Brian Bosse wrote:
Hello Alliop,



It seems problematic to me that in order to know what objects belong to the set called 'good' we must already know in some sense what makes something good. Otherwise, we have no basis to make distinctions as to what objects belong in the set and as to what objects do not belong in the set.

In spite of the above objection, this approach is probably what is most often used. Consider, for example, the field of epsitemology and the attempt to come up with a clear definition of knowledge. Some definitions have been rejected on the basis that the proposed definition precludes some things that we consider knowledge and includes some things we do not think is knowledge. As you can see, to make this judgment we must already have some idea as to what knowledge is.



If applied consistently, this would make things very explicit. However, it does not correspond well to human experience.



This is a good point.



This depends on the circumstance. If you are using a strictly formal system, then there is no interpretation and everything follows from the axioms (defintions) and rules of inference. However, when someone attaches an interpretation to the system (models), then things are different. If the model does not correspond to what you are trying to capture (that which you know intuitively perhaps), then adjustment of the definitions (axioms) may be called for. However, it may be that the axioms are good and that the non-intuitive results are correct.

Brian