Argh! Stupas as Wireless masts...

I just came across some ruined Stupas in Wiang Kum Kam (South Chiang Mai) that AlaskaDave marked as “Tower” (in potlatched nicely marked as wireless antenna). I had marked the ruined wats already with the “tourism” “ruin” symbol, planned to add the names. While “stupa” is an accepted tag for “tower” according to the wiki, I find it rather disconcerting to see an antenna symbol on these stupas. can’t we just leave it with tourism markers (ruin, etc.) and avoid that “tower” tag?

Well, we could do that but they are indeed towers, are they not? And some are quite high. I was the one who added stupa to the tower:types page expressly for the purpose of tagging those tall wat towers.

That said, the problem is with the rendering, not with the classification. A bigger problem, at least with the Lambertus renderings, is that wats show up only as small dots. I would like to have both of these changed but, so far, it’s too big a job to add to my other OSM chores at the moment LOL. It’s weird too, because there is a perfectly adequate icon for “temple” in Garmin’s icon collection that does not get used.



As Dave said, it doesn’t really matter how Potlatch/JOSM/iD/Mapnik shows it - what matters is that the data is correct.

The renders will always be behind in a project like OpenStreetMap where everyone can invent their own tags for special purpose maps and new features. Though lots of the tags we add today doesn’t get rendered, it may very well be in the future (Think highway tags like lanes, width, lit, surface, etc).

For a long time a Buddhist temple was rendered as a grey square box and though it was not the correct icon, we still tagged them because the data was correct.


Thank you for both comments, and nice to see that the AlaskaDave is here. I sure appreciate all the efforts that you are putting in, few people seem to work on Thailand.
Indeed, the original data should be correct, and not necessarily the rendering. Still, I can only think “arggh” when I see the Antenna… ;).
While I’m a newcomer here, I have been working with classification and standardization issues in my work, so I’m familiar with the issues and the difficulty of bringing everybody’s wishes under one hat.
So, I would like to present some arguments, why I think the “tower” designation is not that great for stupas, or chedis as many should be labelled in Thailand.

Stupas are originally burial mounds - from wiki “mound-like or semi-hemispherical structure containing Buddhist relics, typically the ashes of Buddhist monks”. In fact, many of these are more similar in shape to burial mounds/pyramids in china or round pyramids in Latin America. Given their variety, many would not be considered “towers”. Thus, this would create an inconsistency in labeling stupas/chedis i.e. some are towers, some not.
Ideally, a separate tag for stupa, instead of tower, would be nice.
In it’s absence, perhaps a generic “monument” with subset “stupa” would work. monument could then also have subsets like, pyramid, etc. etc.
So, is there currently an alternative for labeling stupas (not as towers)?

BTW, the “ruin” tag I used is also not 100% satisfactory, since effectively it’s just a condition of any building. Essentially, any building/monument could have the condition tag “ruined” attached, i.e. in Wiang Kum Kam, they would be “Wats, ruined”. But I guess this would imply too big a revamp of tags.

Thanks for your thoughts,
Happy Holidays.

That is indeed some pretty good arguments, and if tower:type=stupa is not generally used and isn’t supported by any renderes/tools then we could probably change it to something else - if people feel like it.

It also looks like the tag was introduced in the Wiki by AlaskaDave himself in February, which means “Thailand” invented the tag and could possibly re-invent it if required.

So maybe it is time for some research, what other options do we have? What are used in other countries, I am sure this is not unique to Thailand. I prefer to find standards that are either already in use if possible.

I might take a look at India, I think they may have similar structures.


My mistake, the tag wasn’t introduced by AlaskaDave, I think he just added the nice picture.

The tag tower:type=stupa is used 31 times in total.

monument=stupa is used 91 times.

Actually, AlaskaDave did introduce the subtag stupa. (as well as the photo) I was looking for a way to tag, what to me at least, were obviously towers that dot the Thailand countryside. The actual term “stupa” was suggested by someone on the OSM Help site. I would have preferred “chedi” but that term had never been used. In any case, now that it seems to be an issue, let’s see if we can come up with something better.

That said, I would argue that these religious structures are indeed towers in the truest sense of the word. Wiktionary defines a tower as:

. Straightforward and simple.

Wikipedia expands on this quite a bit:

Certainly, those tall structures attached or near many Thailand wats are IMO deserving of the term “tower” and they are certainly man_made. Furthermore, the term has been used to describe structures, both ancient and modern, that were built for religious reasons.

However, if the consensus of this group is to use the more popular key=tag combo of monument=stupa (94 uses), I do not have a problem with that.


I am not necessarily against it, but if it was used thousands of times in many countries I wouldn’t even think it was worth changing it.

I am just helping out with some research and providing some statistics. I’ll let you guys figure out what is best to use in the future, I can see them both as towers and monuments, neither seem completely wrong to me.

So far I have only been able to come up with monument=stupa and tower:type=stupa, and both have very little use.

RocketMan who haven’t said anything yet seems to be using monument=stupa though, so he may have an opinion here as well.

You could try to get more confused by asking on the “tagging” mailing list.

tower:type=stupa looks good given the fact that it’s also used to tag the bell tower on christian churches. As this is the second most common use of this kind of tag (after communication) it looks quite established for me to use it with religious places.

I vote for keeping it as tower:type and opening a ticket for the renderer to evaluate the tower:type value and if not known use a different default symbol.

BTW: Buddhist temples are rendered with the Wheel symbol both on the main site and on There is a thread about adjusting the rendering for Thailand.

Actually, as Alaska Dave also pointed out with the quotations from Wiki, which I left out in my initial post, but did read, a tower is “A tower is a tall structure, usually taller than it is wide, often by a significant margin”.
I also note the openstreetmap definition: man_made tower: A tower is a building, which is multiple times higher than its diameter

And that is the reason why I think it is not appropriate for stupa. There are many forms of stupa, and many do not classify as tower.
Just to give some ideas:
Borobudur is a terraced flat stupa with many small stupas on it
Here we have a small stupa:
Many stupas are small, i.e. for example in Wat Mahathat has many small stupas (Chedis) on its ground, in fact also modern Wats have many little commemorative Chedis on their grounds (platforms in the foreground
There are basically three stupa types: dome, terrace (Borobudur), and tower-like
The Chedis are basically dome shaped derivatives, with terraces as base, while the old Mon style quadratic tiers of the chedi of Wat Chedi Liang (Wiang Kum Kam) are terrace based.

The tower-like stupas are actually the Pagodas of the Far East. (note: pagoda should also be considered in this discussion, it seems not defined so far, and used in OpenSeaMap in a somewhat inappropriate definition (temple), also a few other tags refer to Pagoda )

My point is simply to have a general way of tagging all stupas (whether tower or not) in the same way. A further question would be whether to include Pagodas in this, or treat them separately. While it’s easy to discern the multi-tier pagodas in China/Japan, the Shwedagon Pagoda in Yangon looks like the normal spired Stupa/Chedi, and is officially also called Shwedagon Zedi Daw (note “Zedi”) - I guess the name Pagoda is a misnomer from colonial times.

Can multiple tags stand?
I admit, I am not familiar enough how appropriate it is it double tag the same thing several times. i.e. is it ok to tag some stupas with the tag tower, as well as a second time with another tag (e.g., monument:stupa), which refers to all stupas and chedis? It seems like double work and would lead to conflicts in rendering (which one is valid, or both are shown next to each other, information overload)

I do notice that conventions for towers are rather confused, i.e. multiple ways to define the same thing, e.g.
man_made campanile (bell tower italy)
man_made tower - tower:type=bell_tower.
as mentioned, getting consensus in a world-wide project seems hopeless… :wink:

Perhaps at this point, stupas don’t need to marked at all, since they are usually part of a place of worship.
Perhaps one should discuss how to have a consistent nomenclature to label all historic sites, including Ayutthaya, Sukothai, Kamphaeng Phet, Si Satchanalai, all the Khmer sites (Pimai, Phanom Rung, Preah Vihear), etc.,. If then subfeatures within a site are to be labelled, can then be taken on afterwards.

PS: I just noticed, there is some discussion about renaming ruins.

I noticed that quite a number of stupas/chedis have been just marked with “generic building”. (in ayutthaya for example).

PS: the comparison to “bell tower” is flawed in my opinion, since the structures are of very different initial design and intention. Church towers with bells, or free standing bell towers are really designed as towers originally. Even Wats can have bell towers:

@Endless - If your main argument against stupas as towers has to do with the fact that their rendering is wrong, then I suggest we work on changing that. I still maintain my opinion that a stupa is a form of tower. Furthermore, I tag them not because they have religious importance but because they are visible from a distance and hence make good positional landmarks.

We could, I suppose, create separate categories for all the variations on the tower-chedi-monument scheme but why not use something that exists already? As for using the tag monument=stupa, I don’t favor that approach either. A monument is, to me, something created to commemorate an event, a person, or a place that has historical significance and thus IMO constitutes a special case. Some towers happen to be monuments but only occasionally. The stupas or chedis we’re familiar with were not created as monuments but as places of worship. Consequently, because in general all man_made towers are structures “usually taller than wide”, I cast my vote to stay with it. The various tower:type sub tags allow us to expand on that theme as we see fit and as the need arises for more specificity.

Hello AlaskaDave,
No, it’s not the rendering per se. Perhaps my comments where too long-winded.
Argument: Many Stupas are clearly not tower-shaped, (and many won’t be great as landmark, but that’s not the point). I don’t recall that professional art historical descriptions of stupas refer to them as towers
My point: use a designation that can be applied to all stupas to make it consistent.
at present a number of them are also just marked as buildings (Chedi Luang)

As to monument, no I am not tied to that at all. It just seemed to be a possible solution in absence of something else.
As to the definition of monument, see, - perhaps a bit broader than your definition. A Stupa fits this description very well, because the very first ones were burial places of Buddha’s ashes (derived from burial mounds), and now they are places for important relics (tooth of Buddha etc.,), or other important relics, or ashes of famous monks, or ashes of people. That’s why they are actually places of worship, like any tomb/site of a famous person can be a pilgrimage site/place of worship.
So, a stupa is very much a monument. - but just to clarify, I’m not tied to it.

Thinking more about how to label ancient sites in Thailand, it occurred to me that there is a legitimate use for “tower” in Khmer sites. Art historical descriptions of sanctuaries clearly refer to the central sanctuaries as towers, e.g. , the central five towers, or, central five towers.
So, no, I am not in principle against towers, just where appropriate. :slight_smile:

Well, seeing as we seem to be going around and around on this topic, what do you suggest?

Sorry for the delay in replying.

First, for sake of expedience, I would suggest a single Point marker is used, as you did (i.e. the tower).
I have seen a number of chedis/stupas marked out in outlines as buildings, an elaborate example is the Chedi Luang in Chiang Mai, but marking all stupas as outlines seems an excessive work at present.

So, then the question is, which tag is most suitable as marker for all stupas.

Monument ? Memorial ? Place of Worship,
or introduce a new tag : Stupa/Chedi/Pagoda for mainly the Asian regions that have them? Perhaps this could be broadly discussed?

@Endless: I don’t believe mappers should be restricted to marking stupas with either a single node or a closed way. That’s entirely up to the mapper and the particulars of the stupa being mapped. If the structure is large, a closed way can be more descriptive, even more accurate. The important thing is to get it on the map. Consequently, we need not, and should not, make such a rule. A similar situation applies to amenity=fuel. Most are mapped as nodes, some as ways. Most mappers don’t take the time to outline all the buildings in a fuel station and frankly it’s not necessary to do so.

When you ask “which tag is most suitable as a marker for all stupas?” and follow that with “Monument? Memorial? Place_of_worship?” you demonstrate in your own question why it is not possible to use the same tag in all situations. There simply is no “right” answer, no one tag will cover all. And if you delve into the arguable distinctions between tower, chedi, stupa and pagoda, you’re going to complicate things to the point that most mappers will settle for something simple and to the point anyway.

I’m following a similar discussion in the tagging group about how to classify road surfaces and the quality of those surfaces. Phew, trying to come up with a scheme that isn’t overly subjective yet accurately portrays the situation in terms of usability for all types of wheeled conveyances from roller blades and wheelchairs to 4x4 vehicles, is a huge chore. One fellow was suggesting measuring the frequency and average depth of holes in the road surface as a measure of, here they introduce a term used in Australia, trafficability! My feeling is that when mappers see something that complicated, they will opt to simply say paved or unpaved and let it go at that.

I want to encourage all involved to not make this scheme overly complex. If most of us want to add more subtags to the tower:type set, I’m fine with that. If we want to add completely new tags like stupa=yes/no, or pagoda=yes/no, man_made=pagoda, or something similar, perhaps we should introduce this topic to the tagging group or at least to a wider audience.

to alaskadave.
No, it was not the intention to restrict stupas to a single node. to me at present, given the huge white parts on the map of thailand, speedy mapping with single nodes is just preferred of course.

And no, I want to make things simple. just one tag. stupa/chedi was not meant as distinction, it was meant as being the same, after all, its the same. pagoda is somewhat different, it does have a well defined description. Confusing is only the probably from colonial times confusing use of pagoda also for stupas, such as the shedagon pagoda in yangoon,

I totally agree that one should not make it complicated like your road example, my idea was exactly to make it simple.
As to the different shapes of stupas, I think this does not matter, a stupa is a stupa, no matter how it looks. Like a museum, no matter whether it looks like the guggenheim, or the hermitage. Or a church, no matter if it looks like the munster in Ulm, or is little more than a little boat house (

the list of
“Monument? Memorial? Place_of_worship?” was meat, as choice. Could one agree on one of these as a single solution?
a) Monument
b) Memorial
c) Place of _worship

d) A new stupa tag

e) other suggestions

perhaps this should be taken out into a new thread to all read it…, or make it also broader (other countries), as suggested


I repeat again, no it is not possible to agree upon choosing one of those as a “single solution” because there is no single solution.