This morning’s reading has lead me to a nice quote:
Model building is the art of selecting those aspects of a process that are relevant to the question being asked – J.H Holland*
What I like about the quote, is that it not only highlights the principle of parsimony (as the Einstein quote did), but it highlights that the question being asked is the element of the scientific problem that should be referenced to determine if an aspect of the model should be kept in or kicked out.
In a world where we might identify ourselves as a landscape ecologist, a toxicologist, or even an expert in neural networks – consider this: there are unlikely to be any discipline specific guides to parsimonious model building. And my reason for wanting to catalog the different types of questions was that this could, possibly, serve as a useful framework; where the same types of questions share the same types of guiding principles regarding how best to achieve parsimony.
Holland, JH (1995) Hidden Order. Addison-Wesley, New York, USA.