Discovering Complexity
Discovering Complexity
复制标题
发现复杂性
DOI：
10.1007/bf03039614
复制
发表时间：
1996
影响因子：
2.5
通讯作者：
L. Darden
中科院分区：
文献类型：
--
作者：
L. Darden
Bechtel and Richardson say in their preface that they set out to write a book about alternatives to philosophers’ formal model of derivational theory reduction. Along the way they reconceptualized the problem, and the notion of formal theory reduction all but disappeared from their book. We are fortu-nate that this change occurred because, as a result, they have produced a much more creative and useful book. They focus on strategies for theory change and theory development aimed at constructing “mechanistic explanations" to account “for the behavior of a system in terms of the functions performed by its parts and the interactions between these parts”(p. 17). This view, they claim, is “what scientists understand by the term reduction"(p. xi, italics theirs). Thus, their goal is to find problem-solving strategies for explaining the behavior of a system. They propose an ordered set of strategies, from simple decomposition of the system into component parts, followed by localization of the behavior in a single component, to the postulation ofsuccessively more complex interactions among the pans of the system in the production of the behavior.This book adds to thetrend in philosophy ofscienceofdiscussing strategies or heuristics for discovery and theory construction (Darden l99l, Ch. I5; Kleiner l993, Ch. 6; Schaffner 1993, Ch. 2; Wimsatt 1980; 1987), in contrast to analyses of theory structure or irrational theory change. Also, Beclttel and Richardson’s work draws on the work of Herbert Simon and others who view science as a problem-solving activity and scientific problem-solving as a special case of general, human problem-solving (p. 6). Such a view of science naturally leads to a discussion oflleuristics for discovery. As students of Wimsatt, Bechtel and Richardson also focus on the issue of biases or systematic failures introduced by using particular heuristics. For example. they recognize that attempts to localize a function in rt single part often fails, because no single part alone produces the function. Such a failure of simple
