Issue 635: property quantification mismatches

ID: 
635
Starting Date: 
2023-02-06
Working Group: 
3
Status: 
Done
Background: 

Post by Wolfgang Schmidle (6 February 2023)

 

Dear All,

There are some properties where the text version and the number version of their property quantification do not match. Since both should express the same information, either the text version or the number version (or both) must be wrong. Martin and I have worked to resolve these cases.

Results:

P10  many to many, necessary, dependent (1,n:1,n)
P81  many to many, necessary (1,n:0,n)
P89  many to many, necessary, dependent (1,n:1,n)
P99  one to many, necessary (1,n:0,1)
P161  many to many, necessary (1,n:0,n)
P187  one to many, necessary (1,n:0,1)
P188  many to many (0,n:0,n)
P191  many to one, necessary (1,1:0,n)
P198  many to many (0,n:0,n)

Details see here:
https://docs.google.com/spreadsheets/d/1p4Ew-vKigz5Lm3H69nrlbk2Xfe8gioun87COJmKFfXY/edit#gid=307539608

Best,
Wolfgang

Current Proposal: 

In the 56th joint meeting of the CIDOC CRM and ISO/TC46/SC4/WG9 &49th FRBR/LRMoo SIG, the SIG reviewed and approved the property quantifications proposed by WS & MD. 

The set of properties concerned immediately follows. Details of the new quantifications can be found in the attached document: 

  • P10 falls within (contains) [D: E92 Spacetime Volume, R: E92 Spacetime Volume]
  • P81 ongoing throughout [D: E52 Time-span, R: E61 Time Primitive]
  • P89 falls within (contains) [D: E53 Place, R: E53 Place]
  • P99 dissolved (was dissolved by) [D: E68 Dissolution, R: E74 Group] 
  • P161 has spatial projection (is spatial projection of) [D: E92 Spacetime Volume, R: E53 Place]
  • P187 has production plan (is production plan for) [D: E99 Product Type, R: E29 Design or Procedure]
  • P188 requires production tool (is production tool for) [D: E99 Product Type, R: E19 Physical Object]
  • P198 holds or supports (is held or supported by) [D: E18 Physical Thing, R: E19 Physical Thing]
     

Concerning property P191 had duration (was duration of) [D: E52 Time-Span, R: E54 Dimension]: there’s an ongoing discussion among WS & MD on this. Once they have reached a decision, they will start an evote through the listserv. 

Τhe correct quantifications will appear on the definition of CIDOC CRM v7.2.3.

Nb. Request that these edits make it to the ISO version –contact PM & EC to take care of this. 

 

Crete, May 2023
 

Post by Martin Doerr (13 September 2023)

 

Dear All,

Regarding as last item of issue 635 the property P191 had duration:

Issue 559 was resolved as:

"In case the instance of S15 Observable Entity is more specifically an instance of E18 Physical Thing, using the property O12 has dimension (is dimension of) is equivalent to using the property P43 has dimension (is dimension of). In other words, using the one implies the other."

Sadly, P43 has quantification "one to many, dependent (0,n:1,1)", but we use O12 now for relative dimensions between multiple things  in CRMsci, AND we use Dimension iP191 had duration (was duration of) from a Timespan, which is not an instance of E70 Thing, with quantification "one to one (1,1:1,1) ". Note that P191 is NOT a subproperty of P43, but E54 Dimension pertains to both.

It appears to me that P43 should have quantification "many to many (0,n:0,n)" , and P191 should have quantification "one to one (1,1:0,1)" , but this leaves P43 without the important semantics of dependency.

Taking relative dimensions into account, it should be clarified that an instance of E54 Dimension is dependent on the combination of references to it. This is a task for an FOL or so, isn't it? Otherwise, we would need to specialize E54 in CRMbase, not really nice.

Opinions?

Best,

Martin
 

Post by Wolfgang Schmidle (12 October 2023)

 

Dear Martin,

Just trying to understand:

Time-Span P191 had duration E54 Dimension:
was "one to one (1,1:1,1)", which is contradictory, and both are wrong

in this issue it was tentatively (1,1:0,n):

(1,1:_,_): Martin: I think the phenomenal time-span has exactly one duration, which can be given with inner/outer bounds. The declarative one has exact durations.

(_,_:0,_): Not all dimensions are durations of a time-span.

(_,_:_,1): Does a duration always belong to a single time-span?  It seems that although P191 had duration is not a subproperty of P43 has dimension because of their incompatible domains (E52 Time-Span versus E70 Thing), the logic should be the same, i.e. a dimension belongs to only one thing, and in particular, a duration belongs to only one time-span. (However, see P179 had sales price.)
(_,_:_,n): Martin: In the case of declarative time-spans, we do not have the identity of the infinitely complex real thing. So, multiple time-spans could have the same duration. Isn't it?

Now you suggest (_,_:_,1): Is that a return to my argument, or just coincidence?

Thing P43 has dimension E54 Dimension:
was (0,n:1,1), this issue did not touch it

now you suggest (_,_:0,n):

(_,_:0,_): What would be an example of a dimension that is not associated to anything?

(_,_:_,n): The resolution "559 [52nd sig-decisions].docx" still says that O12 has (0,n:1,1), which would be compatible with P43. Is that outdated?

Best,
Wolfgang

Outcome: 

At the 57th CIDOC CRM & 50th FRBR/LRMoo CRM SIG meeting, the group reviewed issue 635 and decided to close it, on the grounds that it had given rise to two subtopics, namely: 

  • The idea that a dimension can pertain to multiple timespans, only holds if these timespans are identified by temporal primitives. 
    • This can be conveyed in CRM through the property quantification (1,1:0,n), taken together with an FOL statement relaying that the condition that n only pertains to declarative timespans. 
    • The FOL constraint should be rendered in prose in the scope note. Redrafting the scope note of P191 to be handled in a separate issue (664)
  • The property quantifications of P43 has dimension, O12 has dimension (and possibly P179 had sales price) have incongruent semantics with the specific case of relative dimension. Harmonizing them is to be explored in a separate issue (665). 

Marseille, October 2023