I would like to submit a request for a revision in the Thailand wiki concerning the classification of villages versus hamlets. Since the latest Thailand wiki update, it is suggested that all officially recognized villages be labeled as place=village, while smaller non-official settlements are typically categorized as place=hamlet.
The challenge arises from the absence of official information, unlike higher-level classifications. Village names often lack official sources, making it difficult to ascertain this information. Consequently, many mappers opt to downgrade a village to a hamlet based on the global wiki definition, which relies on their smaller size (Tag:place=hamlet - OpenStreetMap Wiki).
However, the criteria are ambiguous and can vary by country. Therefore, I propose the establishment of a Thailand-specific rule for place=hamlet, regardless of official classification, given the limited availability of such information in 99% of cases.
Since we also lack population information, and settlements can exhibit diverse demographics, ranging from elderly individuals staying alone to thriving communities with large families (2-4 children per household), I suggest considering the definition of a maximum number of houses (e.g., 10 or 20 or ?).
How would you explain the method to someone not familiar with the situation. Especially when the data used to distinguish the two settlement types is either not reliable or missing.
As always, hard to define a rule when the “status” and even boundaries are vague to the Western mapper who relies on the only info available - often a road sign, and an aerial view.
The reality is we already have a map with a mixture of tagging methods, from landuse relations covering several villages/hamlets, to nodes which sit at the heart of a city. Its not consistent.
Julian created this post in response to me correcting a node originally tagged as hamlet, that was then revised by a mapper to village with the comment stating it needed “better visibility” (presumably by the render). The node covered about a dozen houses, AND was sitting in the one of the areas included in a relation tagged as a HAMLET, hence my correction. It rattled cages and I don’t take criticism well !
However I accept the sentiment that some more clarity is required.
As I am unable to track down the official status of any settlement, I use the aerials to guide me on whether its hamlet or village. I don’t have hard and fast rule, but generally the no of houses, and the size of area they exist in, does work as a rough guide. Remember mapping is an art not a science, although many data purists would disagree.
I took issue with the fact that “it was decided” that the tag hamlet was not appropriate in Thailand, to which I disagree. The Wiki is probably OK as it is, being that it does “allow” the use of the hamlet tag if you don’t know the status of a settlement, and its small. I’m fine with adding a guide to the Wiki based on number of houses, but how many ? My thoughts are, say around 25-50 houses.
And if a smarter person than me decides it has an administrative level, and should be village, then go ahead and revise it. Just adding a note or admin level tag, tells me not to mess with it as its someones else’s considered & perhaps researched work. … but “coz I wanna see it on the big map” just doesn’t work for me !
Thanks for your input! I checked out Hamlet (place) - Wikipedia, and based on that, here’s my updated suggestion to throw into the wiki:
If the official status of a village is unknown (no admin_level tag present), and the village lacks essential infrastructure (schools, churches/temples), and exhibits fewer than 50 visible rooftops, the village is eligible for tagging as place=hamlet.
Feel free to share any objections or suggestions for improvement!
The source of all officially recognized villages has been mentioned before:
There should be no ambiguity what is to be mapped a village or a hamlet.
All villages in the dla-database are place=village with admin_level 10. Other settlements not in the dla-database are place=hamlet without admin_level.
Only if you think every government village record can be a perfect match with an OSM place.
The sites mentioned only let you search by district/SAO, which most mappers don’t know much about. And even if they do, I’ve found that village names and locations in OSM can be way off, so you might still end up with the wrong info.
GPS coordinates are only available for the main SAO and subdistrict municipality (องค์การบริหารส่วนตำบล), so there’s no surefire way to check individual villages against the official list.
Sure, it might work in most cases, but it’s a real hassle and not something regular mappers would want to bother with. That’s why we need more details in the Thailand wiki.
The best fix would be adding the admin_level to the existing village data, even I thought about doing it before and lost interest because there are no GPS coordinates at the village level.
Why should this not be the case? For every village in the database there is a village on the ground. I have mapped many villages and the official source was always correct.
Most villages will not need to be looked up in an official list. Only for doubtful cases like the one mentioned by Russ Mc D should someone clarify the tagging. I did a quick search for this village on the noplink website. It is not listed as a village in the Chiang Dao district and is therefore a hamlet. If you want to add admin_level there is no way around it. Only an official list will give you that information.
Don’t assume all village names in OSM are spot-on—I’ve surveyed many in Northern Thailand, and many were off.
The official source only gives names, no GPS coordinates. So, there’s no bulletproof way to confirm if an OSM place matches a government record just by the name.
Take the village we talked about with Russ—it might possibly have the wrong name. The real one could be in the official Chiang Dao district, meaning a different admin_level and place tag.
Bottom line: Don’t take the OSM village names for granted. We might need extra sources, like signs, to be sure.
This may be true for non-local mappers. I have no problem to identify a village with 100% certainty. For me a rooftop limit is just a temporary solution for people who can’t be bothered (your words) to find official information. In the end someone will have to look at official data and determine whether it is a village or not. But if the consensus is that this proposal:
“If the official status of a village is unknown (no admin_level tag present), and the village lacks essential infrastructure (schools, temples etc.), and exhibits fewer than 50 visible rooftops, the village is eligible for tagging as place=hamlet.”
is a good idea, no objections here. It will not concern my mapping.
