Abstraction, Coordination, and Situational Awareness: Implications for use Centered Design

Two strategies that scientists use to reduce the complexity of the natural world are reductionism and abstraction. Reductionism deals with complexity by describing the system in terms of a small number of fundamental elements. In physics the search for fundamental particles and in psychology the search for independent information processing stages or for local neural mechanisms reflect this strategy for attacking complexity. Abstraction deals with complexity by searching for global relational properties that exist somewhat independently of the elements whose behavior they govern. In physics, the idea of “fields” (e.g., gravitational or electromagnetic) are prototypical examples of the strategy of abstraction. In psychology, the cybernetic hypothesis, that human behavior may be constrained by the same kind of stability constraints as a servomechanism is an example of abstraction. This is an abstraction in the sense that within a closed-loop system stability is not associated with any particular element of the system. Rather, stability depends on global relations between forward loop gain (sensitivity to error) and time delays. Recent work using nonlinear dynamics to model phase transitions in rhythmic behaviors such as walking also reflect the strategy of abstraction (e.g., Kelso, 1995). Both reductionism and abstraction have proven value as strategies of science. However, some problems are better approached using one strategy or the other. A fundamental commitment of our research program is that problems of situation awareness and situated cognition that require human operators to adapt to changing demands and contexts in dynamic environments are fundamentally problems of coordination. As problems of coordination, we believe that they are best attacked using the strategy of abstraction. An important tool for implementing this strategy will be the state space. This talk will demonstrate how the field of constraints on information and action can be visualized as boundaries and regions in state space. Recent empirical work on collision avoidance in low altitude flight will be used to illustrate the utility of the state space framework. Finally, the focus on constraints in state space will be linked to the recently articulated framework of “Use Centered“ Design (Flach & Dominguez, 1995)