David Woods
Cognitive Systems Engineering Laboratory
Institute for Ergonomics
The Ohio State University
Draft in progress
July
2000
Do not quote or distribute without
permission of the author.
The study of cognitive systems in context is a process of discovering how the behavior and strategies of practitioners are adapted to the various purposes and constraints of the field of activity.
This
single sentence both describes the enterprise and points to the difficulties
and challenges embedded in that enterprise. In studying cognitive systems in context we are out to learn
how the more or less visible activities of practitioners are parts of larger
processes of collaboration and coordination, how they are shaped by the
artifacts and in turn shape how those artifacts function in the workplace, and
how they are adapted to the multiple goals and constraints of the
organizational context and the work domain.
These factors of complex artifacts, dynamic worlds, cognitive work, coordinated activity, and organizational dynamics do not come to us pristine, isolated, one at a time. Rather they come in association with each other, embodied in the particular, cloaked by some observer’s vantage point. In other words, cognitive systems in context come in a “wrapped package” as a complex conglomerate of interdependent variables (Woods, 1993). Additionally, developing skill at cognitive work hides or obscures the underlying adaptive web across the cognitive systems triad. As the fluency law of cognitive work states: well adapted cognitive work occurs with a facility that belies the difficulty of the demands resolved and the dilemmas balanced.
‘Spartan’ lab experiments would simplify such conglomerates into more manageable units for experimental manipulation, but such efforts fall prey to reductionistic or oversimplification biases which eliminate the very phenomena of interest in the process of simplification (Feltovich et al., 1997). Simplifying dynamic processes into a series of static snapshots or treating a highly interconnected set of factors as separable represent a retreat from complexity that values the means of experimental tractability over the end of adding to our understanding of the phenomena of interest.
The target phenomena of interest for distributed cognition are reflected in the basic founding slogans of adaptations directed at coping with complexity, how things make us smart or dumb, and how to make automated and intelligent systems team players. Patterns on these themes in cognitive work exist only at the intersections of people, technology and work.
Understanding
the deeper patterns in this type of complex conglomerate is difficult in part because
activities and strategies of people are adapted to the constraints in the
environment of demands and the affordances of artifacts – a case of
agent-environment mutuality
Agent-environment
mutuality, after von Uexkull’s Umwelt (1934) and Gibson (1979), means that each
can be understood only in relation to the other and that each changes through
adaptation to the other. Agent and
environment are mutually adapted (when the environment is stable, i.e., when
sources of variability are not changing).
When the environment is changing (i.e., when sources of variability are
changing) adaptation ensues until a new equilibrium is reached.
Understanding
what an environment affords to agents (given their goals) and how agents’
behavior is adapted to the characteristics of the environment and how this
linkage changes is functional analysis.
Hence, the study of cognition in context is develops a functional model
that captures how the behavior and strategies of practitioners are adapted to
the various purposes and constraints of the field of activity. In the case of cognitive systems in
context this mutual adaptation is a three way interplay as captured in the
cognitive systems triad (Woods, 1988).
Partial
decomposability in this triad means that methods can be used to build up a
model of how any one is adapted to the conjunction of the other two. In other
words, the methodological challenges build from the fact that two of the three
factors will be confounded in any technique to tease open the interplay. This
is one of many constraints on method that produces the emphasis on
bootstrapping multiple converging methods in studying cognitive work in the
wild.
Bartlett F. C. (1932). Remembering: A study in experimental and social
psychology. Cambridge,
England: Cambridge University Press.
Feltovich, P.J., Spiro,
R. J. and Coulson, R. L. (1997). Issues of Expert
Flexibility in Contexts Characterized by Complexity and Change. In Paul J.
Feltovich, Kenneth M. Ford & Robert R. Hoffman (eds.) Expertise
in Context
Cambridge, Mass. : MIT Press.
Gibson, J. J. (1979). An Ecological
Approach to Perception.
Hutchins, E. (1995a). Cognition
in the wild. Cambridge, MA:
MIT Press.
Von Uexkull, J. (1934). A Stroll through the Worlds of Animals
and Men. In C. Schiller (ed.), Instinctive Behavior, New York, International
Universities Press, 1957.
Woods, D. D. (1988).
Coping with complexity: The psychology of human behavior in complex systems. In
L.P. Goodstein, H.B. Andersen, and S.E. Olsen, editors, Mental Models, Tasks
and Errors, Taylor &
Francis, London, (p. 128-148).
Woods, D. D. (1993).
Process-tracing methods for the study of cognition outside of the experimental
psychology laboratory. In G.
Klein, J. Orasanu, R. Calderwood, and C. E. Zsambok, C. E. (Eds.) Decision
making in action: Models and methods
(pp. 228-251).Norwood, NJ: Ablex.
Woods, D. D. (1994).
Observations from Studying Cognitive Systems in Context. In Proceedings of
the Sixteenthp Annual Conference of the Cognitive Science Society, August
1994.