Online Environments for Supporting Knowledge Management and its Social Context*
Thomas Erickson and Wendy A. Kellogg
IBM. T.J. Watson Research Center
The issue of how to support the re-use of knowledge under rubrics such
as organizational memory, knowledge management and expertise management
has received increasing attention over the last decade. In this chapter we take
a strongly social approach to the issue, arguing that knowledge (and expertise)
is created, used, and disseminated in ways that are inextricably entwined with
the social milieu, and therefore that systems which attempt to support these
processes must take social factors into account.
Our approach to managing knowledge or expertise is to do it on-line, via multi-user
networked environments that support group communication and collaboration. That
is, we are interested in designing on-line environments within which users can
engage socially with one another, and, in the process, discover, develop, evolve,
and explicate knowledge relevant to shared projects and goals. We refer to online
multi-user environments used in these ways as "knowledge communities."
This chapter consists of two sections. In the first section we make the case
for a deeply social approach to knowledge management. We begin with an example
that depicts a number of ways in which the production and use of knowledge is
fundamentally entwined with social phenomena. We note that this socially situated
view of knowledge is supported by research in a number of disciplines, and also
has made its way into the business discourse that surrounds knowledge management.
This view raises a challenge for those designing technology: knowledge management
systems must take into account, either explicitly or implicitly, the social
context within which knowledge is produced and consumed.
In the second part of the chapter we argue that one way of addressing this
challenge is via the sorts of online multi-user systems that we call knowledge
communities. We describe some examples of systems that currently function as
knowledge communities and then turn to our own work on designing infrastructures
for knowledge communities. Our general approach is to design online environments
that, by making users and their activities visible to one another, can enable
a variety of social phenomena that support social and work-oriented interaction.
We describe a system called "Babble," which we have designed, implemented,
and deployed to about twenty workgroups over the last four years. We report
on our experience with Babble, and conclude by discussing some of the general
issues we see for designing online environments that support a socially-oriented
approach to the management of knowledge and expertise.
2. Knowledge Work as Social Work
Knowledge management is often seen as an information problem: how to capture,
organize, and retrieve information. Given this perspective, it isn't surprising
that knowledge management evokes notions of data mining and text clustering
and databases and documents. This is not wrong, but it is only part of the picture.
We suggest that knowledge management is not just an information problem, but
that it is, as well, a social problem.
2.1 An Example
One of us once interviewed accountants at a large accounting and consulting
firm about their information usage practices. The goal was to find out how they
thought they would use a proposed database of their company's internal documents.
In the course of the investigation, an unexpected theme emerged: the accountants
said that one of the ways in which they wanted to use the documents was as a
means of locating people. The accountants' claim that they wanted to use
a document retrieval system to find people was, at the time,
quite surprising. However, in the course of further interviews, it came to make
sense: It was only through the people that the accountants could get some of
the knowledge they needed. As one accountant explained, 'Well, if I'm putting
together a proposal for Exxon, I really want to talk to people who've already
worked with them: they'll know the politics and the history, and they can introduce
me to their contacts. None of that gets into reports!'
For our purposes, there are five important points here. First, as the accountants
observed, some types of knowledge tend not to get written down. Sometimes it
may be that the knowledge is too politically sensitive: people shy away from
recording gossip and innuendo, even though knowledge of it may be very helpful
to someone about to do business in that environment. Sometimes knowledge
in the form of comments, opinions, or conjectures may not be written
down because the resulting records can be potentially be subpoenaed. And sometimes
knowledge that may seem too trivial to be recorded when first encountered
that the CEO is a teetotaler or a Scotch fancier can prove quite valuable
in the process of establishing a relationship. Because this knowledge is often
quite useful for social purposes, we will refer to it by the rubric "social
The second point is that the accountants were not just tapping into social
knowledge; they were also getting access to social resources such as
contacts and referrals. One accountant explained that the worst way to approach
a company with a proposal was by making a "cold call". It is much
better if the accountant, let us call him Charles, can begin a call to a new
contact by saying My colleague, Jil Smith, suggested I chat with you.
Being able to say that one has been referred by a mutual acquaintance is a frequent
and powerful facilitator for interpersonal interaction and this is true
even if the relationship is only a few hours old. Charles, by virtue of having
permission to assert a relationship with Jil, can draw on to some extent
Jils reputation and standing with the person with whom he is trying
to open negotiations. Notice, by the way, that social resources cant be
extracted from a person and embedded in a database: opening the conversation
by saying I found your name in the corporate knowledge base isnt
the same as saying Jil Smith said I should call.
The third point we take from this example is that people dont necessarily
need access to an "expert." It may be that Jil Smith has had only
one previous engagement with Exxon, and that, in terms of facts, she may have
far less expertise than an outside consultant. Nevertheless, Jils experience
may be sufficient to provide Charles with the social knowledge and social resources
necessary to gain entry into the Exxon environment. In fact, it may be preferable
for Charles to talk with Jil, because, as a colleague who shares the same work
context, she will understand more about what he needs to know, the situations
in which he will use the knowledge, and how he is likely to go about using it,
than someone traditionally construed as an expert. That is, Jil has social and
contextual expertise, in contrast to an outside consultants factual expertise.
The fourth point we take from this example is implicit in the previous ones:
networks of personal relationships, which are created and reinforced through
interpersonal conversation, are critical in supporting knowledge sharing. Let
us return to the example of Jil and Charles. Assuming that Jil's assistance
was helpful, Charles has now accrued a small debt or obligation to Jil. When
Jil needs assistance, she is likely, in turn, to come to Charles with questions
or requests for social knowledge that falls within his domain. Even if the required
information is outside of his domain, she may seek to obtain access to his social
resources a referral to one of his contacts, for example. Thus are professional
relationships established, and thus do social networks grow. In the long run,
if not the short, it may be more valuable for an enterprise if its members seek
knowledge and social resources from one another thus building a web of
mutual knowledge and trusted relationships than if, for instance, employees
are given instant access to a top-notch external domain expert.
This brings us to our final point, which has to do with the centrality and
importance of conversation in knowledge sharing (see Fitzpatrick, this volume).
It is no coincidence that both social knowledge and social resources are best
shared through talk. It is the time spent discussing apparently trivial social
knowledge that suggests that a relationship goes beyond the purely professional
that there is more in play than just a purely instrumental professional
exchange. It is the disclosure of politically sensitive information that indicates
a degree of trust between two people. It is the ability of one person to take
generic information and apply it on the fly to the other's problem
that increases the reputation of the giver and creates an obligation for the
receiver. This sort of talk and the exchange of knowledge and social
resources it involves both requires and strengthens networks of personal
relationships in workplace.
2.2 The Social Construction of Knowledge
This sort of situation is not the exception, it is the rule.
A wide variety of research programs for instance, ethnographies of workplaces,
social studies of science, critical theory, organizational memory, the sociology
of knowledge point to the deep connections between knowledge management
and social context.
For example, ethnographic studies of workplaces reveal a
wide array of social practices implicated in the production and dissemination
of knowledge. Lave and Wenger have developed the notion of a community of practice.
They note that one way in which people come to master a body of knowledge is
through a sort of apprenticeship or "legitimate peripheral participation"
in the activities of a group of practitioners (Lave and Wenger, 1991). Wenger
(1998) describes the daily work in an insurance claims processing office, and
shows how it is entwined with social relationships and processes. Similarly,
in an ethnography of copier service technicians, Orr (1996) reveals that technical
knowledge is socially distributed across a network of technicians, and that
it is tapped into and disseminated through oral processes such as storytelling.
A similar sense of the social nature of the production and
dissemination of knowledge comes from the field of social studies of science
see Latour and Woolgar (1979) and Latour (1987). For example, Traweek's
ethnography of particle physicists (1988) examines some of the social phenomena
that structure the practice of high energy physics. She notes the impact of
social relationships on the placement of graduate students, the evaluation of
experiments, and access to equipment and facilities. Her comments on the role
of conversation are particularly interesting:
"...talk accomplishes diverse tasks for physicists: it creates,
defines, and maintains the boundaries of this dispersed but close-knit community;
it is a device for establishing, expressing, and manipulating relationships
in networks; it determines the fluctuating reputations of physicists, data,
detectors, and ideas; it articulates and affirms the shared moral code about
the proper way to conduct scientific inquiry. Acquiring the capacity to gossip,
and to gain access to gossip about physicists, data, detectors, and ideas
is the final and necessary stage in the training of a high energy physicist."
(Traweek, page 122)
At a more general level, Brown and Duguid (1995) note that
even documents, which appear to be fixed, immutable public entities whose very
purpose is to transcend social contexts, "play an important role, bringing
people from different groups together to negotiate and coordinate common practices."
Documents, they suggest, in their production, use, and distribution, have their
own social life, and function as mediators of and catalysts for social activity.
2.3 Social Capitalism
An awareness of ways in which work is bound up with social
factors has assumed a prominent place in business discourse regarding knowledge
management. Often referred to as organizational learning in these contexts,
knowledge management in the organization is seen as a collective process in
which teams create and share knowledge (e.g. Senge, 1990; Nonacka and Takeuchi,
1995; Cohen and Prusak, 2001; Boone, 2001). While proponents typically invoke
a systems perspective in thinking about organizational processes, they also
emphasize social factors such as relationships, trust, reputation, and
commitment in their descriptions of how such processes play out. As a
Vice President of Strategy puts it:
Expertise location is a big issue in companies today. The goal is
not only to provide access to information, but to provide access to people
who have the information. ... I don't want raw data, I don't even want information,
I want the judgments of people I trust. (Boone, page 22)
Recently the concept of social capital the "features
of social organizations such as networks, norms, and social trust that facilitate
coordination and cooperation for mutual benefit" (Putnam, 2000)
and the possible role it may play in the networked organization, has come to
the fore. Cohen and Prusak (2001) explain the connection:
Social capital makes an organization, or any cooperative group,
more than a collection of individuals intent on achieving their own private
purposes. Social capital bridges the space between people. Its characteristic
elements and indicators include high levels of trust, robust personal networks
and vibrant communities, shared understandings, and a sense of equitable participation
in a joint enterpriseall things that draw individuals together into
a group. This kind of connection supports collaboration, commitment, ready
access to knowledge and talent, and coherent organizational behavior. (Cohen
and Prusak, page 4)
Elaborating on the connection between social capital and
knowledge sharing, Cohen and Prusak point out that exchanging knowledge depends
on a social connection "without some degree of mutuality and trust,
the knowledge conversations will not get started; without some degree of shared
understanding, they will not go very far" (Cohen & Prusak, 2001, p.
86). They also note that the knowledge exchanged in spontaneous conversations
"is often social knowledge shared aims and interests discovered,
signals and stories shared that build confidence, trust, and connectionrather
than technical or business knowledge that can be directly applied to a product
or problem" (Cohen and Prusak, pp. 86-87).
2.4 The Challenge
Thus far we have argued that knowledge management is not
just an information problem, but that it is a social problem. That is, weve
suggested that effective knowledge management involves networks of people, relationships,
and social factors like trust, obligation, and commitment. One cant isolate
knowledge from its social context without denaturing it, without stripping it
of the social resources and social knowledge that contribute to its utility.
Taking the social nature of knowledge seriously raises a
considerable challenge for those interested in designing knowledge management
systems. We suggest that the place to start is to stop thinking in terms of
knowledge management, and start thinking in terms of supporting the larger social
context in which knowledge management is embedded. Our response to this challenge
is to explore the role of online multiuser environments. In particular, we are
interested in environments within which users can engage socially with one another,
and, in the process, discover, develop, evolve, and explicate knowledge. We
refer to online multi-user environments used in these ways as "knowledge
communities." In what follows we discuss current environments that function
as knowledge communities, and then turn to our own work on the topic.
3. Knowledge Communities
Knowledge communities have a long history, albeit not by
that name. The idea that networks of computers might provide a medium within
which individuals might come together to share knowledge and expertise dates
back to at least 1960. Perhaps the first vision of this nature was offered by
Simon Ramo (Ramo, 1961), who wrote of "many millions of human minds ...
connected together." Ramo offered a number of scenarios, including one
of an attorney consulting an on-line database that contained more than data:
"Even on the nonroutine legal processes,
the attorney, in the coming intellectronic age, will be able to consult with
the equivalent of a host of informed fellow attorneys. His request to the
system for similar cases will yield an immediate response from the central
store, together with questions and advice filed by other attorneys on those
similar cases -- even as he will add his facts and guidance into the system
for future use by all." page 10.
Simon Ramo, The Scientific Extension of
the Human Intellect. Computers and Automation, Vol 10., No. 2, pp 9--12.
February 1961. (Based on a talk given in 1960).
Over the ensuing decades the idea spread and evolved. From
its beginning as a vague if exciting vision, it took concrete form in the special
purpose DELPHI and EMISARI systems pioneered by Murray Turoff in the early 70s
(Turoff, 1972; Hiltz and Turoff, 1993) and in the PLATO Notes system in the
mid 70s (Wooley, 1993). In the late 70s and early 80s the
idea took off, spreading and evolving, under pressures from application domains
such as education and gaming, into a variety of genres of software ranging from
bulletin board systems to MOOs.
3.1 Some Examples of Knowledge Communities
A complete account of the systems which are used to enable
online groups to share knowledge among themselves is well beyond the scope of
this chapter. Instead, we will take the tack of looking at some representative
examples to give an idea of both the types of systems and the forms of use that
are used in managing online knowledge. It is important to note that we are not
just interested in the software; we are interested in the combination of the
software and the way in which it is put to use by its users we refer
to this comvination of technology and usage as a knowledge community.
One genre of software that supports knowledge communities
is the MOO. MOOs, originally developed as multi-user text-based gaming environments,
have been applied to a number of pedagogical and business ends. Examples include
MOOSE Crossing, an educationally-oriented environment for children from eight
to thirteen (Bruckman, 1997); Pueblo, a school-centered MOO in Phoenix, Arizona
(O'Day, et al, 1996); Tapped In, a distributed community of teachers (Schlager,
et al., 1998; Schlager et al., in press); and a MUD used by employees at Argonne
National Labs for work-related talk (Churchill and Bly, 1999).
Another genre of system that can support knowledge communities
is the electronic mailing list, or listserv. While mailing lists are used for
a variety of purposes, the existence of mailing lists used to share knowledge
among cohesive, long-lasting communities is well documented. In one case, a
community of about a thousand professional journalists have used a mailing list
to help one another with technical problems and to find story-specific information
sources for over six years (Millen and Dray, 1999; Millen 2000). Another example,
the use of a mailing list to support discourse amongst a scholarly community,
is described by Ekeblad (1999). And a third example, the use of a listserv by
a community of soap opera fans, to share knowledge ranging from plot summaries
to character background information, is described by Baym (Baym 1995; Baym 1997).
In addition to the genres of mailing lists and MOOs, which
can be turned to a variety of ends other than knowledge sharing, quite a few
systems have been designed with their principal aim being the support of a knowledge
community. One example is Answer Garden (Ackerman, 1998), a blending of bulletin
board and email systems that makes a network of questions and answers available
to its users, and uses email to automatically route new questions to appropriate
experts whose answers are then incorporated into the network. The Zephyr Help
Instance (Ackerman and Palen, 1996) has a similar purpose providing online
help information but uses a synchronous chat-like mechanism to broadcast
questions and answers to the user community. Another genre of knowledge community
system is the collaboratory. Collaboratories are aimed at the needs of the scientific
community, and provide real time access to scientific instruments along with
synchronous communication channels ranging from textual chat to real time audio
and video (Olson and Olson, 2000). Collaboratories are a highly successful class
of applications, with many in existence that have supported dozens to hundreds
of users for periods of years.
If one examines these systems and the ways in which theyre
used to share knowledge, an interesting commonality emerges: Virtually all of
these systems exhibit a rich array of social phenomena, in spite of the fact
that most provide only textual communication mechanisms, typically synchronous
chat, asynchronous email, or both (as in MOOs). (Even collaboratories, which
are increasingly support ing various forms of high bandwidth synchronous interaction,
functioned well when chat was their dominant communication channel.) Examples
of the social phenomena found in most knowledge communities range from interpersonal
phenomena such as the negotiation of status and reputation or the development
of trust, to the emergence of group norms and conventions. While these systems
bear eloquent testimony to the ingenuity of their users in using textual representations
to support a rich array of social phenomena, we suspect that we can do better.
This brings us to the question which informs our own work.
What would it mean to design an infrastructure for a knowledge community from
the ground up? That is, if we take seriously the charge that knowledge management
is a social problem as well as an information problem, one response is to ask
how we can better support social interaction. How do we go about designing a
system which supports not just information sharing, but that supports the exchange
of social knowledge and resources, the creation and growth of interpersonal
networks and accompanying social phenomena such as trust, obligation, commitment
To address this question, weve developed a system
called "Babble" which weve used as a testbed for exploring these
issues over the last four years. We begin by discussing the rationale that underlies
Babble's design: the notion that increasing the visibility of the presence and
activity of participants in an online environment can provide a foundation for
a variety of social processes and activity. Next we describe the system that
weve implemented, and discuss the ways in which weve come to use
it as part of our daily work practice. Finally, we discuss our experiences in
deploying Babble to other work groups.
3.2 Supporting Online Social Interaction
In the building where our group works there is a door that
opens from the stairwell into the hallway. This door has a design problem: opened
quickly, it is likely to slam into anyone who is about to enter from the other
direction. In an attempt to fix this problem, a small sign was placed on the
door: it reads, "Please Open Slowly." As you might guess, the sign
is not a particularly effective solution.
Lets contrast this solution with one of a different
sort: putting a glass window in the door. The glass window approach means that
the sign is no longer required. As people approach the door they see whether
anyone is on the other side and, if so, they modulate their actions appropriately.
This is a very simple example of what we call a socially translucent system.
While it is obvious why this solution works, it is useful
to examine the reasons behind it carefully. We see three reasons for the effectiveness
of the glass window: First, the glass window makes socially significant information
visible. That is, as humans, we are perceptually attuned to movement
and human faces and figures: we notice and react to them more readily than we
notice and interpret a printed sign. Second, the glass window supports awareness:
I dont open the door quickly because I know that youre on
the other side. This awareness brings our social rules into play to govern our
actions: we have been raised in a culture in which slamming doors into other
people is not sanctioned. There
is a third, somewhat subtler reason for the efficacy of the glass window.
Suppose that I dont care whether I hurt others:
nevertheless, Ill open the door slowly because I believe that you know
that I know youre there, and therefore I will be held accountable
for my actions. (This distinction is useful because, while accountability and
awareness usually co-occur in the physical world, they are not necessarily coupled
in the digital realm.) It is through such individual feelings of accountability
that norms, rules, and customs become effective mechanisms for social control.
We call systems that exhibit these properties of
perceptual salience, awareness, and accountability socially translucent
systems. But there is one other aspect of social translucence that deserves
mention. Why is it that we speak of socially translucent systems rather
than socially transparent systems? Because there is a vital tension between
privacy and visibility. What we say and do with another person depends on who,
and how many, are watching. Note that privacy is neither good nor bad on its
ownit simply supports certain types of behavior and inhibits others. For
example, the perceived validity of an election depends crucially on keeping
certain of its aspects very private, and other aspects very public. As before,
what we are seeing is the impact of awareness and accountability: in the election,
it is desirable that the voters not be accountable to others for their
votes, but that those who count the votes be accountable to all.
We see these three properties of socially translucent systems
visibility, awareness, and accountability as critical building
blocks of social interaction. Notice that social translucence is not just about
people acting in accordance with social rules (see Erickson & Kellogg, 2000).
In socially translucent systems we believe it will be easier for users to carry
on coherent discussions; to observe and imitate others actions; to engage
in peer pressure; and to create, notice, and conform to social conventions.
We see social translucence as a requirement for supporting online communication
and collaboration in general, and knowledge communities in particular.
This brings us to the question of how to support social
translucence in online environments. How can we provide the cues that allow
our socially based processes to operate and which are so ubiquitous and
lightweight in the physical world in online systems? Two obvious approaches
are to use video or 3D virtual environments. However, these have several drawbacks
for our purposes. First, they don't scale well: we would like to support conversations
among fairly large numbers of people. Second, both approaches are best suited
for supporting synchronous interactions, whereas we would like to support both
synchronous and asynchronous interaction. Third, they are both relatively demanding
in terms of processing power, bandwidth, and display space and characteristics:
we would like to be able to support mobile employees working over sub-56K connections
and using devices with smaller displays.
As a consequence, we have taken a more abstract approach
to supporting social translucence. The abstract approach involves portraying
social information in ways that are not closely tied to its physical analogs.
Exemplars of the abstract approach include the Out to Lunch system (Cohen, 1994),
which uses abstract sonic cues to indicate socially salient activity, and Chat
Circles (Viegas, et al. 1999), which uses abstract visual representations. This
approach also includes the use of text to portray social information; as we
have already noted, text has proved surprisingly powerful as a means for conveying
social information in knowledge communities.
3.3 The Babble System
Babble (Erickson, et al., 1999) is an online environment
intended to support both synchronous and asynchronous text-based conversations
within small to medium sized groups. The principle goal of Babble has been to
serve as a platform for exploring ideas about the social effects of supporting
mutual awareness among online groups. However, to do this effectively, we needed
to be able to observe 'real' workgroups using it as part of the daily work process.
As a consequence, Babble needed to be sufficiently robust and lightweight to
be usable by groups who don't care about the technology itself.
In terms of infrastructure, Babble is a client-server system
with both components written in SmallTalk. Babble stores all data, except for
user specific preferences and state (e.g., the user's last location, last items
read, and so forth) on the server and broadcasts it as needed. Babble clients
request the data they need from the server (e.g., when a user switches to a
new conversation the client requests the content), and also notify the server
of events that it will broadcast to other clients. As this architecture suggests,
Babble only works when on a network; when disconnected it has no cache of conversation
text. The Babble server runs on a variety of server-class machines; the principle
client runs on PCs, though we have had, for varying durations, clients that
ran on the Macintosh (in Java) and on the Palm Pilot. Here we discuss only the
PC client, since that comprises the vast majority of our experience.
In terms of functionality, Babble resembles a multi-channel,
text-based chat system in that many users can connect to it, and select one
of a variety of conversations to participate in (or create their own). However,
Babble differs from conventional chat in two ways, both of which stem from our
interest in supporting knowledge communities. First, the textual conversation
that occurs in Babble is persistent: that is, unlike conventional chat where
newly arriving users only see what has transpired since they've joined a channel,
Babble users can see everything ever typed in any existing conversation. These
traces give the system the potential to function as a knowledge store, or what
we prefer to call a "discourse base." Second, Babble makes the presence
and activity of the participants visible by a variety of means, but principally
through what we call the social proxy.
Figure 1 shows the Babble user interface. In the upper left
hand corner is a list of the names of users currently connected to Babble. In
the middle upper pane is the social proxy, which we will describe shortly. In
the upper right pane is a hierarchical list of the Babble conversation topics
(grouped in categories and subcategories). And the pane that occupies the lower
half of the window contains the text of the current conversation (whose topic
name is highlighted in the topics list); within the pane, each 'comment' is
prefaced with the name of the user, and date and time of its creation (recall
that Babble conversations need not be synchronous; indeed, some are asynchronous,
with hours, days or weeks separating comments). Babble provides a variety of
other types of functionality via the menu bar, context-sensitive menus accessed
via right clicks, and keyboard shortcuts. These include functions for creating
messages, creating, changing, and deleting topics and categories, conducting
private, ephemeral chats, and so forth.
The social proxy, in the upper middle part of the window,
represents the current conversation as a large circle, and the participants
as colored dots, referred to, hereafter, as marbles. Marbles within the circle
are involved in the conversation being viewed; marbles outside the circle represent
those who are logged on but are viewing other conversations. What makes the
social proxy interesting has to do with the position of the marbles in the circle.
When a user becomes active, either 'speaking' (i.e., typing) or listening
(i.e., interacting with the conversation window by clicking or scrolling), the
user's marble moves rapidly to the center ring of the circle. If the user stops
interacting, the marble gradually drifts out to the inner periphery of the circle
over the course of about twenty minutes. Thus, when there is a lot of activity
in the conversation, there is a tight cluster of marbles around the center of
the circle. The social proxy shown in Figure 1 depicts a situation in which
five people have been recently active (i.e., speaking or listening) in the current
conversation, and two others have been idle for a while (and an eighth person
is off viewing another conversation).
When people leave the current conversation their marbles
move outside the circle; when they enter the conversation, their marbles move
into the circle. When a person logs onto the system, it creates a virtual wedge
for their marble, adjusting the position of all the marbles in the social proxy;
when they depart, the wedges are destroyed, and the remaining marbles adjust
to uniformly occupy the space. All marble movements are shown with animation,
thus making arrivals, movements, and departures visually salient. Although simple,
this social proxy gives a sense of the size of the audience, the degree to which
the audience is actively listening or contributing, as well as indicating whether
people are gathering or dispersing, and who it is that is coming and going.
In addition to the social proxy (which we refer to as 'the
cookie'), Babble uses additional mechanisms to reveal the presence and activity
of users. In the topic list, to the left of the topic names, are 'mini-cookies',
thumbnails of the social proxy for each topic with a participant in it. So,
in Figure 1, we can see that there is a single person in the second topic, "Amusing
Wendy." Babble also highlights information that the user hasn't yet seen:
the names of topics with new material in them are shown in red (e.g., "Amusing
Wendy" in Figure 1), and comments that have been added to the current conversation
since the user last 'touched' Babble are shown in reverse highlighting.
One of the shortcomings of the cookie is that it only works
for synchronous interactions that is, it shows only the presence and
activities of people who are currently logged on to Babble. This is a considerable
drawback because the majority of the conversations carried on in Babble are
asynchronous, with just a few comments per day (or per week, or per month).
As a consequence, we designed a second, asynchronous social proxy for Babble:
the Timeline (Figure 2).
The basic goal of the Timeline was to provide a way for
a speaker to see that people were listening (or not),
even when the listening was offset in time. The Timeline proxy works as follows:
each user is represented by a row in the Timeline; when they are logged on to
Babble, they leave a flat trace or line, and when they speak they
leave a vertical mark or blip on the line. If the line/blip is in color, it
means that that user was present/speaking in the conversation currently being
viewed by the user of the timeline; if they were in a different conversation,
the line/blip is shown in gray (and the line becomes thinner). As the user mouses
over the Timeline, the name of the topic, the user, and the time being examined
is shown in the upper left corner; the user can scroll back through as much
as one week of activity. The Timeline also provides access to other functionality
via a menu accessed via a right-click on another users row (e.g., private
For example, in Figure 2, we can see that nine people have
logged onto Babble (shown by the presence of lines), and that all of them have
spent some time in the current conversation (shown by the color/increased thickness
of the lines), and that many but not all have spoken (shown by the
blips). The line being indicated by the cursor shows that the user Peter
logged on around 11am, made a couple of comments in the "Commons Area"
conversation, switched to another topic, and then switched back to the Commons
area about 1pm, and then logged off.
3.4 How a Babble is Used by a Group
While one must be wary about drawing conclusions concerning
the usability of software when it is used by its developers, our aim here is
to simply provide a sense for how Babble is actually used by a group. Well
begin by describing the group, and then move on to discuss how Babble is actually
used. In the next section we'll discuss our deployments of Babble to other groups
and some of the phenomenon that weve observed across different deployments.
Our group has used Babble for about four years. The group
is centered around the software development group (AKA "the lab")
that designed and implemented the system, and includes a mix of computer scientists
and social scientists (including the authors). The size of the group has varied
in number over the years from four to nineteen users. Part of the variance is
due to the ebb and flow of people characteristic of groups in large organizations;
and part is due to current members of the lab inviting "associates"
colleagues with whom they had strong social or professional ties
to join Babble.
Geographically, the group of Babble users is about half
co-located in New York, and half distributed. Most of the lab members are located
in the same building, although offices tend to be distributed around the building
so actual adjacency is rare. Three members of the lab are telecommuters,
and spend the majority of their time tens to hundreds of miles away; other members
of the lab frequently work at home. Four of the six associated colleagues (i.e.
those not officially members of the lab, but users of Babble) are remotely located.
Socially, the lab is a cohesive group, with considerable
camaraderie. In addition to work-based collaboration, the lab members occasionally
socialize, although usually within business hours (e.g., going out to lunch)
The associates vary in the strength and number of their ties to the lab members,
some known to almost all lab members, and others known only to one or two lab
members with whom they have shared interests. Conversation in the Babble system
moves fluidly between work and social talk; it is always civil, frequently informal,
and joking, teasing, and other ludic behavior is not unusual.
Overall, the Babble system as used by the lab can be characterized
as a core of relatively synchronous activity surrounded by a constellation of
asynchronous conversations. At the center of activity is a topic called the
"Commons Area," a place where collocated and remote members greet
one another, share news, engage in banter, and ask general questions. Members
of the lab tend 'hang out' in the Commons Area, often remaining logged on for
most of the work day. Comments in the Commons Area tend to be short and informal,
with relaxed syntax and punctuation, use of paralinguistic expressions ("ummm"),
onomatopoeia, emoticons, and playful tropes (for example, the tossing
of cookies to a dog who usually accompanies one
of the participants all done via text, of course). The content of conversation
in the Commons Area ranges from purely social talk (such as the custom of saying
"good morning"), to the posing of general questions, to reminding
people of an impending meeting of general interest, to more technical discussions
about work projects. (In theory, more topic-oriented discussion is supposed
to take place in specific topics; in practice, work talk often grows out of
social discussions, and the recognition that a substantive conversation that
belongs somewhere else is taking place is often not recognized until
after the fact.) Because of the amount of talk that occurs in the Commons, the
content of the Commons Area is automatically archived once a week.
In addition to the commons area, there are a variety of
other topics or conversation areas in Babble. These have ranged in number from
a dozen or so in the early days of Babble to several dozen, the growth being
facilitated by the addition of an expandable hierarchical topic list. These
topics tend to have asynchronous and mostly sporadic conversation, and they
tend to be focused on particular purposes, typically either project-oriented
or person oriented. Examples of topics include personal offices (e.g., "Tom's
Office"), project-oriented topics (e.g., "Babble EthnographiesCB
Babble"), and occasional non-work topics (e.g., "Bad Jokes").
In general, uses of Babble can be grouped into three general
categories: social/ludic; informative; and instrumental. Social/ludic activities
are those engaged in for social and entertainment purposes such as the custom
of exchanging morning greetings, and the topic devoted to jokes. Informative
activities have to do with actions on the system that are addressed to the group
as a whole, or to no one in particular, and generally are done without expectation
of a reply or responsive action. These activities include posting announcements
and other news believed to be of general interest, commenting on project activity,
and keeping on-line notebooks or offices. The third type of activity is instrumental,
that is, activities engaged in with a particular end in mind. These include
starting or participating in focused discussions, posting bug reports, holding
on-line meetings, and asking questions. These activities are often, though not
always, addressed to a particular participant or group of participants.
3.5 Adoption and Social Phenomena across Babble
Over the last four years we've deployed Babbles to about
twenty groups. Weve studied the deployments using techniques ranging from
ethnographic studies see Bradner, et al. (1999) for a study of six Babbles
to studies based on surveys and analyses of log data and conversation
We have had mixed experiences with the adoption of Babble.
Sometimes groups try Babble out, but fail to adopt it (typically about six weeks
pass before it is evident whether or not the Babble is going to be adopted by
the group). Other times groups use Babble for a period of months, and then cease
(either because it was for a particular event or period that has ended, or because
the composition or needs of the group change). It isnt clear how to operationally
define a successful deployment of Babble: the group uses it for its entire existence?
the group uses Babble actively for six months? the group uses Babble to carry
out a particular activity? If we take, as a rule of thumb, that a Babble is
successful when it is used on a more or less daily basis by several people for
more than six weeks, we can say that about half of our Babble deployments have
met with success. As of this writing, we have five Babbles running, all of which
are well past the six week mark, and all exhibiting robust daily activity.
When a Babble is adopted by a group, it usually supports
a variety of communicative purposes and practices (often similar to those described
in the previous section). Here, we describe four social phenomena that weve
observed in a number (though not all) of successfully adopted deployments that
are most relevant to knowledge communities.
One phenomenon is waylay, in which a user watches for a
particular person to become active on Babble (signaled by the movement of their
marble into the center of the social proxy), and then initiates a conversation
(either publicly within Babble, via Babbles private chat mechanism or
by some external means such as the telephone). Because the movement of the marble
occurs when the user has just begun an episode of typing or mousing, it indicates
a opportune moment for contact (since the users attention has just shifted
to communication with the group). Waylay is used for purposes ranging from asking
questions to initiating casual social chat. In general, forms of opportunistic
interaction such as waylay permit the same sorts of requests for assistance
and transfers of social resources that weve observed in face to face knowledge
sharing situations, with the accompanying effect of strengthening of interpersonal
Babble also supports the maintenance of group awareness
through the exchange of social knowledge. For example, when members of a Babble
travel, many report reading through conversations that occurred in their absence
to find out what happened. For someone who is a member of the group
and understands the context, seemingly trivial comments can convey considerable
information about whats going on at the individual, group, and organizational
levels. Thus, a sign off "I have to go to the [project] meeting
now" reveals that one participant is still involved in a particular
project, and a question "Does anyone know how to do a screen capture"
indicates that someone is beginning to write a paper. Babble also supports
group awareness through the Timeline proxy. Babble participants have reported
uses such as: looking to see who has visited a topic in which they had posted
questions; looking to see whether a colleague who hadnt posted recently
had been online; and using the Timeline to get a sense for the activity of the
community as a whole.
Another phenomenon that can be observed across Babbles is
the development of social norms. That is, one participant may develop a particular
way of doing something, and others will imitate it. Examples of this include
what users include in their online nickname (e.g., in some Babbles users append
"@mylocation" after their name), the types of online conversations
created (e.g., some Babbles have categories for "personal places"
or "offices"), and naming conventions (e.g., one Babble uses the term
"chit-chat" to signal that a topic is intended for casual conversation.
Babble groups also evolve various interactive customs, the most common being
to say 'hello' upon logging in (even when no one else is present). Again, the
existence of these norms supports social interaction by providing expectations
about how to behave.
Finally, weve observed that Babbles are typically
regarded as semi-private, "trusted" places. This became apparent when
strangers appeared in various Babble systems. Sometimes the strangers
were unannounced new members, sometimes they were visitors provided access by
an unreflective manager, and, in one case, the stranger was actually an unannounced
conversational software agent. But in all cases, the arrival and presence of
the stranger (reflected in the social proxies along with the presence of the
regulars) was greeted with considerable consternation. In each case, the appearance
of strangers provoked concern about how unguarded conversations might be interpreted
by those from different contexts, and led to the creation of visitor and membership
policies. We suggest that this concern reflects the success of Babble as an
online space that is rich in social context.
One issue that is not clear, so far, is the degree to which
Babbles social proxies contribute to these phenomena. Analytically, it
is difficult to isolate the effects of the social proxies, from the effects
of purely textual cues. Certainly, there are a number of social practices (such
as waylay) which require (or are at least greatly facilitated by) the proxies.
It is clear that the participants, in general, like the proxies and want
them retained as a feature of the system. One user, responding to a question
in Babble, writes:
"Ah, the cookie
we love the cookie
the cookie is
good our colored dots circulate around to make room when
someone new joins the conversation thats fun. And when someones
connection dies, they rather dissemble into the ether, angelic like. Which
is sort of fun to watch.
Also, when Im wondering whether my comments
have fallen on deaf ears, I can tell when a response may in fact be on its
way when someones dot moves back to the center (happens as soon as someone
starts typing). So, yes, we like the cookie it makes me feel like there
are actually people in a room with me
It is also clear that users are able to read
Babble proxies, using them to draw inferences about the presence of individuals
and the activity of the community as a whole. Another user, commenting on the
Timeline proxy, remarks:
"Its a little like reading an electrocardiogram, the
heartbeat of the community. I noticed that I missed Sandy by an hour on Monday
morning.... Pat comes in every so often as a blip. Lynn jumps from space to
Nevertheless, although we have compelling anecdotes and
a large fund of positive comments by Babble users, analytically separating social
benefits conveyed by proxies from those produced by text remains as a challenge
for the future.
3.6. Babble as an Infrastructure for Knowledge
Babble clearly succeeds as a multi-user online environment
where sustained social interaction takes place. But does it support knowledge
communities? Is the social interaction combined with the sharing of information,
social knowledge and social resources via personal social networks that, we
suggest, is a crucial part of knowledge management? This is indeed what we have
observed. In the following, we refer to examples and survey results drawn from
a Babble whose membership is composed of a world wide cross-section of people
in IBM and Lotus interested in online communities.
Perhaps the first point to make is that participants do
feel as though they are part of a community. This is particularly important
to those who are remote teleworkers:
"I work remotely and can feel very isolated when I dont
travel regularly (as has been the case for the past six months because of
travel restrictions). Babble has provided me with a way to feel connected
with a group of people outside my basement walls. It is my portal (so to speak)
"As a home office worker, this is perhaps one of the things
I miss the most the ongoing banter I can have with colleagues who are
focused on a similar work topic as I am."
This is not simply a feeling of a vague belonging to a group;
participants report feeling as though they are hooked into social networks.
One participant reports that participation in Babble strengthened an existing
"Babble has helped me establish a tighter social and professional
relationship with all of them we have much more regular contact with
each other, much as we would if we were collocated, via the Babble connection.
This in turn has built social capital among us which may be of use in the
And these social networks are not just about talk, they
can also be tapped for assistance. The participant continues:
"I have also contacted Vera about getting her input and advice
about setting up a knowledge network, which is part of my real work.
I felt much more comfortable about approaching her with this question as a
result of our frequent contacts via Babble than I would have otherwise."
Another Babble member notes:
"I like the back and forth.
we have a lot of reflective
talk about our own experiences... In at least one case, e.g., a half-joking
comment of mine, "anybody want to fund this?" has led to e-mail,
phone, and face-to-face meetings and now a serious proposal for funding. I
dont know the final outcome yet, but we have found out something significant
about another part of the business and have made a serious attempt to propose
[a] solution to their problems."
These comments are prima facie evidence that knowledge sharing
and expertise management are deeply social processes that people value
informal exchanges with colleagues, and may only venture a non-trivial request
for information or assistance after a social relationship has been established.
A danger in using the summary remarks of participants to
what happens in Babble is that it makes it sound a bit more straightforward
and calculated than it is. It is difficult to convey the way in which these
effects emerge out of a rich melange of social and work talk. For example, one
instance of the transfer of social resources occurred over the course of a multi-threaded,
30 utterance, 17-minute Babble conversation on March 7, 2001. The conversation
consisted of two primary participants (scienceguy and Patrick),
and was composed of four distinct threads. Two threads were related to work
topics (Patrick explaining that he had referred some colleagues to scienceguy
, and a discussion of the use of patterns in knowledge management), and two
were more social threads (one an attempt to identify an earlier participants
real name, another a request by scienceguy for assistance in developing an Irish
accent for an upcoming storytelling performance). The two work related tasks
were treated relatively seriously, even as the two interleaved non-work threads
were used as an excuse for banter. Yet both the social and work threads developed
and played off one other throughout the conversation, which concluded with Patrick
revealing the names of the colleagues whom he has referred to scienceguy, and
scienceguy indicating that he would be happy to talk with them.
(The situation grows more complex when one recognizes
that Babble users are remote from one another, and may be simultaneously carrying
on other work on their computers, via the telephone, or orally with co-located
4. Concluding Remarks
In this chapter weve argued that knowledge management
is not just an information problem, but is, as well, a social problem that involves
people, relationships, and social factors like trust, obligation, commitment,
and accountability. This view raises a considerable challenge for those interested
in designing systems to support knowledge management. Our approach has been
to explore the creation of infrastructures for knowledge communities: on-line
environments within which users can engage socially with one another, and, in
the process, discover, develop, evolve, and explicate knowledge.
In our work on Babble, weve begun exploring ways of
creating infrastructures that support rich forms of social interaction. Weve
found that social proxies are a promising development, and continue to be impressed
with the power of plain text as a means of supporting interactions that are
both complex and subtle. We believe that one of the most important aspects of
a knowledge community is that it can be used as a place for unguarded discussion
among people who know one another, who share professional interests, and who
understand the contexts within which their remarks are being made.
The notion of a knowledge management environment as a trusted
place is an interesting and challenging one. How technically, socially,
and organizationally can we balance the need for a safe and trusting
place with the organizational imperative to share information? One decision
facing us as designers is how and to what extent we "design in" norms
and social conventions. For example, if we build in technical mechanisms to
provide privacy, in addition to the usability impact, we also eliminate opportunities
for participants to show that they may be trusted, or to rely on others to respect
their privacy. The Babble prototype has no technical features for controlling
access: anyone who has access to the client could, in theory, enter any Babble
space. But, because Babble makes users visible, this results in groups noticing,
commenting on, and ultimately discussing how to deal with this issue. We believe
that a greater understanding of how to design systems that permit social mechanisms
to come into play is of great importance in designing future systems for knowledge
Thanks to David N. Smith for creating the Babble prototype,
to Mark Laff , Peter Malkin, and Amy Katriel for implementation work on the
Babble server and clients, and to Cal Swart for critical assistance in the deployment
of a multitude of Babbles. Thanks, as well to members of IBM's Social Computing
and Pervasive Applications groups for productive conversations, and to the many
dozens of 'Babblers' who have shared their insights, responded patiently to
surveys, and, most importantly, used Babble in many of productive (and often
Ackerman, M. and Halverson, C. (1999); Considering organizational
memory: Processes, Boundary Objects, and Trajectories, Proceedings of the
32nd Hawaii International Conference on System Sciences, 1999.
Ackerman, M.S. (1998); Augmenting organizational memory:
a field study of Answer Garden, ACM Transactions on Information Systems,
Vol. 16, No. 3, July 1998, p203-24.
Ackerman, M. S. and Palen, L. (1996) The Zephyr Help Instance:
promoting ongoing activity in a CSCW system. In Human Factors in Computing
Systems: The Proceedings of CHI '96. New York: ACM Press, pp. 268 - 275
Baym, N. K. (1995) The Emergence of Community in Computer-Mediated
Communication. Cybersociety: Computer-Mediated Communication and Community
(ed. S. Jones). Thousand Oaks, CA: Sage Publications, 1995. pp 138-163.
Baym, N. K., (1997) Interpreting Soap Operas and Creating
Community: Inside an Electronic Fan Culture, In Culture of the Internet
(ed. S. Kiesler). Mahwah, NJ: Lawrence Erlbaum Associates, pp. 103-120.
Boone, M. E. (2001). Managing Inter@ctivity. New
York, NY: McGraw-Hill.
Bradner, E., Kellogg, W.A., and Erickson, T. (1999) The
Adoption and use of Babble: A field study of chat in the workplace. Proceedings
of the European Conference on Computer-Supported Cooperative Work (ECSCW
'99), 139-158, Kluwer Academic Publishers, 1999.
Brown, J.S. and Duguid, P. (1995) The Social Life of Documents.
In Release 1.0 (ed. E. Dyson), October 1995. New York: EDventure Holdings
Inc., pp. 1-18. Also at: www.parc.xerox.com/ops/members/brown/papers/sociallife.html
Bruckman, A. (1997) MOOSE Goes to School: A Comparison
of Three Classrooms Using a CSCL Environment. In the Proceedings of CSCL 97,
Toronto, December 1997.
Churchill, E. F. and Bly, S. (1999) "Virtual Environments
at Work: Ongoing use of MUDs in the Workplace." The Proceedings of WACC
'99 (San Francisco, CA, 1999). New York: ACM Press.
Cohen, D. and Prusak, L. (2001) In good company: How
social capital makes organizations work. Boston, MA: Harvard Business
Cohen, J. (1994). Monitoring background activities. In
G. Kramer (Ed.), Auditory display. New York: Addison-Wesley, pp. 439-531.
Donath, J., Karahalios, K., & Viegas, F. (1999). Visualizing
conversation. In J.F. Nunamaker, Jr., & R.H. Sprague, Jr., (Eds.), Proceedings
of the Thirty-Second Hawaii International Conference on Systems Science.
Los Alamitos, CA: IEEE Computer Society Press.
Ekeblad, E. (1999). The Emergence of Multilogue on a Scholarly
Mailinglist, Paper presented in the symposium: "Time and coordination
in a virtual community of learners," at the 8th European Conference for
Research on Learning and Instruction (EARLI), Göteborg, Sweden, August
24-28 1999. Available at: http://hem.fyristorg.com/evaek/writings/earli99/multdyn.html
Erickson, T., Smith, D.N., Kellogg, W.A., Laff, M.R.,
Richards, J.T., and Bradner, E. (1999) Socially translucent systems: Social
proxies, persistent conversation, and the design of Babble. CHI 99 Conference
Proceedings: Human Factors in Computing System, 72-79, ACM Press, 1999.
Erickson, T. and Kellogg, W.A. (2000). Social translucence:
An approach to designing systems that mesh with social processes. ACM Transactions
on Computer-Human Interaction, Vol. 7, No. 1, pp. 59-83.
Erickson, T. and Laff, M.R. (2001) The design of the Babble
A social proxy for visualizing group activity over time. In Extended Abstracts:
The Proceedings of CHI 2001. New York: ACM Press.
Hiltz, S. R. and Turoff, M. (1993). The Network Nation
(Revised edition), Cambridge, MA: MIT Press.
Latour, B. (1987) Science in Action. Cambridge,
MA: Harvard Universtiy Press.
Latour, B. and Woolgar, S. (1979) Laboratory Life:
The Social Construction of Scientific Facts. London: Sage. (2nd
edition; Princeton: Princeton University Press, 1986).
Lave, J. and Wenger, E. Situated Learning: Legitimate
Peripheral Participation. Cambridge: Cambridge University Press, 1991.
Millen, D. (2000) Community Portals and Collective Goods:
Conversation Archives as an Information Resource. Proceedings of HICSS - 33rd
Annual Hawaii International conference on Systems Sciences, January 4-7, 2000.
Millen, D. R., & Dray, S. (1999)
Information Sharing in an Online Community of Journalists. In Proceedings
of Esprit i3 Workshop: Ethnographic Studies in Real and Virtual Environments:
Inhabited Information Spaces and Connected Communities. Edinburgh, January
Nonaka, I. and Takeuchi, H. (1995) The Knowledge-Creating
Company : How Japanese Companies Create the Dynamics of Innovation. New
York: Oxford University Press.
O'Day, V. L., Bobrow, D. G., and Shirley, M. (1996) "The
Social-Technical Design Circle. In the Proceedings of CSCW 96. (November 1996,
Cambridge, MA). New York: ACM Press.
Olson, G. and Olson, J. (2000). Distance matters. Human-Computer
Interaction, Volume 15, Vol. 2-3.
Orr, J. E. Talking About Machines: An Ethnography of
a Modern Job. Ithica, NY: Cornell University Press, 1996.
Putnam, R. D. (2000) Bowling Alone: The Collapse and
Revival of American Community. New York: Simon and Schuster.
Ramo, S. (1961) The Scientific Extension of the Human
Intellect. Computers and Automation, Vol 10., No. 2, pp 9-12. February
1961. [Based on a talk given in 1960].
Schlager, M., Fusco, J., & Schank, P. (1998). Cornerstones
for an on-line community of education professionals. IEEE Technology and Society
Magazine, 17(4), 15-21/40.
Schlager, M., Fusco, J., & Schank, P. (in press).
"Evolution of an On-line Education Community of Practice." To appear
in K. A. Renninger and W. Shumar (Eds.), Building virtual communities:
Learning and change in cyberspace. NY: Cambridge University Press.
Senge, P. M. (1990) The Fifth Discipline : The Art
and Practice of the Learning Organization. New York: Doubleday.
Traweek, S. Beamtimes and Lifetimes: The World of high
energy physics. Cambridge, MA: Harvard University Press, 1988.
Turoff, M. (1972) Delphi Conferencing: Computer-Based
Conferencing with Anonymity, Technological Forecasting and Social Change,
Vol. 3, 159-204.
Viegas, F.B. and Donath, J. Chat Circles. CHI 99 Conference
Proceedings: Human Factors in Computing System, 9-16, ACM Press, 1999.
Wenger, E. (1998) Communities of practice: Learning,
meaning, and identity. Cambridge: Cambridge University Press.
Wooley, David. (1994). PLATO: The Emergence of Online
Community. On -line version at http://www.thinkofit.com/plato/dwplato.htm;
an earlier version appeared in Matrix News (ed. J. Quarterman), January