Disambiguation between statue and sculptural group

[dcapillae (Daniel Capilla)]

I propose to abandon this discussion to return to the main topic of this thread. I’m going to try asking you directly about a question.

I am interested in knowing how a native English speaker would refer to a sculpture of the type I am here referring to as a “sculptural group”. @Kovoschiz proposed some alternative denominations: statue set, statue group, sculpture group. I understand that it would be most convenient to use the most common English denomination.

Assuming someone would use a tag to identify this type of sculptures (no one is proposing to create it unilaterally), what value would you use?

[Matija_Nalis]

@dieterdreist

I would question this definition. Seems it has been copied over from wikipedia in 2018 and should be redacted They have fixed it in the meantime in WP, but not in OSM

Statue - Wikipedia

Hm, I don’t see that change neither on ID you’ve quoted, nor near it, not in latest revision. The current revision seems to say:

A statue is a free-standing sculpture in which the realistic, full-length figures of persons or animals are carved or cast in a durable material such as wood, metal or stone

That “s” in “figures”, “persons”, “animals” denotes plural, and not singular. Also, many of the example pictures in that wikipedia article clearly show multiple persons and animals (e.g. “Hermes and the Infant Dionysus”, “Laocoön and his Sons”, “Moai of Easter Island”, “The Burghers of Calais”, " Marcus Aurelius" etc.)

But that is not even the point how the OSM wiki historically got its definition. What is important is that OSM wiki has been used as a reference in all that years since. So people and editors have been using it by that (plural) definition.

For example, StreetComplete memorial quest picture for “statue” does not show “single person/animal” but two of them, so practically anybody who solved that quest (at last several thousands of changesets last time I looked) were guided by that picture.

@dieterdreist

how many of them will likely be wrong sfter we change the definition? I would not expect so many, but maybe I’m wrong. Not sure if people typically choose the value “statue” from what they expect it to mean (a statue) or by reading the wiki definitions.

I don’t know even how to estimate. Maybe 20-30%? Perhaps more in older countries. Most statues that comes to mind here are persons on horses, or persons slaying dragons, or religious or historical events ones which are mostly multiple persons. Single persons are way more likely to be artwork_type=bust instead (over here).

@dieterdreist

Yes, they would generally become a little bit less reliable

Yes, but the point of proposal to change things should be to make things better, and not worse (even if damage done could’ve been much worse if even worse idea was implemented, it’s not an excuse to deliberately go damaging data with excuse “yeah, but I could’ve damaged it much worse, so it is OK”)

Especially as there are several available alternatives (mentioned before, so I won’t copy, unless you can’t find them) that avoid that problem completely, and only improve by adding such details to the data. Some of them are even currently being used (by just adding extra details, instead of damaging existing data and making it more ambiguous)

[Matija_Nalis]

@dcapillae

I propose to abandon this discussion to return to the main topic of this thread. I’m going to try asking you directly about a question.

Very well, as a sign of goodwill, I will for now close my eyes to non-answers of that last monologue of yours (even if my fingers itch to respond ). Thanks for choosing to invest time on actual tagging proposition.

@dcapillae

I am interested in knowing how a native English speaker would refer to a sculpture of the type I am here referring to as a “sculptural group”.

Searching for that knowledge is well and fine (although “sculptural group” and the like would likely only be used by art majors, not random English-speaking person on the street, nor OSM mapper).

@dcapillae

Assuming someone would use a tag to identify this type of sculptures (no one is proposing to create it unilaterally), what value would you use?

This however, is false dichtomy. There are way more than 4 ways to “disambiguate between statue and sculptural group”, and you should’ve included at least “something else entirely” option in your poll. This way, vote is rigged as all the answers modify artwork_type=* values in some way (which is your desired outcome).

I’ll create a less problematic poll.

[Matija_Nalis]

Here is better poll, which compiles (hopefully all) currently suggested options:

Note: statue_group value may be replaced by any of the preferred alternatives (like sculpture_group, statue_set, sculptural_group etc) – this poll is only about what method is most acceptable, not what exact name of key/value pair should be. After the method is decided, most fitting value can be decided later.

• artwork_type=statue + statue=statue_group (preserves current meaning of artwork_type=statue meaning both single and multiple people/animals, and only adds details indicating there are multiple of them depicted. statue=* is already used 706 times)
• artwork_type=sculpture + sculpture_type=statue_group (Sculpture is more generic term which includes both single statue and statue groups. This only adds details indicating there are multiple people and/or animals depicted. sculpture_type=* is already used 102 times)
• artwork_type=statue + description=sculpture group depicting xxxx and yyyy (advantage: does not make current tagging ambiguous, description key is wildly supported, can add even more data then just single person / multiple persons qualification)
• artwork_type=statue_group (Warning: makes all existing artwork_type=statue ambiguous! Advantage: only invents new value, not a new key)
• artwork_type=statue + sculpture_group=yes (advantage: does not make current tagging ambiguous, adds machine-readable detail)
• artwork_type=statue + statue:people_count=2 + statue:animal_count=1 (advantage: does not make current tagging ambiguous, adds machine-readable detail in even more detail)
• artwork_type=statue + wikidata=Q107338575 (Q107338575=sculptural set. Advantage: wikidata is already used, adding it does not make current tagging ambiguous, adds more or less detail as wanted via both human-readable and machine queryable structured wikidata)
• Something completely different is needed (please add post describing your idea)

[dcapillae (Daniel Capilla)]

@Matija_Nalis

Searching for that knowledge is well and fine (although “sculptural group” and the like would likely only be used by art majors, not random English-speaking person on the street, nor OSM mapper).

I understand. I am mapping artworks in Malaga (Spain). My point of view is not that of an art expert, although I try to go a little deeper than the average person on the street. That is why I am interested.

@Matija_Nalis

There are way more than 4 ways to “disambiguate between statue and sculptural group”, and you should’ve included at least “something else entirely” option in your poll.

I know. My question is open-ended, you can answer whatever you want. I linked the poll just to remind you that it exists.

My interest is to know how you would name these sculptures in English, not what tag to use. After posting the poll I realized that it would have been better not to use tags and just include values. I didn’t want to modify the poll once it was published. Changing a poll once it has been started could be interpreted as an attempt to misrepresent the results, so I left it at that.

Feel free to answer whatever you like. Including a personal choice in a comment is also fine.

[Kovoschiz]

I mistakenly voted when I wanted to click the links…
statue:people:count= gives precise info. It has an added advantage of showing this is counting the number of characters, not some number of statue depending on the definition of this. But we don’t always want to add a number. So something to show a multiple is still needed before (unless you accept eg statue:people:count=multiple), keeping in mind they are different levels of details, not mutually exclusive.
As said, currently I prefer generic =sculpture + sculpture= first. However further to the above, my other point from the beginning is =artwork needs to be defined in different situations.

1. A fused statue is certainly 1 =artwork
2. What about when the characters on a common platform but separate?
3. When they are only near each other on the pavement?
4. A pair of statues on the sides of a walkway, gateway, or something

Until that and type=site or =cluster are solved, this can’t be represented consistently.

[Matija_Nalis]

@Kovoschiz

I mistakenly voted when I wanted to click the links…

You can click on Show votes / Remove vote to undo your mistaken vote (as well as re-cast it again for some other option if you changed your mind).

[Kovoschiz]

Oh, I thought votes can’t be changed sometimes. Forgot the option is buried in the “Show results”.

[Matija_Nalis]

@Kovoschiz

As said, currently I prefer generic =sculpture + sculpture= first

So that would be second option from the top. Note that specific names of key/values pairs don’t matter at this stage (as noted at the top of the poll) - e.g. you would choose that option even if you preferred sculpture=statue_group instead of sculpture_type=sculptural_set

[dcapillae (Daniel Capilla)]

Thank you for your collaboration!

I have already commented on my opinion of all but two of the options proposed in this poll.

Summarizing:

The artwork_type=statue_group option conflict with artwork_type=statue. The ambiguity in the current use of the keys sculpture, sculpture_type and statue make me think that they are not appropriate. The counting of statues, I have also explained why it does not seem to me appropriate for differentiating this type of sculptures.

Regarding the new options, using the wikidata code would be giving a solution to the problem outside of OpenStreetMap. And regarding the sculpture_group=yes option, would be a new key to be defined.

[Matija_Nalis]

@Kovoschiz

1. A fused statue is certainly 1 =artwork

agree.

2. What about when the characters on a common platform but separate?

I’d say the same. But if people want to micromap them as separate statues and connect them via relation, it is also fine with me.

3. When they are only near each other on the pavement?
4. A pair of statues on the sides of a walkway, gateway, or something

I’d use an OSM relation. Although alternative is give each of them unique wikidata=Q* tag, where all those are instances of parent wikidata key which denotes whole group.

Until that and type=site or =cluster are solved, this can’t be represented consistently.

Those are tags on OSM relation, right? I forgot to add that in the poll, as relations seems to me as an orthogonal issue to this one which covers your points 1,2 (and maybe 3) - i.e. you could have relations of multiple “architectural groups” too, and relations of relations.

[dcapillae (Daniel Capilla)]

@dcapillae

Reducing the difference between a statue and a sculptural group to the number of statues is a conceptual error. Sculptural groups are not just sculptures representing more than one person/animal. They are also the composition or action represented, the interaction between the figures. They are not a random number of statues placed together in one place. They are a singular artwork in their own right, different from a statue.

I share the definition of “sculptural group” included in the Dictionary of Cultural Heritage of Spain:

El grupo no debe ser confundido con la estatua que se compone de una única figura en bulto redondo. A veces, en escultura, ciertos grupos no se consideran más que como estatuas, en particular cuando algunas de las figuras representadas son de pequeñas dimensiones o cuando no se apoyan sobre el suelo. Así, la Virgen con el Niño se designa comúnmente como estatua y no como grupo. Uno de los más famosos grupos escultóricos es el del Laooconte, descubierto en Roma en 1506. El grupo representa al sacerdote troyano Laooconte junto con sus dos hijos, envueltos por dos serpientes que habrían surgido del mar durante la celebración de una ofrenda a Neptuno.

Machine translation:

The group should not be confused with the statue which is composed of a single figure in round bulk. Sometimes, in sculpture, certain groups are only considered as statues, in particular when some of the figures represented are of small dimensions or when they do not rest on the ground. Thus, the Virgin and Child is commonly designated as a statue and not as a group. One of the most famous sculptural groups is that of the Laooconte, discovered in Rome in 1506. The group represents the Trojan priest Laooconte together with his two sons, enveloped by two snakes that would have emerged from the sea during the celebration of an offering to Neptune.

Example 5. Laocoön and His Sons by Hagesandros, Athenedoros, and Polydoros. Source: image by Jean-Pol Grandmont (CC BY-SA 3.0) available on Wikimedia Commons.

[Matija_Nalis]

@dcapillae

Thus, the Virgin and Child is commonly designated as a statue and not as a group. One of the most famous sculptural groups is that of the Laooconte, discovered in Rome in 1506. The group represents the Trojan priest Laooconte together with his two sons

Interesting, that makes it even more complicated. Does it go into detailed reasoning?

in particular when some of the figures represented are of small dimensions or when they do not rest on the ground

Why are two persons (Virgin and Child) seen one single statue, and three persons (Laooconte and 2 sons) and sculptural group? Is it then not about number of persons, but instead of how the figures are connected (or not) by material/base? And their size?

[dcapillae (Daniel Capilla)]

@Matija_Nalis

Why are two persons (Virgin and Child) seen one single statue, and three persons (Laooconte and 2 sons) and sculptural group?

I imagine it is because some images of Virgins are always represented with the Child. The Child would be part of the usual iconography of the Virgin.

[Matija_Nalis]

@dcapillae

I imagine it is because some images of Virgins are always represented with the Child. The Child would be part of the usual iconography of the Virgin

Hmm, while it is somewhat common, there are many instances where she is not having Child.

More convincing sounds the argument from your translation of other persons/animals “not touching the ground” (but being connected only to the body of “main” person, which might explain equestrian statues e.g. person-on-horse being statues even if there are two of them).

However cursory research reveals that it seems that many sculptures of Virgin and the Child are still considered statues instead even if they fulfill that requirement of only one touching the ground and other being small (e.g. Virgin and Child from the Sainte-Chapelle - Wikipedia)

@dcapillae

In the case of Laooconte and His Sons, the sons are not part of Laooconte’s usual iconography.

Conflictingly, wikipedia for Laocoön and His Sons - Wikipedia seems to call it “The statue of Laocoön and His Sons”, and talks about “The figures in the statue” etc.

Perhaps wikipedia entries linked to group of sculptures - Wikidata might shed some more light on it, but there are only available in 4 languages (none of them being English), and while all have different information which would be useful, they all seem to translate quite badly using both Bing, Google and DeepL. (e.g. they often don’t distinguish between statue / sculptural group and translate them the same, even if original language it is clear that different words are used, and such nuances are quite critical for this task)

[dcapillae (Daniel Capilla)]

@Matija_Nalis

Conflictingly, wikipedia for Laocoön and His Sons - Wikipedia seems to call it “The statue of Laocoön and His Sons”, and talks about “The figures in the statue” etc.

Another similar case is Three Soldiers. Perhaps we are facing a cultural difference and the problem is that in some countries they use indistinctly “statue” for what in others is known as “sculptural group”.

[Kovoschiz]

Nesting them is fine. I meant the problem is whether to put =artwork + this proposal, or eg site=sculpture_group, on the =site; or use =cluster (not needing any feature tags). The members are already using =artwork as an individual feature.
There may even be cases where a member is a “sculpture group” proposed here. Then there are 2 levels of grouping that needs to be solved.

Putting this aside, statue:people= =multiple / =group / =set is another possibility.

[dieterdreist]

Nesting them is fine. I meant the problem is whether to put =artwork + this proposal, or eg site=sculpture_group, on the =site; or use =cluster (not needing any feature tags). The members are already using =artwork as an individual feature.

I don’t think a “site” of several artworks is very appropriate to represent a sculptural group, because the latter may often be considered a single piece of artwork, the sculpture is the collectivity of the individual parts.

[Kovoschiz]

That’s why I asked that question above, especially list point 3 and 4. The sole actual example of site=scultpure_group used in Relation: ‪NYPL library lions‬ (‪11520253‬) | OpenStreetMap are already a pair of =artwork on each side. This shows the meaning of “group” has both a closely and loosely related interpretation.

[dcapillae (Daniel Capilla)]

Another example of a sculptural group, in this case depicting a group of seagulls in flight. Note that they are a single sculptural group. With the current tagging scheme, this sculpture should be tagged as tourism=artwork + artwork_type=statue.

Example 6. Seagulls in Flight by Jaime Pimentel (1962), an example of sculptural group. Source: own work (CC BY-SA 4.0) available on Wikimedia Commons.