Issue 574: Scope note/range clarification for E80 and P112
Post by Erin Canning (29 Nov 2021)
Hello,
I am hoping you might be able to help me with a small confusion -
The scope note for E80_Part_Removal states that "This class comprises the activities that result in an instance of E18 Physical Thing being decreased by the removal of a part." This reads to me as if the relationship would then go: E80 > P112_diminished > E18, as P112 creates the connection to the thing being diminished (having something removed from it), as opposed to P113_removed, which is for the connection to the thing that was removed.
However, the range of P112 is E24, not E18, and the scope note for P112_diminished reads "This property identifies the instance E24 Physical Human-Made Thing that was diminished by an instance of E80 Part Removal."
Meanwhile, the range of P113_removed is E18, as is the range of P31_has_modified, the superproperty of P112.
It seems to me therefore that either I am reading the scope note incorrectly (very possible!) or that there is an inconsistency between the two, perhaps in the range of P112?
I looked in the Issues history for anything about this, and wonder if the discussion around the change of P31 (Issue 191) might have relevance to either the range or language around it, as in that case the range of P31 was relaxed from E24 to E18. Although, that being said, this perceived E18/E24 inconsistency as described above exists as far back as v4.0, the earliest version available on the website, so perhaps it is my reading of the scope note that is backwards...
To summarize, my question is - Which is the correct range of P112: E18 as stated in the E80 scope note or E24 as stated in the P112 scope note and range; or am I reading this set of notes/relationships incorrectly?
Thanks for your guidance on this,
Erin Canning
Post by Thanasis Velios (29 November 2021)
Hm, yes, this is confusing. We might need a new issue to update the scope note. I think the correct class is E24 as it seems that E80 Part Removal does not cover cases such as cutting a stalactite off in a cave.
Thanasis
Post by Rob Sanderson (29 November 2021)
Good spotting! I agree with Thanasis that there is any issue, but I think that the range is wrong for P112, which I would argue should also be E18.
For example, I find a tree and break off a branch. The tree is not a Human-Made Thing, it's an E20 Biological Object. Or I break a piece of obsidian (I would argue an E19) into two. Or if the obsidian is part of a lava flow, then it would be a physical feature ... and thus we end up at E18 as the common ancestor.
I think that E18 remains correct for P113, given the described use of removal of a part from an E78 Curated Holding. If I remove a meteorite fragment from the collection of a natural history museum, the meteorite is (again, I would argue) an E19. Now ... can it ever be an E18 Physical Thing but not an E19 Physical Object? It can't be a Feature, as they cannot be removed, ruling out E26 and below. However E78 is an E24 Physical Human-Made Thing, but not an E19 Physical Object. If we use E78 to model sub-collections, and sub-collections can be removed from their parent, then yes, here is a case where we need E18.
Rob
Post by Thanasis Velios (29 November 2021)
Yes this is a logical position. I guess the way I have been reading it is that the object that is diminished was indeed intentionally made by a human and therefore it *can* be diminished. If it is any thing then who judges if it is complete and has been diminished? There is no agency on its original "production".
The reason we have E18 as range is because the removed item's identity is not that of a human made object. I.e. Part removal is not a Production which I think is the reason the following sentence is in the scope note:
"In cases where the part removed has no discernible identity prior to its removal but does have an identity subsequent to its removal, the activity should be modelled as both an instance of E80 Part Removal and E12 Production."
hence the removed part is pushed up to E18.
T.
Post by Rob Sanderson (29 November 2021)
Isn't "diminished" just a label, rather than a value statement? I don't think something needs to be complete for a part to be removed (I can break a twig off the branch, which I previously broke off the tree). I read it as "was made smaller by" in that some part was removed.
I agree that Part Removal is not always a Production -- the same part could be added and removed several times, and clearly not all of them are Productions. [Personally, I would never say it was a Production, but that a Production might have a Part Removal as a part of it]
Here's a related question... Can an E78 Curated Holding have a Physical Feature as part of it? Conceptually, yes ... but E78 is a physical aggregate, not a conceptual thing. Does the collections system of Arches National Park have records for the rock formations? Surely it must. And if the National Park boundaries were changed, then the arched rock formations that no longer fell within the protected area would have to be removed from the E78. So I think I even retract "you can't remove features" given the physicality of E78.
Rob
Post by Thanasis Velios (30 November 2021)
Hm, I do not consider it as a value statement, but as indication of the intension. Breaking a tree branch which is worth putting in your collection is a production of a Human-Made Object (as well as a Biological Object). So when you break the twig off the branch, you are removing a part from a Human-Made thing. In other words you cannot remove a part unless it is included intentionally on the original thing in the first place. Does this make sense?
T.
Post by Rob Sanderson (30 November 2021)
It makes sense for the twig from the branch, but not from the branch from the tree (or stalactite from the cave, fragment for study from a meteorite, etc etc).
Removing the branch/fragment from the tree/meteorite results in a Human-Made Object (via the Part Removal / Production), but the tree/meteroite that is P112 diminished by the activity does not become a Physical Human-Made Thing in the process. It stays an E20 Biological Object/E19 Physical Object, just a smaller one.
R
Post by Thanasis Velios (30 November 2021)
I completely understand the reasoning and I agree that intuitively a tree with a broken branch is a diminished thing. It is just that the scope note and all of the examples in E80 Part Removal are for Human-Made things so I worry that the class has been designed for Human-Made things only, i.e. breaking off the original branch may not be E80. Part Addition and Part Removal are designed to allow us to track the use of a component integrated intentionally in multiple objects over its history, so it may be that a thing needs to be added before it can be removed. If we care about the tree prior to cutting the branch then it may be only a modification. Am I taking it too far?
Having said that, pushing the property higher in the hierarchy, although I am told we should avoid it in general, in this case it may not cause too many problems.
T.
P.S. Amazingly, the inconsistency between the scope note and property range existed since version 3.4.
Post by Martin Doerr (5 December 2021)
Dear All,
Actually the class was also designed for cutting parts from archaeological objects, natural history stuff etc. We had a long discussion if, in the very instant, a part is broken from a natural object, e.g. for sampling, the diminished becomes "human made". We later ultimately decided that this violates identity criteria of classes. It just leaves a human-made feature on a natural object.
Therefore, we need to revise wherever this logic had been applied before.
Best,
Martin
Post by Thanasis Velios (13 December 2021)
In which case I suppose the proposal to discuss at the next SIG is:
1) change the range of P112 from E24 Physical Human-Made Thing to E18 Physical Thing
2) fix the reference to the property under the scope note of E80
3) add an example to E80 and a corresponding example to P112 for non-man-made things.
Could we assign a new issue number to this?
All the best,
Thanasis
Post by Thanasis Velios (25 January 2022)
Dear all,
It turns out that we might also need to worry about P110. The HW for both is included here to discuss and vote at the next SIG:
1) Change the range of P112 diminished:
From: E24 Physical Human-Made Thing
To: E18 Physical Thing
And update the property scope note
from: "This property identifies the instance E24 Physical Human-Made Thing that was diminished by an instance of E80 Part Removal. Although an instance of E80 Part removal activity normally concerns only one instance of E24 Physical Human-Made Thing, it is possible to imagine circumstances under which more than one item might be diminished by a single instance of E80 Part Removal activity."
to: "This property identifies the instance E18 Physical Thing that was diminished by an instance of E80 Part Removal. Although an instance of E80 Part removal activity normally concerns only one instance of E18 Physical Thing, it is possible to imagine circumstances under which more than one item might be diminished by a single instance of E80 Part Removal activity."
2) Update property under the scope note of E80 Part Remove
From: P112 diminished (was diminished by): E24 Physical Human-Made Thing
To: P112 diminished (was diminished by): E18 Physical Thing
3) Example for E80 diminishing a natural object
the removal of the Porite coral specimen from the Cocos Islands by Charles Darwin in April 1836
4) Example for P112 diminished
The coral of the Cocos Islands (E20) was diminished by The removal of the Porite coral specimen by Charles Darwin (E80).
Refs: https://data.nhm.ac.uk/object/e1bfb1ab-e94e-4e0a-a13c-bc54e03f22e5 https://www.nhm.ac.uk/discover/charles-darwin-coral-conundrum.html
Extra HW for P110:
1) Change the range of P110 augmented:
From: E24 Physical Human-Made Thing
To: E18 Physical Thing
And update the property scope note
from: "This property identifies the instance of E24 Physical Human-Made Thing that is added to (augmented) in an instance of E79 Part Addition.
Although an instance of E79 Part Addition event normally concerns only one instance of E24 Physical Human-Made Thing, it is possible to imagine circumstances under which more than one item might be added to (augmented). For example, the artist Jackson Pollock trailing paint onto multiple canvasses."
To: "This property identifies the instance of E18 Physical Thing that is added to (augmented) in an instance of E79 Part Addition.
Although an instance of E79 Part Addition event normally concerns only one instance of E18 Thing, it is possible to imagine circumstances under which more than one item might be added to (augmented). For example, the artist Jackson Pollock trailing paint onto multiple canvasses."
2) Update Class E79 Part Addition:
Reference to property P110:
From: P110 augmented (was augmented by): E24 Physical Human-Made Thing
To: P110 augmented (was augmented by): E18 Physical Thing
Scope note update:
From: "This class comprises activities that result in an instance of E24 Physical Human-Made Thing being increased, enlarged or augmented by the addition of a part."
To: "This class comprises activities that result in an instance of E18 Physical Thing being increased, enlarged or augmented by the addition of a part."
3) Example for E79 augmenting a natural object:
the carving of the Culpa Dendroglyph on the Culpa tree (Buhrich et al., 2015)
4) Example for P110:
The carving of the Culpa Dendroglyph (E79) augmented the Culpa tree (E20). (Buhrich et al., 2015)
Ref: https://www.tandfonline.com/doi/abs/10.1080/03122417.2015.11682048
Looking forward to comments and the discussion.
All the best,
Thanasis
Post by Martin Doerr (29 January 2022)
Dear Thanasi, all,
I agree with all, except:
3) Example for E79 augmenting a natural object:
the carving of the Culpa Dendroglyph on the Culpa tree (Buhrich et al., 2015)
4) Example for P110
The carving of the Culpa Dendroglyph (E79) augmented the Culpa tree (E20). (Buhrich et al., 2015)
I' argue that this example is a production of a human-made feature ex initio on the tree. I'd argue that the meaning of E79 is that a pre-existing thing has been added. Otherwise, it comes in conflict with production, and the tracing of things that become part of another and then travel with it through the world.
I propose to modify the scope note of E79 to make this clear. I think cases in which the P111 added thing is not a "Physical Object" can only be sort of collections, in which the definition of the whole under consideration is expanded to comprise another feature, such as real estate properties.
The removal is not completely symmetric. It says that something has been removed, but the removed matter may have a unique identity only from the time of removal on, and then should be also a Production event.
The inverse, a part addition in which the added part looses its identity within the whole it augmented (and then be a destruction event??) may probably be too exotic (Frodo's Ring not withstanding).
All the best,
Martin
Post by Thanasis Velios (29 January 2022)
Yes, this is a fine point and I struggled to find an example for such a case of P111. However there is also this:
https://www.demilked.com/bronze-hand-squeezed-trees-sculpture-giuseppe-p...
which is rare but matches the case?
T.
In the 52nd CIDOC CRM & 45th FRBRoo SIG meeting,the SIG decided to change the range of P112 diminished from E24 Physical Human Made Thing to E18 Physical Thing, on the grounds that removing a part from a natural object, does not render the latter a human made thnig instead.
A series of minor changes will be implemented as a consequense of the change of range to E18, namely:
- the scope note of the property P112 (to reflect the change of range)
- the FOL inference (it should point to E18)
- add an example involving a physical thing that got diminished as a result of a part removal
- E80: edit the range of the referred property (P112 to E18)
- E80: add an example of part removal that involves a natural object (i.e. not E24, but E18 instead).
The SIG also decided to change the range of P110 augmented from E24 to E18 on the grounds that adding some object to a natural thing, does not render the lattern human made instead.
A series of minor changes will be implemented as a consequense of the change of range to E18, namely:
- the scope note of the property P110 (to reflect the change of range)
- the FOL inference (it should point to E18)
- add an example involving a physical thing that got augmented as a result of a part addition
- E79: edit the range of the referred property (P110 to E18)
- E79: add an example of part addition that involves a natural object (i.e. not E24, but E18 instead).
These changes are to inform BOTH v7.1.2 (ISO) and v7.2.1.
Last, the SIG decided to start a new issue where to discss what counts as an instance of E79 Part Addition
A summary of the decisions can be found here.
February 2022
In the 53rd CIDOC CRM & 46th FRBRoo SIG meeting, the SIG decided to formally close the issue on the grounds of all the points raised in it havng been addressed. All the changes have informed CIDOC CRM v7.1.2 and v7.2.1.
May 2022