Skip to content
jie zheng edited this page Aug 14, 2015 · 2 revisions

Current issue

When a group of people met to discuss BFO2 in december 2012, it was proposed (and IIRC accepted) that we use proper parthood as our parthood primitive, with the rationale being that this is what most people meant, and that reflexive parthood, while making some axioms simpler, in general led to query results that were unintuitive and undesired by users. Therefore I don't subscribe to my previous comments, below, anymore. However I don't think it is beneficial to include both proper and improper parthood in BFO, for reasons of simplicity.

Previous Conversation

Alan wants to deprecate this. What's the justification?

someone said: Is the plan to make part_of ==> proper_part_of? I think this would make a lot of mereological axioms harder. Consider the role chain for overlaps

Alan: I looked over the ro properties, saw proper_part_of, and didn't recall anyone actually using it. In the interest of slimming BFO I made a note suggesting we consider dropping it from BFO main. I imagine we can have additional BFO-based modules, but think it would be good if we trimmed BFO to what seemed essential.