Working paper Open Access
There are two reasons not to consider the whole of an object.
Either the object is global i.e. "big" : global warming, global human exchanges, etc.
Or the relationship between the researcher and his object has qualitative constraints.
What will be met, exchanged, perceived, described is only partial, made of "islands of knowledge".
We describe the D&R method = dimensions & referentials.
We make the point about multidisciplinary versus multi-referential.
We take a didactic and appealing example with Cervantes' Don Quichote.
We show how the method was used for a PhD dissertation.