Judith Kleinfeld's paper

From:
Date: Mon Jul 15 2002 - 17:12:34 EDT


As mentioned in Barabasi's "Linked", Judith Kleinfeld's paper has been
published. See a related article in Discover magazine.

http://www.discover.com/june_02/featworks.htmlSix Degrees of
Speculation
  Even in a small world, there's room for disagreement
  By Karen Wright

  You probably don't know Judith Kleinfeld. She's a psychologist at
  the University of Alaska in
  Fairbanks, and you can contact her by calling the university
  switchboard or finding her e-mail
  address online. But could you get to her through some extension of
  your own social
  network-by mailing a letter for her to a friend who might know
  someone who knows someone
  who knows her?

  Who cares?

  Judith Kleinfeld does, for starters. She's part of a
  growing cadre of scientists reviving the so-called
  small-world problem, a social-cum-mathematical
  conundrum formulated in the last century to
  characterize the interwoven webs of acquaintance
  among friends, neighbors, colleagues, and kin. In the
  mid-1960s the legendary social psychologist Stanley
  Milgram asked randomly selected citizens of Kansas
  and Nebraska to try to connect with social "targets" in
  Massachusetts by mailing letters to likely
  intermediaries. The average number of links between
  strangers turned out to be surprisingly small. Milgram
  claimed we're all connected, on average, by half a
  dozen interpersonal avenues- a numinous network
  popularized by the phrase "six degrees of separation."

  Now some mathematicians say the dynamics of such
  networks can describe a host of natural and
  technological systems. But Kleinfeld, who has
  reviewed the Milgram archive at Yale University, says the small
  world may have no basis in
  fact. "It's not a robust social-science finding," she asserts. "It's
  a very odd one that has
  never been replicated."

  The original small-world question was posed in the 1960s by
  political scientist Ithiel de Sola
  Pool of the Massachusetts Institute of Technology and IBM
  mathematician Manfred Kochen.
  How many acquaintances, they wondered, connect any two people chosen
  at random? Kochen
  and Pool's calculations put the average number at three. Milgram
  devised an experiment to
  test that surprising prediction. He recruited volunteer "starters"
  in Kansas and Nebraska
  from mailing lists and newspaper ads and told them the name,
  address, and occupation of a
  target in Massachusetts. He asked the starters to forward a document
  to someone they knew
  on a first-name basis who might be able to reach the target. In a
  1967 report of his results,
  Milgram described how one folder had found its way to the wife of a
  divinity student in just
  four days and three steps: from a Kansas wheat farmer to an
  Episcopal minister, from the
  minister to a colleague in Cambridge, and thence to the divinity
  student's wife. In that case,
  Milgram wrote, "The number of intermediate links between starting
  person and target person
  amounted to two!" The average number of intermediates was five- six
  steps, in other words,
  connected Milgram's distant strangers.

  As a graduate student in the late 1960s at Harvard, where Milgram
  worked from 1963 to
  1967, Kleinfeld had known of Milgram's studies. But it wasn't until
  the 1990s that she
  decided to try to repeat his study as an exercise for her graduate
  students. When she went to
  the Yale archives to find out more about his methods, Kleinfeld saw
  that the rates of
  completion in Milgram's studies were lower than she'd realized. In
  the main study, using
  almost 300 starters, only 29 percent of the documents reached a
  Boston stockbroker- and
  100 of the Nebraska starters owned blue-chip stocks. In a pilot
  study- the one that yielded
  the anecdote of the divinity student's wife- just 5 percent of 60
  documents reached the
  target, and they passed through an average of eight people. And
  Kleinfeld couldn't find any
  exact replications of Milgram's experiments in the literature.

  "I was really upset and discouraged," she says. "I thought, 'Oh,
  Stanley, how could you do
  this?'"

  Milgram, whose controversial studies of obedience to authority
  earned him celebrity, died in
  1984. But his small-world concept still has its defenders. "It's a
  solid phenomenon,"
  maintains Thomas Blass, a Milgram scholar at the University of
  Maryland, Baltimore County,
  who says Kleinfeld's criticism is overblown. Many investigators
  don't report the results of
  pilot studies, Blass says. And Milgram's low completion rates don't
  necessarily mean that
  his participants hit dead ends. "It's as likely that the low
  completion rates are because people
  just lost interest. If that's the case, it doesn't damage the
  small-world phenomenon."

  Support for Milgram's idea has come from another camp as well. In
  the mid-1990s,
  mathematician Steve Strogatz of Cornell University and his graduate
  student Duncan Watts
  began constructing theoretical models that explained how millions of
  people could be linked by
  relatively short social routes. Watts and Strogatz's models show
  that the members of a big
  network can be connected by short paths if the network comprises
  clusters of close
  associates joined by occasional, far-reaching links- "random"
  elements in otherwise
  structured groups.

  To test whether such networks actually exist, Watts consulted one of
  the few impeccably
  documented social circles known to science: the parlor game Six
  Degrees of Kevin Bacon. The
  game, housed online by the Oracle of Bacon at Virginia Web site and
  supported by the vast
  Internet Movie Database, involves finding the shortest route between
  any given thespian and
  actor Kevin Bacon. Links are forged by appearing in a movie with
  Bacon, or with someone
  who has appeared with Bacon, or with someone who has appeared with
  someone who has
  appeared with Bacon, and so on. Among the hundreds of thousands of
  actors living and dead,
  the average "Bacon number" is 2.918, and no actor has been found to
  be more than 10 degrees
  of separation from Bacon.

  Watts learned that the Bacon network conforms to his models of a
  small world: Groups of
  closely associated actors are joined by random "linchpins" that link
  far-flung clusters.
  Linchpin actors may have long careers, like Green Acres's Eddie
  Albert, thus joining clusters
  in time. They may link the films of different cultures, as Bruce Lee
  does. Or they may span
  genres, like Nicolas Cage (comedy and suspense). There's reason to
  believe that ordinary
  social networks operate according to similar principles, says
  Watts. And he's found the same
  kind of connectivity in systems ranging from the neural wiring of a
  nematode to the electrical
  power grid of the western United States. Small-world networks may
  also describe the spread
  of disease, fads, and social movements.

  "Even just a little bit of randomness in any network is going to
  generate short paths between
  different nodes," says Watts. "Milgram showed that people can
  actually find those paths-
  and that's a different sort of problem. The models we developed to
  show that these short
  paths actually exist don't explain how you can find them."

  How people recognize the routes to their distant affiliates is the
  next big small-world
  question. Computer scientist Jon Kleinberg at Cornell has proposed
  an algorithm to explain it
  (although Milgram himself conceded that "there is nothing more alien
  to mathematically
  untutored intuition than this form of thinking"). Watts is trying to
  trace both the reasoning and
  the routes through an e-mail replication of Milgram's
  experiment. He's got 50,000
  participants so far, with targets in India, Europe, Indonesia, and
  Australia as well as the
  United States. "We've already had some globe-crossing chains- from
  Sydney to Siberia in
  four links," he says. "It can be done."

  Kleinfeld thinks an e-mail experiment misses the point, because
  computers link people who
  already have a lot in common- such as income and education. "I think
  scientists live in a
  small world," says Kleinfeld. "People who are low-income or
  minorities aren't connected."

  Watts concedes that the e-mail network is a practical
  compromise. "In terms of getting very
  large numbers of people [to participate], nothing can beat the Web,"
  he says.

  Blass gives Milgram the final word. "Milgram would be very happy
  about it," he claims,
  "because he was one of the first people to appreciate the
  communicative value of PCs."

  RELATED WEB SITES:

  Kleinfeld, Judith S. "Could It Be a Big World After All?"
  Forthcoming in Society (2002). See
  www.uaf.edu/northern/big_world.html.

  The Oracle of Bacon at Virginia Web site:
  www.cs.virginia.edu/oracle. Sign up for Duncan
  Watts's e-mail experiment: http://smallworld
  .sociology.columbia.edu/target.html. Also see
  Watts, Duncan J. Small Worlds: The Dynamics of Networks Between
  Order and Randomness.
  New Jersey: Princeton University Press, 1999.



This archive was generated by hypermail 2b30 : Tue Jun 09 2009 - 05:00:06 EDT