« Back and Better Than Ever! |
Main
| Geoff Cohen sees the future of Flash Mobs »
September 14, 2003
Codifying Relationships
Posted by Liz Lawley
One of the problems that plagues the "YASNSes" (as Clay calls the growing number of social networking systems) is how to define or codify relationships.
On the one hand, trying to make all relationships equal and bidirectional, as
Friendster and
LinkedIn currently do, is clearly problematic. As I wrote on
Joi Ito's LinkedIn wiki page:
I'd also like to be able to differentiate between (at the minimum) two types of contacts--those whom I'm willing to receive referrals from, and those whom I'm willing to have make referrals on my behalf. There are far more in the first category than the second. I'm more than happy, for example, to have Meg Hourihan or Anil Dash send someone to me. But since I don't have extensive working relationships with either one, I'm not sure I'd want them to be the first line of introduction for me to someone else--for that, I'd be more comfortable with someone like Joi or Clay Shirky or someone I've worked more closely with.
But today I was playing with a pre-alpha version of a new system that does in fact allow me to define types of relationships, and as others have pointed out, that has its own set of problems. In the system I was looking at, I was given the following options:
* I am a close friend of this person
* I am a friend of this person
* I am an acquaintance of this person
* I know this person (by reputation)
* I know this person (in passing)
* I am related to this person
* I would like to know this person
I was trying to categorize my relationship to another system user, a well-known Silicon Valley entrepreneur. I've met the person at a party, and had a brief conversation, but I have no idea if the person remembers me. I'd like to get to know the person better. So...I _might_ be "an acquaintance," I do "know the person in passing," I definitely "know the person by reputation," _and_ "I would like to know this person" better. What do I choose? (I ended up giving up, btw, and not choosing anything.)
This is where
David Weinberger's concerns about making the implicit explicit become most relevant for me. Relationships are complicated. Expressing them algorithmically is terrifically difficult. Reducing the complexity takes something important way from the relationship. And forcing users into these choices without a clear and compelling payoff for doing so (payoff for the users, that is...clearly the marketers and demographers get a payoff!) seems doomed to failure.
Comments (4)
+ TrackBacks (0) | Category:
- 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. Marc Canter on September 14, 2003 3:11 PM writes...
Thank god I'm a SF entreprenuer and would never be caught dead in Silly Valley (unless it's a Reid Hoffman, Ross Mayfield, Jon Abrams fest at Stanford - Tues night.)
Anyway - one concept - the next one to build - is that relationships are in fact dynamic. As are interests, beliefs, emotions - even content (as Dave Winer as proven - much to Mark Pilgrim's chagrin.) So why get so hung up about algorythmically defining relationships - when it's TOTALLY within our grasp - to have these (tentative at best) relationships - change over time - which is (in fact) how the real world is.
Permalink to Comment2. Alex on September 14, 2003 7:22 PM writes...
I think that you potentially both lose and gain from making tacit relationships explicit. But to take off from Marc's comments, I would guess that we are not particularly good at estimating such relationships. I think this is why friendster sometimes seems juvenile. Especially when you are growing up you are trying to come to terms with what it means to have a personal relationship and how your relationships compare to one another.
The big question is, why ask the user about these relationships? Can't they be inferred from the networks of exchange and the content of those exchanges?
It would not be particularly difficult (with the willing participation of the members) to note the frequency of exchanges of emails. Moreover, a very broad linguistic analysis could distinguish between more formal language and, for example, email sex. (I'm not sure there is such a thing. If there is, it is exchanged among people far more patient than I.) In the end, IT could tell YOU who you have relationships with, which might potentially be far more useful.
Permalink to Comment3. Liz Lawley on September 15, 2003 8:12 AM writes...
Marc, you've put your finger on what danah boyd has identified as a major difference between women and men in this social networking context--women tend to be much higher self-monitors, and spend a lot of time (not just in electronic systems) categorizing and understanding the nuances of their relationships. They tend to be more aware of the impact of defining those relationships explicitly.
(This is not to say that men are not...simply that this behavior is more common in women.)
If you're going to build a system that's predicated on defining relationships, it's not unreasonable to expect your users to want those definitions to reflect their mental models. One way around it is not to allow granularity. Once you go down that path, however, you have to work a lot harder to address those "algorithmic" concerns.
And I keep coming back to David's concerns. Why _should_ I codify this in a YASNS? It's already in my head. What do I gain by encoding it? I'm hitting the saturation point on the number of places I'm willing to enter my personal information (Friendster, LinkedIn, Ryze, Tribe, etc, etc).
Permalink to Comment4. Ross Mayfield on September 15, 2003 7:03 PM writes...
Sounds like a great way to gather lots of meaningless data. Everyone will have different understandings of what each category of relationship means. What's worse, it fragments the liquidity of the network from the get-go.
Permalink to Comment