes") and start looking for experimental subjects. Read through likely communities, press the "Random" button on the home page to view a random journal, read comments on posts to find other people who sound interesting.
Then friend them, comment on their posts, read their RSS feeds, whatever. (If you friend someone, it's best to (a) comment on their posts -- who wants a lurker? -- and (b) post stuff in your own journal so they have something to read.
But of course don't hide who you are; there are already plenty of grown-ups on LJ [like me!], it's not all whiny teens =)
Permalink to Comment3. Seb on January 14, 2004 12:31 PM writes...
I reckon you're right, Clay. Wish I hadn't taken out the "well-" I'd typed in before "interconnected" just before I hit publish. Indeed, just a hint of clique-breaking links and you get one large connected component. Which might explain how those quizzes rip so quickly through Livejournal...
Permalink to Comment4. Clay Shirky on January 14, 2004 3:27 PM writes...
So here's the funny thing about those quizzes: when I was testing my LJ spider (this was a couple of years ago), I had it output the average cluster size with each iteration, to watch the thing in action, and for the longest time I thought I had a bad bug, because I would pick one user at random, who usually had ~5 friends, and the thing would get to work, crawling all of that users friends, and then their friends, etc.
The program would output a single number -- total users examined / total friends listed -- with each iteration, and the list would start out 5.0, 5.2, 5.2... like that, and then at a certain point it would go nuts -- 25, 35, 50, 100, 150.
I would stop, look for bugs, couldn't find anything, restart from a different initial user, same pattern. Finally I output the user names as well, and what I found was that I had inadvertantly developed a hill-climbing algorithm for popularity.
Anyplace you start on LJ, if you move one degree out, then one degree out from all those users, you will, by definition, be liklier to come across those users who are listed more often. Random samples, in other words, are strongly biased towards well-connected users. (This is exactly the pattern you alluded to in your "Dont link up" comment on the "Inequality" post.)
And by finding those users who are linked more often, you will, if you go one degree out from them, be reaching hundreds of people at a time, sometimes even 1000+, instead of 5 at a time.
So a quiz, if it gets taken by some random subset of any given users friends, and then some random subset of the friends' friends, will necessarily crawl up the poplarity index, until high-flow users are advertising the test.
And yes, it only takes a tiny, tiny amount of connectivity to do this. (For some reason, I always think of the tiny amount of water you have to put into corn starch to turn it from a powder to a gel -- just a little bit of connection changes the way the substance behaves.)
Permalink to Comment