The search for the Tara

By MIKE WILSON

© St. Petersburg Times, published June 30, 2000

Can a small group of e-mail users get a message to a complete stranger even if they have only her first name and a few sketchy biographical details? Our experiment aims to find out. Click here for follow-up story

Three decades ago, social psychologist Stanley Milgram conducted an experiment to see how connected we are.

Milgram mailed a packet containing the name and address of a stockbroker in Boston to more than 100 people living in the Midwest. He asked the Midwesterners to send the packet to the stockbroker if they knew him. If they did not know him, they were to move the packet closer to its destination. If they had cousins in Boston, for example, they might forward it to them.

How many hands would the packet have to go through to get to the stockbroker? Hundreds? Thousands?

The astonishing answer: Six, on average.

E-mail into the void As we prepared for the "Where's the Tara?" experiment, we decided to try a smaller experiment as a trial run. The focus this time was forwarded messages, the collections of jokes and other trivia that people are forever passing around the Internet.

From this experiment came the phrase "six degrees of separation," the idea that a person can be connected to any other through six acquaintances or fewer. (These days the phrase is often heard as "six degrees of Kevin Bacon," in which movie buffs connect Bacon to other actors.)

We wondered if there was a way to conduct a version of Milgram's experiment in the Internet age. Is anything new to be learned about our connectedness, given all this expensive technology?

That is how we came up with our e-mail experiment: The Search for the Tara.

A weeklong look at everyday life with the Internet. Stories

This morning we are sending an e-mail to a half-dozen people who have agreed to take part. These six people are the Internet-era equivalent of Milgram's 100 or so Midwesterners.

Our note tells them about someone we met recently while visiting another state. Her name is Tara, and she is, of course, our stand-in for Milgram's Boston stockbroker.

But in our experiment there is a twist. Instead of giving a name and address, as Milgram did, our e-mail supplies only Tara's first name and a few biographical details: approximate age, town in which she grew up, marital status, the U.S. region she lives in, the industry in which she works and some family background.

These details are not specific enough to make her easily identifiable. Our six e-mail recipients have no connection to her that we know of.

Their goal is simple: get an e-mail to The Tara. (She's not just any Tara, she's The Tara, the one in our experiment.)

If they recognize her from the few details we give, they are supposed to send an e-mail asking if she is indeed The Tara.

If they don't know her, they should move our e-mail closer to its target, just as Milgram's people did with his packet. They might send it to someone in her hometown, or someone in her industry, or someone who went to the same college she attended.

Of course, some participants may decide to forward our e-mail to everyone in their address books. That's up to them. We are asking them, though, not to play detective. Finding The Tara that way would skew the results.

Finally, we are asking everyone who receives our message to let us know what they did with it. This will make it easier for us to track its progress. Milgram's subjects had a packet and a pencil; ours have e-mail and a reply button.

When our target receives an e-mail from someone asking if she is, indeed, The Tara, she will let us know, and we will work backward to see where the e-mail came from and how long it took to arrive.

Every experiment begins with a hypothesis. This is ours:

The six original recipients will forward our e-mail to several people each, who will do the same, meaning the message will spread rapidly and exponentially. It won't take long, perhaps less than a week, for someone to send Tara an e-mail asking if she is The Tara.

Why do we think this will this happen so fast? Well, we have no reason to believe people are any more closely intertwined today than they were before the Internet. E-mail, we think, has not reduced the six degrees of separation to three or four. Our social circles are about the same size they always were.

The Tara will get a message quickly, we think, because e-mail allows us to transmit information to large numbers of people with incredible speed. Think of the lists of dark jokes that flash across the Net after any tragedy. Think of the creepy urban myths that have taken hold through sheer ubiquity.

Our e-mail is going to zip toward The Tara in an eyeblink. (For evidence that we are dead wrong about that, see related story.)

We are aware that when you start an e-mail chain, there's no telling where it will end, and that worries us. That is why our e-mail explains that we will publish the results of our experiment between now and Aug. 13 (on the theory that we should know something inside six weeks). Folks can also check at http://www.sptimes.com/experiment. When we publish a story with the results, this page will link to it.

Everyone who gets our e-mail will see that there is no point looking for The Tara after Aug. 13 because she will long since have been found.

That's the idea, anyway.

We asked The Tara what she thinks will happen.

Her best guess was that it will take a week for our e-mail to reach her. She shares our view that e-mails bounce around like atoms and believes the message will end up in her computer because of sheer volume.

On the other hand, The Tara thinks our message will have to pass through more than six hands before it gets to her. She does not consider herself extremely well-connected socially. She isn't one of those people who knows everybody.

So how many people will touch the message before she gets it?

"I would guess between 10 and 15 people," she said. "But what do I know?"

Nothing. Same as us.

© St. Petersburg Times. All rights reserved.