More than six degrees separate us
Perception and motivation influence social
networks.
8 August 2003
PHILIP BALL
An e-mail experiment has confirmed the famous 'six degrees of
separation' of human social networks, but revealed that individuals don't
necessarily benefit from their connectedness.
 |
| The original experiment was with posted
packages |
| © GettyImages |
|
|
The ease with which two people can hook up depends also on their
perceptions of social structures and their motivation to get connected, say
Duncan Watts of Columbia University, New York and his colleagues1.
This new study injects some psychology into a topic that has recently
become highly mathematical. It shows that we need "far more information on what
people know about their networks and how they use this knowledge during
searches," says social-networks pioneer Mark Granovetter of Stanford University
in California.
Omaha to Boston
The founding experiment in the science of social networks was performed
in 1967. Stanley Milgram of Harvard University asked randomly selected people
in Omaha, Nebraska, to send packages to a Boston stockbroker identified only by
his name, occupation and rough location.
The initial recipients were told to send the package to someone they
knew who might be in a better position to guide it to the addressee. People
guessed that it might take hundreds of these steps for the packages to reached
their destination. It took, on average, between five and seven.
This idea of 'six degrees of separation' between any two random
individuals has entered the cultural lexicon. The phrase forms the title of a
1990 play by the American John Guare. Interest in this small-world phenomenon
was revitalized in 1998 when Watts, together with Steven Strogatz at Cornell
University in Ithaca, New York, mathematically described a type of network that
could give rise to it.
In the new study, Watts' team invited 61,168 volunteers from 166
countries to recreate Milgram's experiment by e-mail. Targets ranged from an
Ivy League academic to an archive inspector in Estonia.
Of more than 24,000 chains started, only 384 found their target.
Successful chains were, on average, about four steps long, although this number
was biased by the greater likelihood of shorter chains being completed. A
typical chain length was indeed between five and seven, consistent with
Milgram's earlier findings.
But most e-mails did not reach their targets. This, it seems, is mainly
because of lack of interest from participants, rather than because of their
inability to think of an appropriate person to pass the message to.
The success of chains reaching one target - a professor at a well-known
US university - was by far the biggest. The researchers think that this was
because more than half of the volunteers were professional, college-educated
Americans, who probably perceived this target as more easily reachable than
others. In fact, his degree of connectedness was probably akin to that of the
other targets.
The exercise seems to shows that even if global social networks can be
searched quite easily, a searcher may not exploit this asset unless he realizes
the strength of his connectedness and has sufficient motive to make the effort.
|
