From OpenStreetMap Wiki
Jump to navigation Jump to search


Nice additional tag, thanks. As a newbie to OSM I expected such a tag and was surprised to find that it had only just been added.

For the UK it would be good if mappers could add the appropriate link to the UK National Inventory of War Memorials page for the memorial. Cheers -- John (Daytona2·Talk·Contribs) 09:27, 29 September 2008 (UTC)


I would suggest a tag inscription=* for reproducing any inscriptions on the memorial. Any comments? --abunai 17:32, 17 May 2009 (UTC)

I like the idea. Perhaps even inscription:XX=* for different languages if that's the case. --Nighto 23:23, 22 February 2010 (UTC)
Tagwatch shows inscription=* is used, however memorial:text=* is more widely used (1979 uses) --GabeH 18:03, 12 December 2010 (UTC)

By the way, do you think that this is the right tag for sculptures like busts? --Nighto 23:23, 22 February 2010 (UTC)

If the bust is set up as a memorial over a person, why not? If the purpose of the bust is more decorative, than find something else. Isn't there any tags for sculptures in general? That could be needed for parks like Vigeland Parken in Oslo, Norway, which displays several sculptures by Gustav Vigeland. The entire park can be seen as a memorial over the artist, not each single sculpture. --Skippern 01:47, 23 February 2010 (UTC)

Very long inscriptions may be added to WikiSource, and tagged here with, say, inscription_url=*. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 11:33, 20 November 2011 (UTC)

I'm using the URL tag for photos of the inscription too, I hope that's fine. Of course I store the photo on Commons. See --Nemo (talk) 23:16, 10 July 2016 (UTC)
Seeing this now after many years. For the record, the established tags are inscription=* and inscription:url=*. --Dieterdreist (talk) 13:28, 24 February 2017 (UTC)

Commemorative trees

Please see discussion at Talk:Tag:natural=tree#Commemorative trees. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 11:29, 20 November 2011 (UTC)


Some mappers at Germany are using memorial:name=* within the Stolpersteine-Project. name=* was not fitting, because a Stolperstein is only dedicated to a victim. This might also be applicable to some boards without names itself, but which are dedicated to a Person. I think this key-name is not a good choice, because it seems to be a duplicate to the name=*-Tag (another proposals for the same thing is person:name, which is not satisfying, too).

It's worth looking at wikipedia=* and particularly the proposed wikipedia:subject=* for memorials to notable people (or events; there are also memorials to named animals) who are the subject of Wikipedia articles. Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 12:08, 24 September 2013 (UTC)
So the Wikipedia-Proposal implies a tagging like memorial:subject=*. As an alternative: I thought about memorial:designation=*, which might be not very specified. So +1 to memorial:subject=*.--Cracklinrain (talk) 14:09, 2 November 2013 (UTC)
See also subject:wikidata=*.
--Pyrog (talk) 12:26, 18 November 2019 (UTC)
This is still an unresolved issue that has widespread effects. It is very common to misuse the name=* of memorials to indicate the subject of the memorial instead of its actual name. For example, it is very common to set the name of historic persons such as politicians or artists that are the subject of a plague or statue - especially if there is no clear true name of the memorial itself. memorial:subject=* has almost 700 uses in mid 2020 while memorial:name=* has twice as many (certainly many from Stolpersteine but it is used for other memorials as well). How shall we go forward: mention both on the wiki page (how? deprecating one?), start a vote, ...? --Stefanct (talk) 12:44, 11 August 2020 (UTC)
--Pyrog (talk) 13:54, 11 August 2020 (UTC)
I am not a big fan of relying on external databases for key functionalities. Also, IDs are not human-readable as is. The other (albeit related) disadvantage of solely using wikidata IDs is that there is little wiggle room if needed because of unclear/minor cases where no clear ID exists. You mention collections of subjects as for war memorials. In my country there are lots of war memorials dedicated to those fallen in WWI or WWII of individual small villages... adding and referring to them by a wikidata ID (that possibly does not even pass Wikidata's notability requirements) does not make a lot of sense. I acknowledge the big advantages of tagging the ID though (if available or useful otherwise)! Why not allow/promote both?
--Stefanct (talk) 14:33, 11 August 2020 (UTC)
> Also, IDs are not human-readable as is.
Well, with iD or JOSM, you get automatically the label (even in your language)
> where no clear ID exists
Just create it :)
> Why not allow/promote both ?
Duplicate data mean more work ;)
--Pyrog (talk) 16:43, 11 August 2020 (UTC)
While it is perfectly fine to link to wikidata (additionally), I also agree with Stefanct that we should have and promote a proper OpenStreetMap tag for everything that we consider relevant (includes subjects of memorials and artwork). If the issue is duplication then I would prefer refraining from third party references. —-Dieterdreist (talk) 17:39, 11 August 2020 (UTC)


memorial=* has been massively changed to memorial:type=* by user malenki in Germany, thus taginfo usage statistics are 99% influenced by a single contributor. This variant is not the result of a general consensus, neither the usual tagging rules used in OSM (for instance, "building=school" is not "building:type=school", etc) and is therefore not recommended in the main documentation. --Pieren (talk) 16:17, 19 February 2014 (UTC)

I agree. memorial=* seems more logical to me, and keys with ":type" or "_type" are prone to misspellings because it's difficult to memorize when to use colon (e.g. tower:type, maxspeed:type, seamark:type, mtb:type) and when to use underscore (e.g. artwork_type, map_type, reservoir_type, border_type). --Fkv (talk) 19:10, 12 August 2014 (UTC)
As for "malenki removed memorial=stolperstein": Thats not the case. See for example , where malenki removed a wikipedia tag (and rightly so, because the wikipedia article is for the general project "Stolpersteine", and not for this individual stolperstein. Or see the one round the corner where he did exactly the same thing: . I think because memorial:type=* is widely accepted in germany at least for the stolperstein project, this use should be documented on the wiki accordingly. --Gormo (talk) 06:47, 1 October 2015 (UTC)
@Pieren: you wrote:
memorial=* has been massively changed to memorial:type=* by user malenki in Germany, thus taginfo usage statistics are 99% influenced by a single contributor. This variant is not the result of a general consensus,
Please prove this wild speculation -- malenki 09:22, 1 October 2015 (UTC)
@Malenki - surely it is for you to show that your changes have consensus? Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 23:01, 15 November 2015 (UTC)
@Andy: Consensus on what? -- malenki 12:27, 16 November 2015 (UTC)
Please provide a link to a stolperstein node (or OSM IDs) where malenki has performed the action you critizise. You can - for example - use the deep history tool at to check which particular user changed what tags in what changeset for a specific object. I think you are misinterpreting OSM object histories. --Gormo (talk) 14:36, 16 November 2015 (UTC)
Obviously neither Pieren nor Andy "Pigsonthewing" Mabbett are able to or want to substantiate their allegations/assumptions with facts. It seems the former has quite contributing to OSM altogether. -- malenki 11:12, 15 February 2016 (UTC)
I think they are misinterpreting the object histories. I think this is an unfortunate misunderstanding on "their" part, it's not malice but incompetence/errors in understanding. I think this should not escalate. --Gormo (talk) 12:19, 15 February 2016 (UTC)
I have looked at the tag history and the key memorial was at no time higher than memorial:type, rather the latter was used from the beginning and "memorial" only gained recently. Still there are significantly more memorial:type than memorial tags in the database as of now. --Dieterdreist (talk) 10:06, 11 January 2017 (UTC)
Indeed, but most uses of memorial:type are with stolperstein, in Germany (18 395 out of 21 911 uses). So I guess that the Stolpersteine project has started using it and didn't change it. The RedBurn (talk) 08:36, 27 January 2017 (UTC)


I would like to add artist_name to the page for tagging the artist like on tourism=artwork, can that be done without rolling out a proposal? :) Hakuch (talk) 21:04, 27 January 2016 (UTC)

I see no reason why you could not. --Escada (talk) 08:16, 17 February 2016 (UTC)

Real persons and events only, right?

Am I right, that historic=memorial, being a tag within historic=* key, should always reflect objects, commemorating real people or events, including abstract ones (like, "Memorial of Unknown Soldier", which commemorates all real fallen soldiers, whose remains haven't been identified)? If yes, shouldn't it be explicitly told in tag description? There are sculptures of fictional characters or artworks, inspired by fictional events, which took place in non-documentary films or literature, and I think, these should always be mapped as artwork, not as memorials. --BushmanK (talk) 18:27, 4 June 2016 (UTC)

I'd say yes. A "memorial" to Gandalf the Grey (from Lord of the Rings) or Lord Vetinari (from Discworld) would be an artwork, not a memorial. The historic-namespace should be reserved for historic events that, according to history, really took place. There might be grey areas (a memorial to something Jesus Christ did somewhere, according to the Bible), which I would see less strict (because there are historic interpretations of the Bible and other religions' texts). --Gormo (talk) 06:41, 6 June 2016 (UTC)
Bible and its interpretations are not a part of evidence-based science, so I don't see why it's "grey area". However, memorial from your example isn't an artwork for tourists either, so it's another question, how to tag it properly. --BushmanK (talk) 14:51, 6 June 2016 (UTC)
I have 2 memorials that are to animals. One to Mathew Flinders cat Trim the other to Australian horses in WW1, there are more of these WW1 horse memorials within Australia. So these real events, fictional events it is a possibility. Oh "dog on the tucker box" ... a 'legend' humm 'dedicated to the pioneers' Humm I am certain there will be a monument to some myth somewhere. Warin61 (talk) 20:50, 19 January 2017 (UTC)
What about memorials to events considered as real by some? Religions, myths, propaganda, confusion, better historical research can cause all of that, sometimes it is clear whether something has ever happened, sometimes it is object of ongoing politics. Also, at least animals and plants may also have memorials. Mateusz Konieczny (talk) 09:59, 12 June 2022 (UTC)
Also, in many cases things are confused - lets take,_October_2020_(1).jpg depicting a god that is considered as not existing by vast majority of people. But it is copy of sculpture created by people who believed in it and still has some worship, maybe even serious one (note candles at bottom of sculpture). But it is also can be considered as reminder of a deeper past and roots of a local culture. Is it memorial or not? Mateusz Konieczny (talk) 10:03, 12 June 2022 (UTC)
Yes, it is. maro21 12:51, 26 June 2022 (UTC)

memorial= values are orthogonal

Values of memorial= are mainly representing a physical form of a memorial, such as a statue, a plaque, and others. However, memorial=war_memorial indicates an event, which a memorial is dedicated to. Obviously, it creates the false dichotomy, when a mapper has to make a choice, what to indicate: a physical form or a dedication. This is totally wrong, and I suppose, that was a ground for introducing memorial:type= key. Currently, there is no other way to describe a war memorial and its form, except to use both memorial:type= and memorial=. --BushmanK (talk) 23:21, 7 January 2017 (UTC)

You should take a look at the German page. There is a statistics table, which at least at Stolerstein has an overweight for memorial:type=stolperstein. This can not be so easily changed with the rules of OSM. An have a look at Historical_Objects/Map_Properties. I would be happy if the things are arranged more clearly. --geozeisig (talk) 07:24, 8 January 2017 (UTC)
First, I just wanted to mention this obvious issue of this scheme. Second, I don't think we should worry about being able or not being able to change anything by "rules". Changes are made by changing (clarifying) documentation and editor presets. Not immediately, but there is nothing happening immediately in OSM. --BushmanK (talk) 18:02, 8 January 2017 (UTC)
+1 with BushmanK on both the scheme issue and the migration as a way of change. memorial:type=* is not very descriptive. Tagging could be something like memorial:event=* (war, WW2, WW1, death, rescue etc) and memorial:form= * (plaque, bench, statue, monolith etc). This is more descriptive than the over used and misunderstood 'type'. Warin61 (talk) 22:24, 18 January 2017 (UTC)
There is now subject:wikidata=. --Shoepuppet (talk) 03:14, 16 July 2018 (UTC)
>Currently, there is no other way to describe a war memorial and its form, except to use both memorial:type=* and memorial=*.
You could use memorial=plaque;war_memorial

For information, in France a war memorial is a type of memorial usally called "Monument aux morts" and build after WWI.
They ususally look like this:
--Pyrog (talk) 12:20, 18 November 2019 (UTC)

Name vs. subject

Most memorials don't actually have names. But they generally commemorate a subject such as a person, an event, a place, an animal or a thing. The subject generally does (or did) have a name, or at least a description. So the name=* tag doesn't really make sense. I'm using subject=* instead. Objections? T99 (talk) 07:03, 24 September 2019 (UTC)

See subject:wikidata=* above.
--Pyrog (talk) 12:22, 18 November 2019 (UTC)
Some memorials are statues or sculptures of something other than the person (or people, etc.) that the memorial is for (e.g. node 9235086641).
I think in that case it makes sense to use subject=*/subject:wikidata=* for the subject of the artwork and
memorial:subject=*/memorial:subject:wikidata=* (which has 1'412 uses on taginfo as of 2021-11) for the subject of the memorial.
In cases where the two are the same (and the memorial actually depicts the subject) memorial:subject=* could be omitted.
—— AlephNull (talk) 11:08, 8 November 2021 (UTC)
subject=*, inscription=*, subject:wikidata=*, subject:wikipedia=* all can be useful Mateusz Konieczny (talk) 09:53, 12 June 2022 (UTC)


A Khachkar is memorial. There are hundreds of khachkars worldwide, many of which are memorials to commemorate the victims of the Armenian genocide. Any opposition in adding this as an additional type of memorial? Uip (talk) 18:31, 27 October 2021 (UTC)

I would see these as a specific case of stele, my suggestion would be historic=memorial and memorial=stele combined with stele=khachkarDieterdreist (talk) 08:05, 28 October 2021 (UTC)
A Khachkar is usually a stele, so memorial=stele. So I don't think memorial=khachkar is a good choice. Wikipedia: "A khachkar is a carved, memorial stele". maro21 19:38, 28 October 2021 (UTC)
Resolved: there is now note in article mentioning Khachkars Mateusz Konieczny (talk) 09:53, 12 June 2022 (UTC)

historic=memorial on relations

The page currently suggests that historic=memorial should not be used on relations. But there are cases when a memorial consists of multiple objects - like a railway car, a memorial stone, a stele and perhaps a few other objects. Would it make sense to map these as relations, or is the best still a simple area that has no physical object mapped on it, and then individual objects get no historic=memorial? --Richlv (talk) 22:13, 30 January 2023 (UTC)

pavement_plaque good or not? Brightj (talk) 15:24, 6 February 2023 (UTC)

Calling monuments memorials

I think asking people to call monuments memorials, and memorials monuments, is "editing for the renderer" pure and simple.

The Jefferson memorial is a memorial. Boundary monument 34 is a monument.

If you want to know their importance, then have people give that as a additional parameter. Same goes with size.

Otherwise it's all very mind-bending. Jidanni (talk) 17:00, 21 February 2023 (UTC)

Is there a difference between a monument and a memorial? Brightj (talk) 19:48, 21 February 2023 (UTC)

See also Talk:Tag:historic=monument#Calling_monuments_memorials.

Also don't blame users in future purely visual editors when an icon used for something is a perfect match, and they tap it. (No, they can't always cross a river to see what some nameplate on it says, if anything.)Jidanni (talk) 20:41, 22 February 2023 (UTC)