On Making the Abstraction Hierarchy More Abstract

Gavan Lintern

General Dynamics
Dayton, Ohio


Work Domain Analysis has a special role in cognitive engineering. It is directed at describing the workplace in functional terms and does not actually describe cognitive activity or cognitive events (which is Work Analysis or Cognitive Task Analysis). However, there are numerous possibilities for describing a work domain and if Work Domain Analysis is to be of value, the resulting knowledge representation must reflect or relate to the structure of cognitive work. This particular point, although rarely discussed, is key to understanding what Work Domain Analysis is about. Guidance on how to construct and interpret the Abstraction-Decomposition map (the representational product of Work Domain Analysis) is typically based on definitions but, as with all definitions taken out of context, these do little to resolve the ambiguities. Even cognitive engineers with an impeccable pedigree in work domain analysis (a pseudonym for that constellation of cognitive engineers who have been students or close associates of Kim Vicente) are inconsistent in their use of the terms.

Here I argue that Work Domain Analysis has an important role within cognitive engineering but it needs a coherent cognitive and/or design theory to stabilize that role. I will outline some elements of a suitable theory and discuss the extent to which we can claim that it is well enough established to support this endeavor.