Pedantic
Legend
It definitely is impressionistic, but I think the most useful way to understand that is to look for how it breaks. What causes players concerned with "realism" to find pieces of the fiction dissonant or to reject them? You're not going to get to a universal principle that way, but I think you can approach a reasonable set of norms.Right, that makes sense. In terms of what I was asking it'd fall under reasonable inference from fiction and genre. This seems to be more than not purporting to be grounded in scientific reality but, as you put it, encouraging incorporation of fantastical features.
The mechanism of this doesn't yet seem entirely clear. Picture an array of fictional facts, some as yet not appearing or untested, others extant or in mind. An intuition I've been pursuing is that the former are not fantastical until in joining the latter they are said to be so. That they be fantastical is not resisted, but neither is it automatic: a myriad of details are accepted as being just like they are in the real world until they are modified (as encouraged.)
I think something like that likely because although we can easily pull out fantastical facts -- dragons etcetera -- in truth they are the minority of facts implied in play. Of course the minutiae is largely elided, but when we do scrutinise it the real world supplies the common experience which is our go to. I suppose in saying so, you might remark that a scientific appreciation of those details isn't encouraged (suggesting, were it not already obvious enough, the presence of tacit principles that guide us) so the "grounding in real world" may be impressionistic.
I tend to find consistency and causality are most important. You can set exceptions to the common understanding of reality, so long as they remain consistent, even as the circumstance changes, and so long as they don't break the flow forward in time. Casting a spell is a cause that leads to an effect, the scope of spell effects is knowable, if not known.
Parsimony is important, as you note. That anything can be changed doesn't establishing the expectation that any specific things will be changed, and established fantastic facts are subject to repetition and testing; once established they continue to be normative and must be reliable as norms.