Cognitive Artifacts in Complex Work - Ambient Intelligence for Scientific Discovery - page 166

of “discovery dialogue” requires access to deep sources of both related and di-
vergent literatures. The structure or depth of such analogies may differ widely
by discipline, but the access to related concepts presents itself as a design op-
portunity in the processes of discovery.

4.3 Artifacts and Resources in the Information Ecology

An institution's information ecology [22] provides infrastructure and resources
for scientific information practices, as well as constraints which guide and permit
certain actions and not others. In this infrastructure, scientists develop regular
patterns of use, as described in this model. Even as the institution adds re-
sources and services (through growth, acquisition of new content holdings, and
new services on the Web), the post-Internet information ecology now provides
a stable set of resources. With this stability, little demand is exhibited for new
services or features beyond those currently available. Instead of the dozens, if not
potentially hundreds of possible information resources available in the natural
sciences, we find two resources (PubMed and Google) consistently used, in the
specified context of research projects.
Another indication of resource stability in this ecology shows in the regu-
larity of information tasks performed by scientists on an everyday basis. Table
1 describes six information tasks found across all participants. These tasks are
regular, repeatable discrete information seeking actions. Each task represents
a different common objective, with scientists locating specific information arti-
facts associated with the task. These tasks are driven by seeking information
artifacts, following current theories in information seeking that identify a puta-
tive information need. But the task's cognitive drivers are not located in this
context. The cognitive work required by the research project demands specific
information objects, irrespective of artifact.

Habituated Stable Behaviors. An important finding disclosed by this model
shows these six information tasks are habituated; people resist switching to alter-
native resources or information seeking methods after cognitive accommodation
to sucient tools. Habituated information tasks are cognitively economical ,re-
quiring minimal conscious attention to perform successfully. These tasks tend to
become internalized as routine operations, demanding from the user only mini-
mal, if any, reflection on their interaction. Because they are stable, continually
performed, transparent tasks, we predict such tasks may translate as complete
functions in an information ecology enhanced by ambient systems, since the
resources and resulting artifacts will be recognized by all participants. This find-
ing has significant implications for design. First, these six “canonical” stable
tasks might be incorporated into information ecologies that distribute the arti-
facts used in research discovery. Second, because these information tasks do not
directly support research discovery, we identify a substantial gap between the
current ecology of information resources and the potential for supporting the
cognitive demands of distributed discovery. This gap delineates the boundary of
cognitive artifacts used in discovery, and the limits of technology.