The Textual Dynamics of Disaster

As Petroski concludes, trial and error is fundamental to a notion of engineer-ing experience that builds upon past events in order to design structures that will solve problems in the future. In order for engineering experience to provide a reliable index of future action, engineers must assume that responses in the future will be predictable that actions in the past can predict future consequences and thus, that particular features of the material environment can act as predictable indices of future actions. MSHA investigates accidents to identify the technical cause (or causes) of accidents so that agencies can prevent similar occurrences in the future. But engineering experience must be adapted to new sites and situations in order to provide a warrant for action. 

Gate conditions steadily worsened as panel extraction advanced toward the third panel. Excessive abutment loading on the gate, due to poor overburden caving and a relatively low pillar support capability, resulted in severe floor heave, a deteriorating roof mass, and pillar dilation to the extent that cribs 

Severe and unpredictable natural conditions provide some of the most difficult challenges that longwall planners face.. . . Solutions which work at one mine can seldom be applied, ctKikbixik style, to other operations. None of the specific designs Idescribed in the paper] has an inherent superiority. Sound engineering, based on past-experience and site-specific knowledge, is needed to develop optimal solutions, (p. 37) 

Despite this dynamic uncertainty, engineers must construct plans to control risk and predict hazard in the everyday working operations of the mines. Roof-support engineers, for example, must map underground faults, determine stress and load levels, and construct plans that outline the minimum support methods and requirements for each section of the mine. These support requirements include a broad plan for overall roof support as well as site-specific rules for pillar width and additional support. But engineers must also be prepared to respond to change. 

TTiirty-four other non-fatal, injury type roof fall accidents were identified in which rock fell from between roof bolts striking persons standing under permanently supported roof during mining.. .. Remote-control continuous miner operators, miner helpers, section foremen, and others observing the mining procedure were involved in these accidents. This type of accident, however, is not necessarily related to the depth of cut. From the information in the record, and discussions with roof control supervisors in the districts involved, an accurate determination could not be made as to whether extended cuts were used at the time of the accident [Italics added].  

---

The present analysis focuses on representations of roof support because roof control represents one of the most dangerous and uncertain aspects of safety in a coal mine. To help readers understand the uncertainties inherent in roof control, I have provided a brief overview of three methods of roof support in U.S. and British mines. This overview will set the stage for a more theoretical discussion of the nature of warrants grounded in experience, the effect of warrants on risk decisions and risk outcomes, the rhetorical incompleteness of written instructions and procedures, and finally, the textual dynamics of disaster. In the conclusion, I look at the implications of this analysis for writers and rhetorical theorists. 

---