|
|
This is a limited free preview of this book. Please buy full access.
development of new ones, and where they appear to exert significant influence to the structure of the argument, these too would be woven into the model. Horowitz similarly argued that if an assumption does not significantly improve a model, it would be reasonable to drop it in order to make the model less restrictive. If the assumption also exerts significant enough influence on the conclusion, the supposition may itself warrant further study: ‘If the model suggests that an assumption is invalid even though it is normally taken for granted, this too will be a useful output of the model’.14
Models and Nested Models
The preceding argument suggests that a model could be multidimensional and multileveled, consisting of different parts, each part in turn generating new sets of information, which, as a result, require new levels of analysis. One way to achieve this degree of sophistication in analysis is to restrict the study of a phenomenon to a narrowly defined parameter, such that a group or an organisation, for example, is referred to as a special case. For instance, it could be assumed that the subject of the investigation possesses characteristics that make it unique if compared to similar entities. The idea is to draw conclusions from a range of variables, which in some cases may not be present if analysis is restricted to a particular field of discipline. The process of developing this kind of model need not be exhaustive, as it needs only to examine the defining characteristics of the variables in each case.
In reality, it is not possible to draw up a comprehensive inventory of such variables. This is because some of the variables we seek to understand or include in the model may not be immediately obvious at the point of analysis, or they may be nested elsewhere in the same model to the extent that their existence may not be immediately recognisable, and their effect on the outcome may be neglected.15 Elements of collective action could exist outside of the known or established pattern, and their manifestation could be missed if analysis is focused on the demands of the existing set of rules.