Administrative level of townships in Maine

Communities United States
41

I am late to the party but wanted to weigh in, perhaps poke some holes in logic. Correct me if I’m wrong, but not all suggested edits have been made? Perhaps consensus has not been reached. I found it odd that Plantations are still admin_level=9 when their existence as an administrative entity is stronger than that of a township that has joined the unorganized territories. Plantations should definitely be admin_level=8

From what I understand an unorganized township (Forest City Township) has the same administrative responsibilities (none) as any of the survey blocks (T10 R12 WELS), the only difference being that once upon a time, the unorganized township had some administrative responsibilities, and now maintains an identity on maps with a name instead of a survey code. When a town disorganizes, they sell off government infrastructure. While they could go through the process to organize again, they will no longer have access to the administrative buildings they once owned, and will have to create entirely new facilities or buy back old ones. For these reasons, I view Townships as equal to the survey blocks. I don’t consider them to have any authority on their own. The advantage they have is if they are ever successful in organizing, their borders are predefined.

In Vermont, it looks like our gore’s and unorganized areas are admin_level=8 (Example: Warner’s Grant). However, in NH they are admin_level=9 (Example: Chandler’s Purchase). Aligning the three states as 8 or 9 should be a goal. From my previous research on this, VT/NH/ME have these strange designations, where the land maintains some level of identity thru naming. This differs from out west where they are looser and land is either within ‘city limits’ or part of the county. CT/MA/RI are more civilized and have (I think) allocated 100% land to incorporated towns/cities.

It was mentioned earlier, but when a series of Maine’s unorganized townships are adjacent to each other, the Census combines them and names them. This can be confusing for pretty much everyone. Since other census boundaries are in OSM (for better or for worse), I’d suggest additional coding be added to the Maine ‘blocks’ to stave off confusion, in which we can clearly designate those boundaries as census boundaries and not an admin_level.

As an aside, Prentiss Plantation recently disorganized, and is adjacent to two other unorganized townships. It’s predicted that in 2030 the Census will create a new ‘block’. Very exciting.

One pitfall to be careful of during this process: the inclusion of township/plantation in existing documentation may not represent the legal administrative structure of the entity. Magalloway Plantation disorganized and joined the unorganized territories in 2018. Codyville Plantation disorganized in 2019. However, in OSM these two entities showed up different. Magalloway is still listed as a Plantation but Codyville was updated to a Township. More recently, Drew Plantation disorganized, but is still listed as a Plantation in OSM. Perhaps these types of mistakes will be caught overtime and it’s not worth slowing down the project to get it 100% right. I think there may be a few other mislabeled Plantation/Townships situations.

Other fringe cases I noticed:

  • Islands in Moosehead Lake are not captured by any admin_level below the county level
  • similarly, there are a few islands off the coast that don’t seem to be captured
  • Reservations do not have admin_level
1 Like
42

In that case, I wonder if boundary=historic might be more technically correct. The image in the infobox even shows a boundary marker for a historic township in Nova Scotia:

Unfortunately, there’s been some confusion about what this tag is for, so any renderer that adds support for it would end up with, for instance, 1911 electoral districts in Ireland/Northern Ireland that probably no one uses as a geographic point of reference. (Fortunately, that import added end_date=* tags, making it pretty obvious to data consumers that this data that belongs in OpenHistoricalMap instead of OSM.)

43

Correct, only a portion of the suggested edits have been made. I was pausing to wait for any other feedback that might roll in. I’ve also been doing some background work to prepare for creating wikidata items to 400+ townships in Maine that don’t have established wikipedia pages along with some boundary validation scripts that I hope to share soon, with the goal of getting this 100% right for New England. So your notes above about incorporation changes are incredibly helpful and I for one certainly haven’t abandoned this effort.

I do agree that whatever we tag the unincorporated townships, plantations, and gores in northern New England, it should be consistent across the three states and my sense is that “everything admin_level=8” is probably the leading proposed solution at this point since they are at the same topological level.

3 Likes
44

I disagree. Each state is its own entity, meaning that “containment and equivalence” only happens within a state, a “key point” that is often overlooked. Those quotes are from @Minh_Nguyen , in the original post of this topic.

Everything else said in @A_Hall 's post (thank you for making it!) is quite agreeable.

@ZeLonewolf 's post is correct as well. I think we have most or all of what we need (@blackboxlogic : an opportunity for your input is going, going…) to make these “suggested edits” and after we do these, Maine, and the process we develop and are using to determine how to tag in this New England state can be used to “clone” (but not directly and exactly) — better inform, actually — other admin_level=* disputes / disharmonies which might exist in other New England states.

It might feel we creep forward (only) millimeters at a time, but I do see forward progress here!

Edit: A point of clarification. If it turns out that the three states mentioned (Maine, New Hampshire, Vermont) do end up with “consistency w.r.t. admin_level=8” it won’t be because “we think that is a good idea,” it will be because the names (like “gores” and peculiar-to-here instances of “incorporated or unincorporated townships”) happen to align with each other, similar to how a “county” in California is pretty much the same thing (a state political subdivision) as in Oregon. But not exactly, and only “loosely coincidentally.” I keep saying it: “this stuff ain’t easy.”

45

I very much like / agree with most everything I see here (and appreciate its clarity and being shared), especially:

Makes sense. If there were to be an historic boundary as Minh asks, I think that might be an OHM datum (right?) while this one would enter as described here: not an admin_level=*-taggable entity, rather a “survey square” with a name instead of a numeric survey moniker. It gets a boundary=place tag, like T10 R12 WELS does (or should). After all, both are unorganized “townships” which really describe what were/are originally “survey squares” unless they become incorporated, where it seems such a “plantation” rightly gets admin_level=8 (by agreement here). And they can wink out (and back in) to being incorporated; that’s when admin_level=8 gets applied or stripped off.

What the Census Bureau does might indeed be exciting, but I find such boundaries less interesting and indeed less authoritative than what it appears the whole world does, namely “tag what is.” We can and do correctly delineate these with boundary=census. Though, there are quite a number of such boundaries, flavored as MSAs, µMSA, PSAPs and on and on. It would be easy for OSM to get carried away entering these, especially as they do not age well, quickly becoming out of date as soon as they are uttered.

The admonishment to “add coding to the Maine ‘blocks’” is easily agreed to, so we get closer. As long as these are unique enough to identify as from the Census Bureau (“in opposition” to what OSM considers a boundary=place); as differing values of place and census enforces this, we’re good.

Vermont having its unorganized areas should similarly be tagged without an admin_level=* tag, ditto if Vermont’s gores are unorganized (that is, use boundary=place on the polygon). The only attempt OSM should make to “align” this “truth in tagging” isn’t to Maine and/or New Hampshire, but to the syntax and semantics of OSM’s tagging (for boundary=* and admin_level=*). It’s difficult at times to do this (misunderstandings and disagreements are frequent), but we enjoy these conversations together and keep getting closer and closer to better and better.

Keeping track of recent disorganizations (like Magalloway and Codyville) means we have our finger on the pulse of updates, either potential or actual. Our map data are a living thing, really. If Drew is disorganized, let our tags reflect that. Thanks to all who update our map, it can feel like a thankless task.

I think as simply as I can state it (for Maine): on incorporated townships, tag admin_level=8; on unincorporated townships, current tagging is correct, EXCEPT for admin_level=9, which should be deleted.

For Moosehead Lake, I ask that somebody (here) “utter what is so” about these, as they fit into the state’s structure somewhere and perhaps admin_level=* tagging somewhere, even if it means they are a “special case.” If so, say so on the Maine row(s) of our border_type=* and admin_level=* wiki. That is what we are hammering out here and now: entries in each cell in Maine’s row(s) so that anybody reading it/these will know what to tag, where and how.

Regarding reservations (and First Nations), it has emerged / is emerging (certainly in USA and Canada) that how admin_level=* tagging happens is “by the people of the land.” In places, we already see this happen: so far, so good.

I think that’s what’s happening here, too: people from New England states largely guide this, there are a couple folks out in California (Minh and I) who have chimed in, but, OK. On the whole, I believe that a blend of “locals know best” combined with “OSM data have a structure and guidelines” is working well.

Green lights ahead from what I can see and read from here, ladies and gentlemen. Brian, thanks for your evident and wonderful heavy lifting. Heck, thanks to everybody here, really. Excellent work so far!

46

In OSM, boundary=historic is supposed to be for a boundary that’s relevant in the present day – observable through boundary markers, verifiable through word of mouth, expected on maps, etc. – despite no longer being an official administrative boundary. The same boundary in OpenHistoricalMap would be mapped twice: one relation tagged boundary=administrative during the period of significance and another tagged boundary=historic up to the present day.

1 Like
47

While I like “leaning heavily on the border_type=* tag,” (as Zeke) remaining ambiguous (to me) are:

• some may continue to feel “three-state harmony is a goal to be achieved,” while I disagree,

• Zeke’s proposals all contain admin_level=8 tagging, even for unorganized areas, supported by Vermont having all land “contained within one (and only one) city, town, unorganized town, or gore.”

So, rather than what I stated earlier that “flipping into being organized” is whether an admin_level=8 tag is included (or excluded), instead, border_type=township becomes border_type=plantation (or vice versa). But then I see (and agree with) @A_Hall 's assertion that "an unorganized township (Forest City Township) has the same administrative responsibilities (none) as any of the survey blocks, and I’m back to thinking that “no administrative responsibilities = no admin_level=* tag.” But, it’s possible that it’s a “Maine, as a state (4), administrates these as an 8” situation. OK, if that’s correct, that’s how OSM should tag. But it’s hard for me to see that.

Others here (especially in New England): can you more fully sharpen this up, please?

48

I think this is too narrow of an interpretation, and it’s one that’s more suitable in states that have tracts of unorganized land from which incorporated places are carved out of. New England is fully space-filling from top to bottom (4->6->8) (state->county->city/town/town-in-waiting), and we’ve always taken the interpretation that the states have chosen to administer their territory by pre-determining how it will be divided and that the exact nature of governance is not what we are assessing with admin_level. What we are assessing with admin_level >=4 is the vertical location in the topological hierarchy of the primary political division of a state. This is why we’ve stuck with admin_level=6 for RI/MA counties that don’t really do anything, and switched them over to CT’s planning regions when they became the primary political division rather than counties.

I think the fact that people of T10 R12 WELS can chose to vote in or out a local government under its present borders (should anyone choose to move to the abandoned logging camp there) is plenty sufficient to recognize it as the primary political division. The fact that it has a silly name (that would no doubt get renamed once the people outnumber the Moose) doesn’t negate for me that it’s a present-day and not a historic boundary.

2 Likes
49

Thanks for staying up late, Brian!

I can see how a place like the island of Maui, Hawaii “stops at 6” (no further admin_level=* with larger values), meaning that all government in Hawaii is either state or county government. Or, perhaps similarly, even Johnston Atoll (tagged admin_level=2 to represent that the US Department of the Interior administrates at a federal level out to the territorial waters, and again, there are no further admin_level=* with larger values). What I mean is that “the People” (who live on Maui, Johnston Atoll is uninhabited) “have not yet” decided to “subdivide” below 6, so there shouldn’t be a smattering of polygons numbered 8 where they might. Even if these are/were part of a set of survey squares at some point: these aren’t the same thing, even though “they could, if people created them and voted them into a new existence” (organizing).

In Maine, “the People haven’t yet, but could” is a good argument for recognizing that the survey squares already exist, and aren’t really anything yet, but could be with a vote of the People. But that’s not the same thing as saying that admin_level=* is appropriate, as I don’t think it is. I can see how People in New England might wish it to be true, and so “by convention” tag it (with an admin_level=* value, likely 8), but that wouldn’t make it true. Similarly, saying that a Connecticut COG / RCOG “should” get admin_level=6 (despite Connecticut dissolving its county governments around 1960) doesn’t make what isn’t really a “full rainbow government” (with taxing authority, police powers…) a “government” (which I continue to contend is what admin_level=* tagging is all about).

But if New England wants to adopt such “conventions” — I’d argue contrary to how admin_level=* is "meant to denote " — well, I’ve certainly found myself on the other side of those discussions before. I think doing this muddies what OSM (-US) means when it does such admin_level=* tagging, diluting its effectiveness somewhat, but if “that’s what the locals want to do,” I can zip my lip about it, certainly.

I’ll watch and listen more here for now, and participate less for now. Wider discussion can no doubt help.

50

Maui County has not further subdivided its territory and thus there are no higher-numbered admin_level tags. Simple, and I think there’s no disagreement there.

And I and others contend that while these concepts often overlap, it is not what admin_level is about, nor is there any confusion in New England about which boundaries are the primary administrative ones. Different opinions are fine of course, but let’s not pretend there is any authority to appeal to beyond our personal assessments about what tagging means and how concepts equate or don’t between states.

It’s unnecessary to toss in the scare quotes to imply that we’re doing something outside the norm and I strenuously object to the notion that we are adopting unusual and/or inconsistent applications of admin_level. I have also yet to hear of a single negative impact from any data consumer who has been troubled by the precise nature of New England intra-state governments and their association with admin_level tagging. So far all the New Englanders commenting here seem to be in alignment on how our boundaries are organized and tagged and feel its consistent with the rest of the country.

So far nobody has come up with equivalent places in other states that are analogous to the ones in NH/VT/ME. If you have such evidence, please produce it. If you have examples (real or hypothetical) about data consumer interpretations and how applying admin_level in certain ways would impact them positively or negatively, then by all means put that on the table. Otherwise this is simply re-hashing “feelings” about how admin_level “should” be used and it’s just the New England counties fight all over again.

Just to be clear with my board hat on, OSM-US (the local chapter in the United States) takes no position on tagging.

51

I had to look up “scare quotes” as I meant no fear-mongering. I’m discussing what I understand (about admin_level=*, and it certainly has changed / morphed / evolved over the years) compared to a potential new tagging scheme that might or does depart from what I know. As I nod my head that “each (country, state/province, county, township, city, town, village, neighborhood…) level is unique in its own space of the hierarchy,” I understand this at a more fundamental and deeper level, perhaps “like a person from there,” but please understand that we all must on occasion “put on our neighbor’s shoes and walk in them.” No hostility, confusion or rancor, deliberate or not, should arise from these “hm, thinking about it from YOUR perspective, let me try that on for size…”.

There are also real aspects of tags (especially admin_level=*) being “stretched to accommodate.” OSM experiences a fair bit of “the meanings of tags can and do change over the years.” Hopefully not so much that they break, but again, especially with admin_level=* there must be some flexibility. Things bend, they don’t seem to be breaking. I think that’s what we’re seeing here.

Again, patience, sometimes a reiteration that it is valuable to “try on” (and mutually respect) the perspectives of others goes a long way. I thank you for your explanations, and if putting quotes around “conventions” has offended anybody, I apologize. And this is not a court martial, although any of us here is perfectly welcome to “strenuously object.” (A fine movie was linked, if you follow it).

There is something to be said for how a tag “should” be used, just as there is value in allowing tags to be as flexible as we have allowed them to become (bending without breaking). We have found a happy medium, it seems, and I don’t wish to re-hash anything, I certainly am not picking a fight.

52

I don’t agree with this. Vermont towns and gores have existed as administrative entities for hundreds of years. The fact that some are not organized or incorporated does not mean they are not administrative entities. It just means they don’t have a local government based within the administrative area. Instead they are administered by a state representative. I’m less familiar with the other New England states administrative divisions, but the more I read about them, the more similar they all seem. I understand this region has a rather different adminsitrative model than the rest of the country, but that’s how things are (and have been for centuries) around here. Good reading on the topic here: New England town - Wikipedia

The boundary=administrative wiki page says the tag is for “Subdivisions of areas/territories/jurisdictions recognised by governments for administrative purposes”. There is no requirement that the administrative area have its own incorporated government. A subdivision of a larger administrative area is sufficient. Removing boundary=administrative and admin_level from unorganized New England towns and gores is not the right thing to do.

2 Likes
53

Got it, Zeke. This was “clear as mud” to me an hour ago, now it is “crystal.” (Thanks, Brian, that’s another quote from that movie).

1 Like
54

The only thing that gave me pause about using boundary=administrative for the more obscure township boundaries is that we’re skating very close to the verifiability line. After many years, OSM has finally moved beyond the contentious arguments about whether any administrative boundaries should be mapped. I think we got here only with the understanding that, apart from some edge cases, each kind of boundary is somehow partially verifiable on the ground and generally useful for describing a location within it.

The rule of thumb I’ve always used is something like: Would there likely be a welcome sign at this kind of border? Conversely, would a well-informed resident be surprised to discover that they’ve been living “in” this kind of jurisdiction all along (emphasis on the preposition “in”)? I can see how a historic township, previously organized, might easily pass both of these tests. There’s probably still a vestigial sense of community there to boot. But what about a town-in-waiting with nothing but an unwieldy survey-era placeholder for a name?

I would defer to a local (or a regional) on this question, but I think it’s worth paying attention to whether we’re weakening the argument for boundaries in general by tagging the more theoretical boundaries as mostly the same thing. If we can point to examples of general-interest maps that depict these townships, news articles that locate events “in” these townships as a routine matter, and so on, that would help us avoid the slippery slope arguments. It would keep enterprising mappers elsewhere in the country from looking to Maine townships as precedent for tagging their planning districts and such as administrative boundaries. This is not a theoretical prospect.

55

Sure, are these sufficient?

From the Daily Bulldog, December 27, 2023, Rural Maine state representatives eye a seat in Washington, with Republican primaries in June 2024:

Michael Soboleski of Phillips currently represents Maine’s 73rd district in the state legislature, which includes the following towns: Carrabassett Valley; East Central Franklin; Kingfield; Wyman Township; Coplin Plantation; Dallas Plantation; Eustis; North Franklin; Phillips; Rangeley; Rangeley Plantation; Sandy River Plantation; West Central Franklin; Alder Stream Township; Coburn Gore Township; Jim Pond Township; Lang Township; Freeman Township; Madrid Township; Salem Township; Andover; Gilead; Lincoln Plantation; Magalloway Township; Newry; North Oxford; Upton; Adamstown Township; Andover North Surplus Township; Lower Cupsuptic Township; Lynchtown Township; Richardsontown Township; Riley Township; Dennistown Plantation; Highland Plantation; Jackman; Moose River; Northwest Somerset; The Forks Plantation; West Forks Plantation; Attean Township; Bowtown Township; Carrying Place Township; Dead River Township; Holeb Township; Pierce Pond Township; Raytown Township; Spring Lake Township; Upper Enchanted Township.

Or from thecount.me on June 16, 2022, where there’s a blurb about thefts in Bancroft Township:

BANCROFT TOWNSHIP — Tr. Saucier and Tr. Rider responded to a residence under construction in Bancroft when a neighbor reported there were multiple vehicles there stealing building materials. No one was at the property when troopers arrived. Tr. Saucier is working with the owner and the case remains under investigation.
BANCROFT TOWNSHIP — Tr. Saucier took a report from a New Hampshire resident complaining that his camp neighbor in Bancroft parked a trailer on his property and he wants it removed. Tr. Saucier is working on contacting the owner of the trailer.

Regarding:

I found this on Big Lake Township’s Facebook page:

image
image1512×2016 607 KB

In fact, I was surprised to find that even the so-called survey townships that you guys don’t like are signed at the town line in places where there’s actual roads running through them:

image
image800×465 52.2 KB

This is not a good argument. We’re not going to disqualify Maine township boundaries because someone in Illinois once tried to tag a dogcatcher district as a town. I’ve asked a few times in this thread for specific examples of analogous entities that would disqualify New England townships as administrative entities. I’m looking for examples where a named, bounded place has some or most of the following qualities:

  1. Is the primary administrative subdivision of a county or county equivalent
  2. Has fixed boundaries that don’t change with incorporation status (in other words, the act of incorporation does not carve its boundaries out of a larger area)
  3. Is roughly corresponding in relative size to other administrative subdivisions at the municipal level and does not overlap any other administrative subdivision at that level
  4. Receives all of its municipal services from a higher level administrative entity
  5. (Bonus points) has an established political process for establishing itself as an incorporated or semi-incorporated place.
4 Likes
56

Every map I’ve ever seen that shows town boundaries in Vermont includes boundaries and labels for unorganized towns and gores in the same style. Because of this, I was actually unaware that Vermont had any unincorporated areas until just a few years ago. Here’s my Delorme Vermont road atlas showing five unorganized towns and gores (highlighted pink by me) surrounded by organized towns. No styling distinction is made.

Vermont-unorganized-towns-and-gores
Vermont-unorganized-towns-and-gores1077×945 175 KB

New Hampshire’s unorganized administrative entities are maybe collectively called “townships” like Maine, but are also called “purchase”, “grant”, or “location”. Here’s my Delorme New Hampshire road atlas showing four unorganized grants/purchases (highlighted pink by me) surrounded by organized towns. Again no styling distinction. (Interestingly, Harts Location is an organized town, but has “Location” in it’s name just to confuse us :grinning:)

New-Hampshire-unorganized-townships
New-Hampshire-unorganized-townships1115×852 188 KB

I’m much less familiar with maps of Maine, but I’ve just taken a look at my Delorme Maine road atlas to see if and how it shows unorganized townships. Here it shows two unorganized townships (highlighted pink by me) surrounded by organized plantations and towns. Same styling. It does distinguish townships and plantations from towns by appending “TWP” or “PLT” to the name. The systematic township name/code is also shown.

Maine-unorganized-townships-named
Maine-unorganized-townships-named1400×985 271 KB

Moving north, here is an area where all the administrative units are unorganized. Some have human readable names, others have only systematic names.

Main-unorganized-townships-named-codes
Main-unorganized-townships-named-codes1400×938 237 KB

Moving even further north, here is an area where all the administrative units are unorganized and only have systematic names.

Main-unorganized-townships-codes
Main-unorganized-townships-codes1400×822 203 KB

Obviously this is only one map publisher and it’s possible others might choose to omit boundaries and labels for unorganized admin areas. I can say with rather high confidence that I’ve never seen a map of New Hampshire or Vermont that has made this choice. They generally show all town boundaries (including unorganized towns, townships, gores, grants, purchases, locations, etc) or no town boundaries. Maine certainly is a bit different just in the sheer quantity of unorganized townships that it has. So I won’t be too surprised if someone finds an example of a Maine map that shows boundaries & labels for towns & plantations but not for unorganized townships, gores, grants, etc. My Delorme atlas does not do this though.

6 Likes
57

Side note: I grew up in New England and have live most of my life here. My mental model of the world included the assumption that when you cross the boundary of one town you always enter another town. This had always been my experience. The question “what town am I in?” was always answerable. In a few cases it might be called a gore or some other name, but effectively it was the same thing. When I would travel to other states and “what town am I in?” was answered with “oh this isn’t part of a town, it’s just the county” that 100% did not compute in my brain. It took me quite a while to understand that the New England town grids (drawn up by English colonial governors well before independence) were not the norm in the rest of the country. It was a real :exploding_head: moment when I finally understood.

4 Likes
58

I cannot stress enough how accurate this mental model of cities and towns is to New Englanders, and it is also my experience. When we say that our towns are “space-filling”, this is what’s meant. “Unincorporated” is conceptually different here than it is elsewhere - it’s a characteristic of a town, not a description of the absence of one.

4 Likes
59

These explanations, and especially maps! are quite helpful to us Californians (and others, I’m sure) who “just don’t have these.” Yes, “unincorporated” not being head-wrap-around-able is an excellent flip-side of “walking a mile in the other’s shoes.” Thanks to everyone for the great expositions of the truly interesting country (and its admin_levels) that we live in.

1 Like
60

Yes, your examples seal the deal for me. Nothing about these mentions and signs implies a historic or future quality.

That wasn’t my contention. I was suggesting that we have something to back up the heretofore claims of significance, that we can use to show future mappers in other places why they’re doing it wrong. You’ve provided exactly what’s needed. I don’t think we even need to get into the minutiae about the nature of municipal services and political processes when making this determination.

I agree, though I think it’s just a matter of degree. Having grown up in Ohio, I had the same experience one day realizing that, right across the border in Kentucky, the counties aren’t also divided into townships. But then again, Ohio townships have portions that additionally lie within a village and portions that do not.

2 Likes