I just got back from another conference on the future of news, where many cool thoughts were exchanged that will find their way to this blog in due time.
Stephenson studies social networks. She goes into a company–her clients include J.P. Morgan, the Los Angeles Police Department, T.R.W., and I.B.M.–and distributes a questionnaire to its employees, asking about which people they have contact with. Whom do you like to spend time with? Whom do you talk to about new ideas? Where do you go to get expert advice? Every name in the company becomes a dot on a graph, and Stephenson draws lines between all those who have regular contact with each other. Stephenson likens her graphs to X-rays, and her role to that of a radiologist. What she’s depicting is the firm’s invisible inner mechanisms, the relationships and networks and patterns of trust that arise as people work together over time, and that are hidden beneath the organization chart. Once, for example, Stephenson was doing an “X-ray” of a Head Start organization. The agency was mostly female, and when Stephenson analyzed her networks she found that new hires and male staffers were profoundly isolated, communicating with the rest of the organization through only a handful of women. “I looked at tenure in the organization, office ties, demographic data. I couldn’t see what tied the women together, and why the men were talking only to these women,” Stephenson recalls. “Nor could the president of the organization. She gave me a couple of ideas. She said, `Sorry I can’t figure it out.’ Finally, she asked me to read the names again, and I could hear her stop, and she said, `My God, I know what it is. All those women are smokers.'” The X- ray revealed that the men–locked out of the formal power structure of the organization–were trying to gain access and influence by hanging out in the smoking area with some of the more senior women.
This fascinated me because I’m beginning to take a serious interest in Internet trust currencies — everything from eBay trusted merchants to the LinkFilter system of hits and points.
The other day, a poster on the MetaFilter ombudsite MetaTalk suggested a complicated post rating system founded on the principles of battle in online role-playing games:
Metafilter hitpoints! We all get 5000 to start. Once you level up via unattacked thread posting, you can cast healing spells on your favorite, but inexplicably hated MeFi pals, or do double damage with Fireballs. Anybody who reaches zero has their account closed, unless someone ells resurrects you by sacrificing 3/4 of their remaining points.
It inspired a long thread of quality snark.
But there might be a journalism-related nugget in here. I was in a small group session with Jeff Jarvis where we came up with a model for a future news organization that highly resembles some of Robin and my EPIC prototypes from early 2004. (Karen Stephenson, Andreas Neus and I are three of the folks whose names Jeff Jarvis has forgotten in the past 48 hours. Sad!)
One of the four planks of our news model was this idea of trust aggregation:
Let’s say that five people cover the school board. Whom do you trust? It might be the one with the most links, or the most positive reviews, or the most traffic, or the most experience, or the fewest corrections and complaints, or the one who has the contempt of the people in power you hate, or perhaps training, or even editing. It may also be the reporter — staff or independent — who is the most transparent, who tells you how she votes so you can judge her reporting. Trust is your decision. We report; you decide.
The model Robin and I were batting around was a little better, I think, though more complicated. Who has the time to go around picking every news source they trust or don’t trust? And ‘sort by corrections’ seems to lack nuance. Ours was a distributed trust system, involving the weighting of trust (or influence, I’d say) — if I like your stuff, then those whom you like are rated-up accordingly in this system. Anonymous sources became losers in our media environment because without a trusted identity to trade on, they don’t make it into many stories.
I imagine in our system one could also sort by corrections.
But the MetaTalk post inspires me to think there might be even more imaginative trust structures out there we could learn from. Who might be the smokers in EPIC’s trust ecology?