Intentional systems representations are useful alternatives to physical systems representations of fire-related human behavior

Abstract

Fire safety engineering analyses have been mostly confined to describing how the physical environment causes the events, including human behavior, that comprise a fire incident. Despite the use of cognitive and information processing constructs to explain behaviors, the person remains a “black box” that generates behavioral responses when exposed to physical stimuli in the environment. These “physical systems representations” are well-suited to the prediction of some human behaviors, for example, travel times constrained by physical crowding. However, physical systems representations are ill-suited to the task of modeling cognitively-derived human behaviors, because the mediating effects of peoples' intentions result in behavioral responses that are poorly correlated with objective measures of environmental stimuli. Examples of such behavior are delays in initiating evacuations when people are exposed to alarm signals and other ambiguous signs of fire. Because research based on physical systems representations neglects to investigate how individuals use information in their pursuit of goals, the resulting models are often complicated, lack parsimony, and do not generalize well to new settings. This paper discusses how “intentional systems representations” can describe the cognitively derived responses of people in a more parsimonious manner, and can be used to design fire safety systems that capitalize on the adaptive abilities of people. (A fire safety system is viewed as being comprised of all relevant components, including people, that play significant roles in the defense against a fire threat.) In an intentional systems representation, events are driven by the goals that people pursue. Using the framework of physical versus intentional systems representations, various models of fire-related human behavior are reviewed and evaluated. Also discussed is recent research by Groner and Williamson (1998, Scenario-based goal decomposition: a method for implementing performance-based fire safety analysis. In: Proceedings of the Second International Conference on Fire Research and Engineering, 200–211) to develop an approach, called “Scenario-Based Goal Decomposition,” whereby the performance of fire safety systems is analyzed by integrating physical and intentional views into a general systems approach. A pilot study is discussed that used this approach to describe and evaluate the predicted response of a fire safety system to a representative high-challenge fire scenario at a university library. It is also argued that systems approaches that integrate physical and intentional systems representations axe the only way to fully realize the potential of performance-based codes.