第1章：系统建模-Why and How?
1.1 THE SYSTEMS APPROACH AND HOLISM
At the core of building good models in operations management and the management sciences is the systems approach. This has a critical role because of three fundamental features.
1. It recognises systems as wholes - unlike the reductionist approach, which presumes that understanding is best obtained by breaking the wholes down into their constituent parts. Ackoff (1993) commented that “it can be proven rigorously that when each part of an organisation improves its performance independently of other parts with which it interacts, the performance of the whole may well suffer. This follows from the fact that the performance of a system is not the sum of the performances of the parts taken separately, but is the product of their interactions”.
2. It considers organisations as open systems in constant interaction with their environment.
3. It is naturally comprehensible because human beings organise their knowledge into cognitive systems spontaneously (Jackson, 1991).In any model building, though, it is necessarily impossible to include all the elements of a given situation: there are too many and the interactions between them are too complex. The trick is thus to recognise the elements and their interactions not completely but adequately: this is often far from straightforward to achieve. Particularly in the early stages of modelling (Clarke and Tobias, 1995a; 1995b) we need ways of dealing with complexity (Flood and Carson, 1993) and to recognize why we are doing it. As illustrated in Figure 1.1 we will be hoping that, through informed discourse and debate, our modelling - and more generally, problem structuring and analysis – will result in rational decision making and interventions, and hence outcomes that are “desired” and somehow make the world a “better” place.
1.2 DESCRIPTION, METAPHOR AND ANALOGY
The approach in much management research is just to describe. Situation descriptions are framed in long and complicated textual terms, and are intended to provide a complete coverage of every possible feature of the situations without quantifying many (if any) of the properties involved. Indeed some organisational models “do little more than describe or depict” (Burke and Litwin, 1992). The trouble is that assessments of particular real situations against such models tend to be at best debatable and at worst arbitrary.
More succinct than descriptions are metaphors, and these differ from analogies and from similes. Analogies reflect “correspondences between things in certain respects that are otherwise different” (Chambers Etymological English Dictionary) - e.g. “Money is like muck, not good unless it be spread” (Fowler’s Modern English Usage). Unlike similes (which are comparisons proclaimed plainly as such), metaphors are, in conversation at least, mere figures of speech based on comparisons that are tacit - e.g. “He is a tiger when roused” (Chambers).
In management though, metaphors tend to assume the role of models and the use of metaphorical analysis in discussing organisational change has long been widespread (Grant and Oswick,1996; Palmer and Dunford, 1996). The "organisation as machine" metaphor can actually be traced back to Taylor (1911) but more recent studies tend to rely upon studies of biological systems for insights into the management of organisations (e.g. Waldrop, 1992). Indeed it has become popular, when attempting to understand almost anything in management, to refer to the animal kingdom (e.g. Collinvaux, 1978).