The UndernetOrganizations often con

The Undernet
Organizations often conclude that knowledge sharing does not occur
because no one is using the organizational knowledge repository. Knowledge
sharing may in fact be occurring. Often employees simply create their own networks
instead of going through an official or formal organization-wide
network. KM succeeds when it is a grassroots or demand-driven initiative
rather than a top-down technology push.
Knowledge flows well when members perceive that there is a climate of trust,
that the members with whom they exchange knowledge are credible, and that
knowledge exchange is bidirectional. In small organizations, these undernets
bring different specialties together, such as engineering, design, and marketing.
But in larger organizations these specialties tend to separate into their own
groups. When that happens, the communities develop different ways of
working and even adopt different vocabularies, and they no longer understand
each other. Knowledge still flows easily within specialties but not across them
(Seely Brown, 2002).
Social network analysis is a very useful tool, for it provides the means of
identifying the “undernets” in an organization (Weinberger, 1999). The undernet
is defined as the intranets that escape the official gaze of the organization.
They represent how people really share knowledge, and they constitute the
KNOWLEDGE SHARING AND COMMUNITIES OF PRACTICE 133
skeleton of the communities of practice that have emerged. Weinberger aptly
refers to these undernets as the “lifeblood” of the organization. In fact, many
corporate top-down knowledge management initiatives are met with lack of
interest and lack of activity, and investigation invariably turns up the existence
of the “other” network—the one people really use!
The undernet is often referred to as KM’s dirty little secret: however much
you invest in high-tech knowledge banks, employees in search of an answer
tend to make their first port of call the folks they know from the water cooler.
An example is the Kraken vs. KnowledgeCurve system at Pricewaterhouse-
Coopers (PWC) as discussed in an interview with Julia Collins (CIO Magazine,
March 1, 2001, available from http://www.cio.com/archive/030101/
passport_colins.html). Although PWC has considerable investment in formal
knowledge management databases, called KnowledgeCurve, the Kraken is an
informal and unofficial Lotus Notes e-mail list that has been garnering more
attention lately. Named after a mythological sea monster in a poem by Alfred
Lord Tennyson, the Kraken is a sort of global glue, sharing knowledge across
national borders. Kraken is much less sophisticated as a system goes—just
e-mail—so what is the secret of its success?
Knowledge sharing in Kraken is a manifestation of a community of practice.
In fact, 80% of the messages in Kraken begin with a question: Does
anybody know? Has anybody ever done . . . ? Such questions often result in
four- to five-page responses. This is knowledge sharing among professionals
with concrete decisions to make and problems to solve. In order to do so, they
need to connect to their peers, and the undernet is the result of their connections.
Ideally, such grassroots or bottom-up knowledge systems should be
accommodated by the organization-wide systems. Knowledge brokers are individuals
who are able to move among more than one network, and they can
play a key role in putting together a company’s “big picture.” Formal, topdown
KM systems tend to encapsulate more formal, explicit knowledge,
whereas community networks tend to be less formal and more tacit and to
have more “work in progress” content. Ellen Knapp, PWC’s Chief Knowledge
Officer, puts it this way: “KnowledgeCurve is about teaching. Kraken is about
learning. You can’t have one without the other” (Stewart, 2000).