Post by Dominic Oldman (10 December 2025)
Dear SIG,
The scope note for E24 says,
"This class comprises all persistent physical items of any size that are purposely created by human activity. This class comprises, besides others, human-made objects, such as a sword, and human-made features, such as rock art. For example, a “cup and ring” carving on bedrockis regarded as an instance of E24 Physical Human-Made Thing."
Is this right/misleading?
If it includes objects then why can't they be moved? The note includes items that might be considered objects - they are usually defined in E22 or E18 - items which have "physical boundaries that separate them completely in an objective way from other objects." This explains the difference between a carving on a wall and a movable object. If the carving is cut out of the wall then it gets sound physical boundaries and can be moved.
Cheers,
Dominic
Post by George Bruseker (10 December 2025)
Dear Dominic,
Yes this is as it should be. This class is the super set of the human made object and human made feature. As such its instances include both of its child classes’ instances. It represents what they share in common which is essentially being a physical kind of thing and being the kind of thing made by humans. It is also stated by the ontology that this then is from where you can begin to speak of representations. According to crm representations are only made by humans.
So if you need to talk about things that are movable you hop down to e22 and if you are needing to make statements about things that are features hop down to e25.
E24 is a class that likely isn’t invoked much directly but rather serves to support the representation of some things that are common in its child classes.
Linked.art takes the decision to not split the hairs about whether a thing can be moved or not (since ultimately anything likely could be moved with a little imagination) and uses e22. But for some e25 serves useful purposes for indicating the physical objects that inhere in other objects.
Is that helpful or addressing the direction of your question or did you have something else in mind?
Best
George
Post by George Bruseker (10 December 2025)
Maybe a more useful way to say it is that it’s a class that allows you to talk about either a physical thing that is movable or that is not but does not yet allow you to talk about its relations such as being movable or not. For that further detail you need a more precise class.
Post by Dominic Oldman (10 December 2025)
Yes, my point was not about the difference but more about the scope note. The reason I brought this up is because I saw an example of the confusion and it made me wonder whether this came from a reading of the scope note.
D
Post by Christian-Emil Ore (10 December 2025)
Dear both,
I early times before the first ISO version (pre 2004) the name of the class E70 Thing was E70 Stuff. It was changed because, according to Nick Crofts, the term 'stuff' could not be translated properly into French. If you check CIDOC CRM version 3.4.9 you will see that. Personally, I think that was a bad decision. Shakespeare writes "The stuff dreams are made of". "Thing" is a bad choice since it may be confused with "Object".
*in reply to:
Dear Dominic,
Yes this is as it should be. This class is the super set of the human made object and human made feature. As such its instances include both of its child classes’ instances. It represents what they share in common which is essentially being a physical kind of thing and being the kind of thing made by humans. It is also stated by the ontology that this then is from where you can begin to speak of representations. According to crm representations are only made by humans.
So if you need to talk about things that are movable you hop down to e22 and if you are needing to make statements about things that are features hop down to e25.
E24 is a class that likely isn’t invoked much directly but rather serves to support the representation of some things that are common in its child classes.
Linked.art takes the decision to not split the hairs about whether a thing can be moved or not (since ultimately anything likely could be moved with a little imagination) and uses e22. But for some e25 serves useful purposes for indicating the physical objects that inhere in other objects.
Is that helpful or addressing the direction of your question or did you have something else in mind?
Best
George
Post by Franco Nicolucci (10 December 2025)
Although I like "stuff" more than "thing" because it sounds more generic, although they are more or less synonyms, here is a comment from the Cambridge dictionary (see here for the complete text):
Thing
We use the general noun thing more commonly in speaking than in writing.
It is most commonly used to refer to physical objects, but we also use thing to refer to ideas, actions and events
...
Stuff
Stuff is one of the most common nouns in speaking. It is more informal than thing. It is not at all common in writing.
...
In conclusion, there should be another word for the matter, as "thing" may convey a dubious meaning and "stuff" is too colloquial and of difficult translation in other languages ; in doubt, "thing" looks better to me.
In Italian I would use the term "cosa" which applies both to physical and to immaterial things and translates either English terms; probably the same happens to "chose" in French.
Franco
Post by Dominic Oldman (10 December 2025)
The issue was just about the scope note and "objects and sword". I saw an example of E24 being used with E9 and I wondered whether the scope note was misleading.
E24 says:
"This class comprises, besides others, human-made objects, such as a sword, and human-made features, such as rock art."
Thanks,
Dominic
Post by Christian-Emil Ore (10 December 2025)
I think this is not the only scope note where one refers to instances that should better be modelled by a sub class. On the other hand it is a leading principle that one should not refer explicitly to subclasses in the definition of a class. Maybe this is just an ideal and not followed in practice?
C-E
*in reply to:
Yes, my point was not about the difference but more about the scope note. The reason I brought this up is because I saw an example of the confusion and it made me wonder whether this came from a reading of the scope note.
Post by Thanasis Velios (10 December 2025)
Also complexity increases when you have an object E22 which then becomes a feature E25 in another object, e.g. a plank of wood becomes part of a laminated board for a manuscript.
A new identifier would be necessary after the part addition event for the resulting feature. People (including me, regularly) get confused with the classes at that point as the new identity for what seems to be the same thing feels counter-intuitive.
All the best,
Thanasis
*in reply to: The issue was just about the scope note and "objects and sword". I saw an example of E24 being used with E9 and I wondered whether the scope note was misleading.
E24 says:
"This class comprises, besides others, human-made objects, such as a sword, and human-made features, such as rock art."Thanks,
Dominic
Post by George Bruseker (10 December 2025)
It feels like a pedagogical issue. The person _may_ have only read the one scope note and felt that probably that works and done it like that.
Obviously that's complex to figure out how to help the learner make better choices.
One long standing ongoing effort is to update the 'use and learn' material and specifically the diagrams describing aspects of the model. This is the issue here:
We are actually very advanced now thanks to many collaborative contributions to having a full updated set of these diagrams that match 7.1.3. They are done in draw.io and use the libraries developed by CHIN.
What we lack however are good texts for describing the diagrams.
In the particular case that you are looking at there are two diagrams:
Part and Component Information and Location Information which when published on the site and accompanied by useful pedagogical text might be a help to the learner.
We are looking for volunteers to help write these texts.
Another interesting tool available is the draw.io validator created by FORTH. If the learner had their data in draw.io they could validate it and find that they had chosen a wrong property (I think). Of course this could be done with many different tools!
*in reply to: I think this is not the only scope note where one refers to instances that should better be modelled by a sub class. On the other hand it is a leading principle that one should not refer explicitly to subclasses in the definition of a class. Maybe this is just an ideal and not followed in practice?
C-E
Post by Martin Doerr (10 December 2025)
Dear All,
May I remind you that we have deliberately decided to use instances of subclasses as examples, because it is the only way to make people aware what a superclass is. Otherwise, they typically regard it as a complement of the subclasses. The typical questions we have heard dozens of times: "what is the difference between a physical thing and a physical object?" etc. If indeed the E-number of the most adequate subclass is not in parenthesis behind the example, this should be corrected.
Since Dominic's question is about the movability, I propose to add to the scope note, of E18 actually, that :
- "Many instances of this class cannot be moved as a whole independently from surrounding or attached matter, or be detached along well defined boundaries in all dimensions, such as the Temple of Abu Simbel before its transfer. Therefore, the class E9 Move applies only to adequate subclasses as targets of a move."
Further, the scope note of E18 states:
"This class comprises items of a material nature that are units for documentation and have physical boundaries that separate them completely in an objective way from other objects. "
I suggest to add that:
- "Items that have been permanently attached or incorporated into another material environment, but could in principle be detached along the natural boundaries it had before, may still be regarded instances of E18Physical Object. Note that only instances of this class can be target of an instance of E9 Move. If moving requires separating at substantially new boundaries, such as the Temple of Abu Simbel when it was moved, a new instance of E18 Physical Object is created that incorporates the previous item. The latter process may be documented by the class E81Transformation."
I believe such a wider interpretation of the natural boundaries would be useful. Incorporating physical objects permanently without loosing their natural boundaries will be a simple application of E79 Part Addition and the deduction P46 is composed of.
I agree with George that we have a didactic issue here.
The problem Thanasis points out may be solved by offering more specific subclasses, e.g., of both E80, E81 and E12 for directly saying in one activity that the original temple was removed and been made a modified item which was moved. In particular, the specific form of continuity of matter that goes beyond E81 and allows for scientific inferences from the current state back, may be analyzed and modelled more specifically.
This may be a separate ISSUE.
Note that the intuitive concept of incorporated or broken off parts can be very deceiving wrt to the reality we talk about and the reasoning a KB system may apply. We have made many years ago extensive studies of the semantics of such "parts".
See
- Bekiari, C. and Doerr, M. (1999) ‘Documentation and reasoning on parts and potential wholes’, in Fennema, K. and Kamermans, H. (eds.) Making the connection to the Past: Computer Applications and Quantitative Methods in Archaeology (CAA99), Dublin, Ireland, 14–18 April. Leiden: Faculty of Archaeology, Leiden University.
- Doerr, M., Plexousakis, D., & Bekiari, Ch. (2001). A Metamodel for Part and Whole Relationships for Reasoning on Missing Parts and Reconstruction. Yokohama, Japan, November 27-30 (pp. 412-425).