Closely related to the 'separation' problem is the problem of drawing the system's boundaries. By narrowing the boundaries of the system considered, many of indirect effects may be reduced to the effects attributable to the variation in external forcing. This is often convenient, in particular, if the aim is to study responses in a controlled environment. However, the scope for investigations of indirect effects in a system with narrow system boundaries inevitably becomes limited; by widening the boundaries and including more components the potential for indirect effects is greatly enhanced, but as the system's complexity increases, so does the associated uncertainty of any interpretations!
Was this article helpful?