Patrick Waterson, Chris Baber and Chenling Li, Loughborough, Birmingham and Beijing Jiaotong Universities
We consider the manner in which ‘systems’ thinking continues to influence Human Factors methods (Salmon et al., 2016). The question we wish to address is what the benefit of using the notion of ‘systems’ offers if it is removed from a theoretical foundation; it is simple to put boundaries around a set of objects and activities and call this a system but this does not, of itself, lead to better understanding of the behaviour of these objects. Key to systems thinking is the notion of self-organisation and, we argue, this is not something which Human Factors methods address explicitly. In order to explore this we compare and contrast the STAMP methodology (Leveson, 2004) with Stafford-Beer’s Viable Systems Model (VSM, Beer, 1972).
Both approaches describe systems in terms of interacting components, both view these interactions as causal and dependent. However, VSM is also concerned with how the system self-organises itself to achieve a state of stability through aligning its internal complexity with external complexity of the environment in which it operates. The question is whether similar concerns are fundamental to STAMP? Using a case study of the 2011 China-Yongwen Railway Accident, we consider how STAMP and VSM could be used to make sense of the events leading to the accident.
This not only provides an opportunity to compare the assumptions and methodologies underlying these approaches but also raises broader issues surrounding the concept of causality in Human Factors research. In other words, is an accident a ‘state’ of the system rather than a random arrangement of factors.
From this, a secondary question relates to when an accident begins, i.e., how far back in the preceding states of the system can be accident’s ‘roots’ be traced? We relate our findings to other debates within the field of accident analysis and complex systems which include the role played by emergent system properties, recursion and positive/negative feedback in accident causation.