[Air-l] From Social Isolation in America to Facebook

Barry Wellman wellman at chass.utoronto.ca
Sun Sep 10 13:29:04 PDT 2006

SPOILER WARNING: This posting will wind up with Networked Individualism.

The meaning of friendship has been preoccupying me -- and others -- more
than I would have expected. It started with the publication of McPherson,
Smith-Lovin, and Brashears' "Social Isolation in America" in Amer Soc
Rev, June 2006. This evoked a media panic that friendship was dying, based
on a shrinkage in the number of people "who discuss important matters"
from nearly 3 in 1984 to slightly more than 2 in 2005. (And nearly a
quarter of the American adult pop said they had no one.)

I spent a lot of time with the media for the next month pointing out that
research suggests that most very close ties are specialized, and that
"discussing important matters" wasn't the only way that people are close.
I noted that other surveys -- such as our Connected Lives study in Toronto
and the Pew Internet "Strength of Internet Ties" study have found a mean
of 10+ Very Close friends and relatives.

(These numbers probably seem high to non-North Americans. German and
French folk frequently tell me that they are astonished at how easily
Americans call someone "friend" when they, by contrast, take 5 or so years
to admit someone to their charmed cognitive friendship circle.)

I also pointed folks to Peter Bearman (& coauthor's) 2005 Social Forces'
article showing the variety of matters that people think are important to
discuss. While most people have their own construct of what are important
matters, they'd be mistaken to think that others share their zeitgeist. To
me, it is "world peace" (just like Miss Congeniality), the state of the
internet, and various family issues, but Bearman shows that to others it
might be what kind of haircut or tatoo to get, plus the usual boy/girl
friend issues.

This leads me to the Facebook fiasco.

First, as danah boyd says on her blog, a Facebook, MySpace, etc. "friend"
often is different from what most of us otherwise would consider a
"friend". Folks on the AOIR list reported mean numbers of greater than 100
Facebook "friends" in major American universities. By contrast, no survey
has shown such high numbers, even though they have shown larger numbers
than the 2 that McPherson found in analyzing the 2005 US General Social

One problem with almost all social software is that it makes two
false assumptions:
(a) It assumes that relationships are dichotomous: Friend/Non-Friend.
(b) It assumes that friends all belong to the same group.

In reality, friendships not only are specialized in content, but vary in
intensity. It's rare that someone would want to share everything with
almost any "friend".

Our data also show the obvious: it's rare that all Very Close friends (and
relatives) live in the same densely-knit group. The news that we want to
share with friends in one cluster of relationships may not be something we
would want to share with those in other clusters.

In short, we live in a world of "networked individualism" in which we're
forever navigating through complex social networks -- assessing who and
when to tell what.

Facebook got both of these wrong (in addition to foolishly not

It propogated all changes to all Facebook Friends.

Gosh, we could have saved them a bunch of money and reputation.


  Barry Wellman         Professor of Sociology        NetLab Director
  wellman at chass.utoronto.ca  http://www.chass.utoronto.ca/~wellman

  Centre for Urban & Community Studies          University of Toronto
  455 Spadina Avenue    Toronto Canada M5S 2G8    fax:+1-416-978-7162

      You're invited to visit & contribute to the new version of
 "Updating Cybertimes: It's Time to Bring Our Culture into Cyberspace"

More information about the Air-L mailing list