Thoughts, links, and essays from Peter Merholz
petermescellany   petermemes


Archives before June 13, 2001

RSS Feed

Adaptive Path (my company!)

About peterme

Most of the Time
Oakland, CA

American history around the time of the Revolution, figuring out how to marry top-down task-based information architecture processes with bottom-up document-based ones, finding a good dentist in San Francisco Oakland
Designing the user experience (interaction design, information architecture, user research, etc.), cognitive science, ice cream, films and film theory, girls, commuter bicycling, coffee, travel, theoretical physics for laypeople, single malt scotch, fresh salmon nigiri, hanging out, comics formalism, applied complexity theory, Krispy Kreme donuts.

Click to see where I wander.

Wish list
Show me you love me by
buying me things.

Track updates of this page with Spyonit. Clickee here.

[Editor's note: began as a site of self-published essays, a la Stating The Obvious. This evolved (or devolved) towards link lists and shorter thoughtpieces. These essays are getting a tad old, but have some good ideas.]
Reader Favorites
Interface Design Recommended Reading List
Whose "My" Is It Anyway?
Frames: Information Vs. Application

Interface Design
Web Development
Movie Reviews

Social Network Thoughtwander. Posted on 07/29/2002.

So, the interpersonal relationship stuff I wrote about really hit home how much of this internet thing is about communication. Making contact. Keeping in touch. And then yesterday I had coffee with Kevin Fox, who until recently worked on the design of Yahoo! Messenger (though nothing you've seen yet... All still very much in development). And that got me thinking about the mobile phone buddy lists that Katelyn McKenna discussed, which show not only which of your friends is "at their phone," but also where they are. And all of this social connectivity seems ripe for social network analysis depictions of the kind Valdis creates. To the best of my knowledge, though, such diagrams are only used to assess the state of a situation, not as tools or interfaces for interacting with those people. But I think that such visualization could work to help me make my social network more manifest, and let me act on it.

I mean, I'd love it if I could dump my email address book into a social network software tool... I hate looking at my contacts in long lists. It would be great to be able to group them in that fashion... And also have them regroup themselves based on attributes of interest at the time (industry, region, etc.)

Is this one of those applications for the lazyweb? Considering the powerful desire we have to use these tools to connect with one another, and the fact much of this understanding of the connections we have is done through information-poor means such as lists or, usually, in your head, I would think that a tool that successfully demonstrated your connections to others, and their connections to others, and the status of those people (think of one of those social diagram programs, where next to every name, you'd know if they were on instant messenger, or had their cell phones turned on, or, where they were at that moment), could be quite popular, in that it would support your efforts in being connected.

Yes, there are boatloads of privacy issues that would need to be dealt with. Sigh. I tend to think that issues around "privacy" and "security" are scare tactics drummed up by The Media in order to increase sales.

15 comments so far. Add a comment.

Previous entry: "You know, I've never had a milkshake."
Next entry: "Soon I'll be one of those folks who peers at the outside world through parted slats in the window blinds."


It sounds like what you're suggesting is a kind of souped-up address book. Which is not at all a bad idea, but I think the bigger potential for social network software has less to do with personal productivity tools and more to do with mining for macro patterns of social interaction that stretch beyond the immediate sphere of any one end-user.

Imagine if you could visualize the flow of communication within a large organization - say, by exploring patterns of email traffic, calendars, maybe mixed in with some level of linguistic analysis. By looking for early warning signs of changes in the way people communicate and collaborate, a company might be able to identify implicit work groups that haven't yet been made explicit in the organization (which could also be a privacy advocate's worst nightmare, and possibly a CIO's dream?)

Rick Rashid, VP of Microsoft Research, touched on some of these themes at the OReilly conference in his presentation on the "Future of Operating Systems" (here's a link to an audiocast of his presentation).
Posted by alex wright @ 07/29/2002 05:55 PM PST [link to this comment]

A souped up address book is just the start. Imagine being able to look at all your contacts, arranged into groups, connected by relationships ... and recognising that some of the people you know don't know the others. You're in position to introduce them to each other, or maybe even discover a relationship you didn't know about.

This is especially so if you could share your social network maps with your contacts. The old website provided something similar, although flatter in structure measuring only 'distance' (a la "six degrees of kevin bacon"). Then they jumped on the consumer portal bandwagon and went over a cliff.
Posted by Eric Scheid @ 07/29/2002 06:58 PM PST [link to this comment]

Funnily enough I was chatting with one of my friends the other day, where this kind of "thebrain" meets "contact book" and "buddy list" is maybe the sole application I could imagine that could make good use of speedy VR. Documents and relationsships flying around in sheer mayhem. Would be killer app.

Anyone ?
Posted by Frederik Andersen @ 07/30/2002 03:34 AM PST [link to this comment]

Peter --

The system that I've heard of that sounds like it comes closest to this is Bonnie Nardi's ContactMap, see a CACM paper on it at

However I have not used it myself! And just from the screenshots it appears to be limited to the sorts of 25-75 person working groups where you can fit everyone's face onto a screen, not the totally unwieldy 150+ groups that characterize the net.


Posted by Edward Vielmetti @ 07/30/2002 05:12 AM PST [link to this comment]

To Eric's point -- that's exactly where I wanted to go with this. Let's say I'm in a new city. It would be great if my mobile phone buddy list not only told me where people I knew were, but where people two degrees from me were.... And if one was in this new place, and I was feeling nervy, I could contact that person and say, "Hey, I'm a friend of John, and I'm in , and don't really know what's going on."

ContactMap looks interesting, but seems to be only for folks in AT&T.
Posted by peterme @ 07/30/2002 07:07 AM PST [link to this comment]

The possibilities are endless. Here's two that immediately pop to mind:

1) Imagine if you're trying to get a job at company X, and you're at some networking-type event, and there's a recruiter from company X there. But, there could be hundreds or thousands of people, and you can't go up to each individually and ask them where they work or read their nametag. Now, they don't make their information publicly available on this wireless social network (or else they'd get swarms of job hunters attacking them), but your former co-worker's brother works in a different department in company X. Through those social connections you are "granted" permission to access the recruiter's location and can find out which person they are so you can approach them and talk to them about the job, giving you an advantage over anyone without social connections.

2) You're walking down the street one day and see someone that you know but can't remember their name. You know you went to college with them, but as they approach you, you draw a blank. You look up your freshman year roommate on your cellphone, and remember that he has a connection to his ex-girlfriend whose older brother was on the soccer team with this guy. You input this information and it returns to you his name, marital status, occupation, and interests (information he has made available to anyone looking him up on the "public" network), sparing the embarrasment of having to admit you have no idea who he is.

It would be like a mobile myGoogle on 'roids. Serious privacy issues, but fun to think about.
Posted by jefflash @ 07/30/2002 09:29 AM PST [link to this comment]

Following Jeffs #1, remember that the Japanese have this dating gizmo: you set the switches to say you are (say) Male, looking for Female, looking for non-committal fun. You then put it in your pocket and walk about ... if it detects a complementary gizmo it beeps/throbs/whatever, and you take a look around to see if anyone nearby looks tasty to you.

Take this one step further: instead of the gizmo only broadcasting three or four bits of data, have it broadcast your public PGP key, and then also broadcast/swap the public PGP keys of your friends. If a corresponding gizmo recognises any of those keys it could pipe up and you have a new friend of a friend.

You could even prevent promiscuous catch & release of latent PGP keys (and thus impersonating others) by having your friend PGP sign your key, thus confirming the relationship.
Posted by Eric Scheid @ 07/30/2002 10:05 AM PST [link to this comment]

Hm, just thinking about this a bit more -- Ryze does a certain amount of this as far as helping out the friend-of-a-friend problem, though it doesn't currently hook up with any kind of instant message are-you-online tool.

Posted by Edward Vielmetti @ 07/30/2002 01:05 PM PST [link to this comment]

A couple of responses...

Contact Network... where would it run? on the WWW or on your laptop? or on your wireless device? Got to consider screen real estate... InFlow could be easily adapted for the laptop version.

Alex's comment on discovering work groups... we have been mapping emergent work groups/communities in organizations since 1990. IBM uses InFlow to discover hidden Communities of Practice and link together those of similar interests.

Frederik's comment on nodes as people or docs... the nodes could be any object with an address -- click and go.

Obviously with wireless tools we would not be doing much 'network visualization', but a small db could keep track of links and a few simple algorithms to calculate distance/hops between people.

What if your network looks like Erdos' collaboration network? How many contacts would you show in your network? How many contacts of your contacts would you show?
Posted by Valdis @ 07/30/2002 02:00 PM PST [link to this comment]

I implemented something like this on a wiki, with one contact info page per person. A group can be created by simply making a new page and linking to people who belong in the group. Backlinks let you know what groups someone is in. I use it for tracking my relatives' email addresses. Zero programming involved, I just downloaded UseMod.

I thought if something like this became popular, it could be a great way to find people you have lost contact with. Even if you don't remember their name, someone else might have added them in a group that you recall having been part of.

Of course it would be very useful to have openly accessible social network data for politicians, CEOs, etc., in order to better understand the moves being made.

Also, Cameron's Social Network Explorer ought to be mentioned somewhere in this discussion, so here goes the link: (Scripting News' social network)
Posted by Seb @ 07/30/2002 08:18 PM PST [link to this comment]

Sebastion wrote...
" would be very useful to have openly accessible social network data for politicians, CEOs, etc., in order to better understand the moves being made."

Yes! Much of this data is publicly available... the hijacker network was created from 100% open source data.

Another person is using InFlow to map and measure the links between all of the ENRON players.

IMHO, all investigative journalists should learn basic social network mapping to illustrate their stories and discover the real power positions within the human networks they are writing about.

Posted by Valdis @ 07/31/2002 07:48 AM PST [link to this comment]

Sebastion wrote...
" would be very useful to have openly accessible social network data for politicians, CEOs, etc., in order to better understand the moves being made."

ob-link: They Rule
Posted by Eric Scheid @ 07/31/2002 12:06 PM PST [link to this comment]

Which reminds me of...
Columbia Journal Review, tracking media conglomerates.
Posted by Seb @ 07/31/2002 05:46 PM PST [link to this comment]

Another interesting riff would be the the ability to group people together on the fly, say put all of your friends who like movies (this would of course be in their profile), together into a group. You could then message the group - this would be sent to their IM or email or phone, depending on their location and preferences.

Say if you had found something about the new David Fincher film, you could send out a message to this group, creating a sort of mini-newsgroup (or specialized micro-blog), to which everyone could then add their own thoughts, responses. If someone sends the message on to their own friends, then those second-generation message recipients might still be referred back to the original thread, creating a whole new discussion group that you were previously unaware of.

I think, as far as security and privacy go, it wouldn't be too hard to simply base it off of personal preferences. I.e., if you are on IM then everyone knows you're there, but if you only have your phone, only your immediate friends need to know you're still on-line. Similarly, maybe I don't mind broadcasting my propensity for Fincher films out to the wide world, but only a handful of people need to know about my unnatural obsession with spatulas.
Posted by Justin @ 08/01/2002 09:27 AM PST [link to this comment]

TouchGraph: open-source network graphing software (Java) --
Posted by erm @ 08/06/2002 08:00 AM PST [link to this comment]

Add A New Comment:


E-Mail (optional)

Homepage (optional)

Comments Now with a bigger box for text entry! Whee!

All contents of are © 1998 - 2002 Peter Merholz.