[Sorry for the length; there's a lot to talk about.]
The definition of 'has member of located in' reads:
An instance of an Object Aggregate 'has member of located in' an instance of some material entity if and only if every member of that Aggregate is located in the same instance of that material entity.
An initial (minor) issue here is the use of single quotation marks. Since these are not used in the other definitions, I doubt they should be present here.
The main problem I want to focus on, though, involves the use of ‘instance of’. (Note: ‘instance of’ is also used in many other definitions in ERO, but not all these uses are similarly problematic. I discuss this below.) Suppose X has member of located in Y. The definition then indicates that X is an instance of an object aggregate, A, and that Y is an instance of some material entity, M. However, object aggregates and material entities more generally don’t have instances because they aren’t classes (or universals), and hence the definition is unsatisfiable.
The same problem arises in the definitions of ‘process started by’, ‘affects’, ‘process starts’, ‘is temporal region of’, and ‘is affected by’, and a related one arises for ‘caused by’ and ‘is cause of’.
I think my preferred way to fix this problem would be by revising the definition of ‘has member of located in’ as follows
An Object Aggregate has member of located in some material entity if and only if every member of that Aggregrate is located in that material entity.
and making similar changes to the other definitions mentioned above.
Worth noting that there are other definitions in ERO that unproblematically make use of ‘instance of’. For example:
x aggregate_bearer_of y iff x is an instance of Object Aggregate and y is an instance of Specifically Dependent Continuant and z is an instance of Object, such that z bearer of y, and all other members of x are bearers of a unique instance of the same type as y.
Here, due to the lack of an article before ‘Object Aggregate’, that phrase can be taken to refer to the class of object aggregates rather than to a particular object aggregate (as in the ‘has member of located in’ case). (This, of course, suggests an alternative way to revise the problematic definitions.)
However, while its use in some definitions is unproblematic, I suspect it would be better if ERO didn’t use ‘instance of’ at all. For notice that none of the definitions for object properties in BFO use that phrase, nor do definitions from the other Common Core Ontologies. (As an example of the latter point, the definition for ICE begins with ‘A Generically Dependent Continuant’ rather than ‘An instance of Generically Dependent Continuant’.)
[Sorry for the length; there's a lot to talk about.]
The definition of 'has member of located in' reads:
An initial (minor) issue here is the use of single quotation marks. Since these are not used in the other definitions, I doubt they should be present here.
The main problem I want to focus on, though, involves the use of ‘instance of’. (Note: ‘instance of’ is also used in many other definitions in ERO, but not all these uses are similarly problematic. I discuss this below.) Suppose X has member of located in Y. The definition then indicates that X is an instance of an object aggregate, A, and that Y is an instance of some material entity, M. However, object aggregates and material entities more generally don’t have instances because they aren’t classes (or universals), and hence the definition is unsatisfiable.
The same problem arises in the definitions of ‘process started by’, ‘affects’, ‘process starts’, ‘is temporal region of’, and ‘is affected by’, and a related one arises for ‘caused by’ and ‘is cause of’.
I think my preferred way to fix this problem would be by revising the definition of ‘has member of located in’ as follows
and making similar changes to the other definitions mentioned above.
Worth noting that there are other definitions in ERO that unproblematically make use of ‘instance of’. For example:
Here, due to the lack of an article before ‘Object Aggregate’, that phrase can be taken to refer to the class of object aggregates rather than to a particular object aggregate (as in the ‘has member of located in’ case). (This, of course, suggests an alternative way to revise the problematic definitions.)
However, while its use in some definitions is unproblematic, I suspect it would be better if ERO didn’t use ‘instance of’ at all. For notice that none of the definitions for object properties in BFO use that phrase, nor do definitions from the other Common Core Ontologies. (As an example of the latter point, the definition for ICE begins with ‘A Generically Dependent Continuant’ rather than ‘An instance of Generically Dependent Continuant’.)