What Does Systems Thinking Add to Mechanistic Explanation?

William Bechtel, Department of Philosophy and Center for Chronobiology, University of California, San Diego

Mechanistic researchers in biology have developed powerful techniques for isolating phenomena, associating them with specific mechanisms, and decomposing these mechanisms to identify their parts and operations. In many areas of biology researchers have succeeded in identifying many, if not all, of the components of mechanisms, and determining what operations they perform and how the operation of each part affects other parts of the mechanism. When mechanisms are organized sequentially and the operations are linear, it is not hard for scientists to rehearse in their heads how the mechanism will generate the identified phenomena. But often what researchers have found is that organization in biology is not sequential and the operations are not linear. Moreover, many biological phenomena turn out not to be regular but to exhibit complex dynamics. Mental rehearsal is no longer sufficient to determine how such mechanisms behave and researchers have applied or developed methods to represent graphically and computationally the relations between operations in such mechanisms and simulate how they work together to produce the mechanism’s behavior. I will describe some of these graphical and computational techniques and argue that they are becoming ever more important as researchers discover that mechanisms they had investigated separately are in fact highly interactive and modulate each other’s operation. I refer to explanations that integrate such graphical and computational with the products of traditional mechanistic research as dynamic mechanistic explanations. Moreover, I will argue that graphical and computational techniques, just as identifications of parts and operations, pick our real features of mechanisms (their spatial and temporal organization) and are integral to the explanations.