Issue 275: Space primitive

ID:
275
Starting Date:
2015-04-27
Working Group:
3
Status:
Open
Background:

In the 32nd joined meeting of the CIDOC CRM SIG and ISO/TC46/SC4/WG9 and the 25th FRBR - CIDOC CRM Harmonization meetingdiscussing about the name of E93, we discuss that maybe we need a place primitive analogously to time primitive. The E53 Place should not be contiguous. It may consist of finite sets of non-contiguous areas. The CRM-SIG decided to introduce a place primitive. This is an assignment for Øyvind.

commented by Gerald Hiebel on 1/5/2015

On 1. mai 2015, at 23:12, Gerald Hiebel wrote:

> Dear all,
> After a Skype with Martin we ended with a much more elegant Version of Space and Spacetime Primitive getting rid of all new properties, by making Space Primitive a subclasses of Place in the sense of a declarative Place and Spacetime Volume Primitive a subclasses of Spacetime Volume in the sense of a declarative Spacetime Volume. Comments welcome,
> Best,
> Gerald

commented by Oyvind on 7/5/2015

This is an interesting solution. It is a stretch for me for the following reason:

A place in the CRM sense is an extent in space, e.g., on the face of the earth.

A declarative place is a place (an extent in space, e.g., on the face of the earth) which is defined by a language expression, such as a set of coordinates. The coordinates are not the declarative place, the declarative place is rather the real world place specified by the coordinated. In that sense a declarative place is a type of place.

A space primitive is a language expression. It is, e.g., a set of coordinates. These coordinates can specify a place. But what is suggested now is that the coordinates themselves -- the space primitive -- isa place.

How can you explain to a stubborn theoretician that this is not confusing a thing with its name?

Commented by Martin Doerr on  7/5/2015

Dear All,

May be I have also confused things. The question is, if multiple syntactical forms of the same
mathematical construct would be regarded as identical or not by a knowledge base. Then either
we need a property of a (declarative) place to be "defined by" a place primitive, and then a
phenomenal place, possibly spatial projection of a STV, may fall within the former.

The analogy of the Time-Span would mean that the "at some time within" is a sort of "falls within"
between a phenomenal and a declarative Time-Span.

May be a better solution than declaring the primitive IsA Place or Time-Span would be to regard
"at some time within" a shortcut through a not represented declarative time-span, and do an
analogous thing with Place. Then we would need an explicit construct for place and time-span,
an "is defined by: Primitive"

What do you think?

Commented by Eide on 10/5/2015

Instinctly this sounds like a better solution.

Proposed by Christian Emil on  10/5/2015

Dear all,
In the scope note of Exx Space Primitive the term  'double instantiation'  is used. It should be 'double inheritance'.

A dummy's question may be: The examples of Exx Space primitive could have been used in the scope note of Place Appellation. What use do I have of a Space Primitive? Compute distances?

Ex1 Spacetime Primitive:
A spacetime volume can be imagined/is  as  4 dimensional object. The spatial extent is a function of the time dimension. The first example has a static spatial extent. The second may have the same kind if the GML is not parameterize with time.

Spacetime volume have the properties

P160 has temporal projection: E52 Time-Span
P161 has spatial projection: E53 Place

E52 Time-Span P81 ongoing throughout: E61 Time Primitive

Is there any property between a E53 Place and  Exx Space Primitive or is the subclass of identifiable places?
Ditto for Spacetime volume

I see that this may be what you, Martin,  may mean by the last paragraph in your email from May 7th?

Martin, could you please elaborate a little on  what you mean by formulation below, especially what you mean by a mathematical construct in this context, is it for example
3+7
20-10?

commented by Martin on 10/5/2015

On 10/5/2015 2:51 μμ, Christian-Emil Smith Ore wrote:
>
> I see that this may be what you, Martin,  may mean by the last paragraph in your email from May 7th?
>
> Martin, could you please elaborate a little on  what you mean by formulation below, especially what you mean by a mathematical construct in this context, is it for example
> 3+7
> 20-10?
>
> "The question is, if multiple syntactical forms of the same mathematical construct would be regarded as identical or not by a knowledge base"
Yes, this is what I meant, more practically, encoding differences like "2" or "2.0" or "2.00000000000 e0"

M

Current Proposal:

It is proposed  by Gerald Hiebel  about Space Primitive and Spacetime Primitive   on    13/5/2015

Dear All,
I tried to adapt the proposal to your inputs (I hope that I understood them in the right way), thanks very much for them!
Any suggestions, comments, improvements very welcome,

Dear Gerald,

This looks good to me. Two smaller comments:

1) Given the definition of E53 Place: in Exx Space Primitive the concept of “geometries on or relative to earth” should be extended to include things like coordinates on a ship, places on the moon, and areas of a manuscript page.

2) “GML” in Ex1 Spacetime Primitive: this is polysemous and should be written out in full.

Proposed by Gerald Hiebel  on 15/05/2015

Dear Øyvind,
Thanks for the very good remarks. I will include them and send it start of next week.

Posted by Gerald Hiebel on 18/5/2015

Dear Øyvind, all,

I put your comments in and added a sentence how things like a ship, places on the moon, and areas of a manuscript page can be explicitly modelled through the property P157 is at rest relative to going from the defined declarative place to the Physical Thing. I was unsure if this belongs to the scope note of Space Primitive, please delete as appropriate.

Thank you very much for your scope note on E4 Period, that’s great!

There is one sentence I had trouble understanding the “it” in the first read:

However, the real spatiotemporal extent of an instance of E4 Period is regarded to be unique to “it” due to all its details and fuzziness

Reading several times it seems more natural and I did not come up with a different formulation either.

Best,

Gerald

In the 35th joined meeting of the CIDOC CRM SIG and   28th FRBR - CIDOC CRM Harmonization meetingthe crm-sig changed the scope note of E94 Space Primitive, P168 place is defiend by(defines place).

Also new classes and properties have been introduced, these are E95 Spacetime primitive, P169 defines spacetime volume (spacetime volume  is defined by), P171 at some place within , P172 contains.

These changes have been incorporated into the CIDOC CRM version 6.2.2

Also, it is pointed out that revision of the scope note of E61 Time Primitive is needed, and the addition of a new construct about  declarative time span which is a result of a time primitive is needed too.

Prato, February 2016

posted by George Bruseker on 28/11/2016

Perhaps my contribution will be a bit confusing with regards to this issue. I had an old version of this homework which didn’t have the same detail as this last version.

Anyhow, I noted that there were some minor updates to be done to p171 and p172 in order to spell out the long paths of which these properties are shortcuts. I did this and marked it in yellow. It seems a minor textual change unless I did not understand what I was supposed to do.

I also made a go of fleshing out the scope note of E61. Not knowing what the issue was about except that it related to declarative vs phenomenal I tried to rewrite the scope note to match somewhat to E94 upon which, to my understanding, it is now more or less modelled. I also reintroduced the text about the fact that the instances of E61 are not meant to be modelled explicitly in CRM which seemed pertinent although perhaps it was removed for some good reason.

Anyhow, I’m sure Christian Emil can do a much better job, but if any of this text helps, please take it.

To me the paragraph to be added, “Note that it is possible for a time-span to be defined by phenomena causal to it or other forms of identification rather than by an instance of E61 Time Primitive. In this case, this property must not be used for approximating the respective instance of E52 Time span with an instance of E61 Time Primitive.” is of itself is highly obscure without further explanation or context. I wouldn’t know what it was referring to or how it should affect my modelling. Perhaps an example of how not to use or what to do instead?

In the 37th joined meeting of the CIDOC CRM SIG and ISO/TC46/SC4/WG9 and the 30th   FRBR - CIDOC CRM Harmonization meeting,  the crm-sig discussed this issue and made the following proposals and decisions:

(a) assigned to GB to improve his proposal about the scope note of E61 according with the following recommendations:
– Check the proposed scope note (2nd paragraph) is consnisten with the   properties P81 and P82, and if they can be referenced to.
– Check against E59 and if this is not expressed there, this should be moved here.
– Consider if E61 is a declarative time – span (Linked to the overall issues of E52 either an approximation or declarative in sense of plan)
(b) the reference note in the scope note of P169 defines spacetime volume (spacetime volume  is defined by) should be removed.
(c)approved the changes in the scope note of P171, P172 and since in  common documentation practice it might be needed to be expressed  a place that is indefinitely related (P171, P172), assigned to  Gerald to formulate a  question/issue about this in order  to be discussed in the next meeting.

Berlin, December 2016

Posted by George on 1/4/2017

Here is my effort for 275, revise my revision of E61 Time Primitive.

In the 38th joined meeting of the CIDOC CRM SIG and ISO/TC46/SC4/WG9 and the 31st FRBR - CIDOC CRM Harmonization meeting, the crm -sig

(a) reviewed the E61 scope note and  assigned to George to  revise it  according to the  comments and to  the overall issues of E52 Time Span either an approximation or declarative in sense of a plan

(b)  discussed about the missing reference of P169 decided the following:

·        To postpone the reference to an extension from core

·        The reference should be removed from the text

·        To consider again the scope note of P169 under the notion of phenomenal  vs declarative space

(d)  approved the proposed changes on the scope note of P171, P172

(e)  assigned to Gerald to write up a new issue to be discussed in  the next meeting about places that are indefinitely related in common documentation practice.

Details of the accepted changes may be found here

Heraklion, April 2017

Sent it by Gerald on 6/10/2017

I wrote up a proposal for a new Issue about places that are indefinitely
related in common documentation practice and included also a proposal for a new property for this relation.

In the 39th joined meeting of the CIDOC CRM SIG and ISO/TC46/SC4/WG9 and the 32nd FRBR - CIDOC CRM Harmonization meeting,  the sig reviewed the scope note provided by GH about  Pxxx approximates  and decided to accept in principle but need to add a .1 property before installed in standard and to look at how it relates to measurement.  The revised text is here

Heraklion, October 2017

In the 40th joined meeting of the CIDOC CRM SIG and ISO/TC46/SC4/WG9 and the 33nd FRBR - CIDOC CRM Harmonization meeting, the sig  assigned to MD to create the .1 property. GB, OE SS and others will give examples of the actual practice of having approximate locations, since this will allow us to check if the accuracy should be a property on property or if the approximation is related to the event itself. In addition, the crm-sig decided that this property is no longer necessarily accepted for CRMbase, it should  be determined if should go in CRMgeo or it may still go in base.

Cologne, January 2018

posted by Ovind on 1/5/2018

Dear all,

Our job was to find some ”examples of the actual practice of having approximate locations.” Based on the border protocols I have worked with, and in line with other similar documents, I will present some core categories. If needed I can provide specific examples and translate them to English

- at/within: Something is said to be ”at” something else, which usually means within the area of. A house is ”at” a farm. If it is, for instance, a small mine it could also be said the be within the [land of the] farm.+

- between: This is a common category, for instance, the hill between two lakes. The area is closed by the two things the target is between, but undefined/open in other directions.

- [direction] of: For instance, north of. The area is bound in one direction and not in the other. It will fan out in a undefined with. Context will often tell what the relevant area is, but can be hard in historical sources.

- either/or: It is not common but not very rare either that one finds explicit alternatives. ”Either A or B is the border mountain.” ”The farm is either east or west of the lake.”

As a remark: in my best reading of my 18th century material the only exact locations were based on shared denotation, that is, the reader/listener was assumed to share with the speaker/writer an understanding of what a place name referred to. This method is often imprecise but highly functional, especially within smaller cultural groups. Then additional constructs (as the ones above) were used to be more precise or to refer to unnamed places.

In the 41st joined meeting of the CIDOC CRM SIG and ISO/TC46/SC4/WG9 and the 34th FRBR - CIDOC CRM Harmonization meeting the sig decided the following:

• The vocabulary of .1 property could be characteristics  of interest such that… ‘close by / around / centroid’. Potentially could be any place preposition.HW assigned to CEO to work on this.
• We should have two .1   properties for quantification and one for qualificationIn addition, these could be the scope of an approximation property of greater complexity  which could be part of CRMgeo.
• That which cannot be expressed in the approximation vocabulary can be placed in a note. This note should be on the place that is approximated (domain place and NOT range place).
• Related: rules should be placed in CRM about where to put a note. If you don't know where to find them, then this will not be productive/efficient/useful. MD proposes new issue in the guidelines using CRM, where we may put notes. E.g.: someone involved in architectural project and we want to denote more info about this involvement… the note should not go on the person but on the event note (it should be assigned to the most characteristic node.). MD will do the HW.
• The provided examples are accepted as a vocabulary.
• The property will be added in CRMbase. with the name PXX.1 has type: E55 Type

Lyon, May 2018

In the 42nd joined meeting of the CIDOC CRM SIG and ISO/TC46/SC4/WG9 and the 35th FRBR - CIDOC CRM Harmonization meeting, the crm-sig reviewed the scope note by MD for a .1 property on PXX ‘is approximated by’ –namely PXX.1 ‘has type’: E55 Type. MD proposed that the approximations the relevant .1 property evokes (eg. by a bounding box, footprint, polygon, centroid, etc.) be made part of CRMgeo rather than CRMbase –the latter already being overpopulated with material from geo. Seeing as CRMbase must close within this year, the crm-sig resolved to discuss the .1 property on P189 approximates on their next meeting.

Berlin, November 2018