Joi Ito's Web

Joi Ito's conversation with the living web.

Khalid on #joiito pointed me to the following article.

New Scientist
Email experiment confirms six degrees of separation

Despite enabling almost instantaneous global communication, email appears not to have made the world a more close-knit community.

It's an interesting article about how an email six degrees experiment shows we are no closer than when Milgram did his famous experiment in 1967. (Milgram did an experiment which resulted in the assertion that we are only six hops away from anyone else in the world.) I referred to Milgram's famous experiment in my Emergent Democracy paper. When the paper was being reviewed by Shumpei Kumon, he referred me to Six Degrees by Duncan J. Watts and pointed out to me that Watts writes about Judith Kleinfeld who found that Milgram's experiment was flawed. I removed the reference in my paper. Milgram's six degrees experiment is so widely referenced that it has become almost an urban legend, but it DID NOT show that the world was connected by six degrees, it just got us thinking about it. I think the phenomenon is real and the "small-world problem" is a very interesting field, but people should stop quoting the Milgram study as fact. The email experiment referred to in the article is being conducted by Duncan Watts as well and he has a web page with more info.

Duncan Watts
The Psychologist Judith Kleinfeld stumbled onto what nows seems like a classic instance of such misplaced faith while she was teaching her undergraduate psychology class. [...] Remember that Milgram started his chains with roughly three hundred people, all of whom were trying to get their letters to a single target in Boston. The story everyone tells has the three hundred people living in Omaha, but a closer look reveals that one hundred actually lived in Boston! Furthermore, of the almost two hundred in Nebraska, only one-half were randomly selected. [...] The other half were blue-chip stock investors, and the target, of course, was a stockbroker. The famous six degrees is an average over these three populations, and as you might expect, the number of degrees varies quite a bit between them, with the Boston natives and the stock investors managing to complete chains more successfully and with fewer links than the random Nebraska sample.

Remember also that the surprising finding about the small-world claim is that anyone can reach anyone--not just people in the same town or people with strong common interests, but anyone anywhere. So really the only population that satisfied, even remotely, the conditions of the hypothesis as it is usually stated (even by Milgram himself) was the ninety-six people picked out of the Nebraska mailing list. At this point the numbers starts to get disturbingly small: of ninety-six starting letters in that population, only eighteen reach the target.


Urban Myths and FOAF-tales are Modern Legends. They live forever ...

I read about it in the Financial Times on friday. According to Mark Granovetter of Stanford it could help scientist to understand the spread of diseases and police to track down criminals.
Pretty neat!

In the 1960s when I worked as a photojournalist in Fleet Street, I used this kind of technique for impossible assignments. With no leads at all I would phone just anyone, at random if necessary, and ask a question. It got things started and worked surprisingly often. I even acquired a small reputation as a psychic of some sort.

Interesting article - most of the other articles considered the experiment a failure because of how few chains were completed.

I participated in this study - I live in NYC and was given a target in Moscow. but there was no way to track your chain to see if it had progressed.

They have just started a new study ( - you can sign up to be a participant or a target... and unlike the first study, you can now send as many emails as you would like at the top of your chain...better yet, you will soon be able to track the progress of your chain - which to my mind will be more interesting than whether you reach your target!

I also participated in the Columbia study and, actually, it was quite easy to trace the progress of my two chains. The sign-up proceedure furnished initiators with a target and password and URL specifically for tracking. It was easy to log into their site, type in my email address and password and read the succession of links. I reached a technician in Punjab in 5, my attempts to get to the policeman in Australia fizzled when the 3rd link never responded.

However, I got the news by telephone, before it actually was completed, before I would have gotten a chance to look it up: The third link called the second link to announce that his choice, the fourth contact, knew of the fellow's cousin and would be able to complete the chain! That news traveled back up the chain very quickly back to me.