WeRelate talk:Places



Call for volunteers

The WeRelate place index was generated by combining information from several sources: the Getty Thesaurus of Geographic Place Names, the Source:Family History Library Catalog, and Source:Wikipedia. This has given us the most comprehensive index of current and historical inhabited places available. However, sometimes places were merged incorrectly, or weren't merged when they should have been. Also, there are still places that are not in the index but should be. We are looking for a few people to take charge of regions they are interested in, as TomChatt is in charge of Scotland, see Place talk:Scotland. We are looking for help with the US, Canada, England, Ireland, Europe and others. Please leave a comment here if you are interested. Thanks --sq 15:05, 15 January 2007 (MST)

Not sure if this is too narrow, but I can do Mississippi, US. --Joeljkp 15:32, 15 January 2007 (MST)
That would be wonderful. Go to Place:Mississippi, United States. You should be able to browse the pages from there. Do you have any friends that would be interested in other areas? One other thing we were thinking about for sometime next year is to have a research flow chart for each region. Something that would help beginners understand the nuances of a particular region. What do you think?--sq 21:46, 15 January 2007 (MST)
Well, maybe I should clarify a bit first. I live in Mississippi, but all of my research is in New England. I could take care of Mississippi if what we want is an accurate set of place names, etc., but I wouldn't be able to offer anything in the way of research advice. If that's the goal, I should probably help out with Massachusetts. --Joeljkp 22:38, 15 January 2007 (MST)
It would be great to have the place names. We need to make sure the counties are counties and town are inside counties not inside towns. The research flow charts are a long way from production right now. Thanks.--sq 08:43, 16 January 2007 (MST)

FYI I've created a couple of pages for the Scotland project that may be a useful example for organizing some of the other regions as well. First, I did a page explaining how places are (ought to be) organized in Scotland. Then I did a How You Can Help page, with a sign-up / status sheet for the counties, and some step-by-step instructions how to "adopt a county". Check it out! --TomChatt 02:01, 18 January 2007 (MST)

I would like to research the place names for Place:Franklin, Illinois, United States. I have already added one Place:Dog Prairie, Franklin, Illinois, United States because this was the location cited on the marriage record of my great grandparents. Having a category "inhabited place" is a great idea because it can include small named communities such as Dog Prairie or Palestine where a group of families and homes might be grouped around a store, school or church.

Is the fact that US counties have two different categories a problem related to what you're talking about? For example, Category:Alexander County, North Carolina consists largely of sources. Category:Alexander, North Carolina, United States has just the Place article for the county. Is this something we can fix when we find it or is it in need of some larger fix? --Amelia.Gerlicher 12:03, 28 April 2007 (MDT)

This is in need of a larger fix. The problem is that most of the source pages contain places of the form "X County, State", which place is now redirected to "X, State, United States". But this causes the sources to be placed in the wrong category. We need to go through and update all of the sources to use the redirected-to place names. We'll be doing this hopefully about the end of May, at the same time that we standardize place titles for non-US and Canada places.--Dallan 09:58, 30 April 2007 (MDT)

We're getting ready to rename all non-US and Canadian place pages to their "full" names (including all levels in the jurisdictional hierarchy) starting October 27th, just as we did for US and Canadian places last Spring. Before we start the renaming, we could use your help to review the "located-in" field for places to make sure it's correct, since once we add all of the levels to each place's title, an incorrect located-in field will be much harder to change; it would require renaming the place page and all of its contained places as well. Please see WeRelate:Place review for more information.--Dallan 15:41, 11 October 2007 (EDT)

Historical Place Names

What is the protocol for assigning events to historical places? For example, Thomas Rogers died in the winter of 1620/1621 in what he would have known as Plymouth Colony, Kingdom of England. However, the closest I can come to this is to assign the place to the present-day county of Plymouth, Massachusetts, United States. Would it be possible to create a new page for Plymouth Colony, and link it somehow to these newer places? How should this be handled? --Joeljkp 10:44, 6 March 2007 (MST)

Wikipedia associates "Plymouth Colony" with the modern day city of Plymouth, in the county of Plymouth, Massachusetts, US. This city has a place page on WeRelate. Here is the link to its place page: Place:Plymouth,_Plymouth,_Massachusetts,_United_States. You will see in the Place Information box on the right of the screen that an alternate name is listed as "Colony of New Plymouth". I would use this place as the death place for Thomas. Anyone else have any ideas? --Wrhelp 17:01, 6 March 2007 (MST)
I suppose my point above was that using that non-linkable alternative name system means that in the person page, the death date is listed as modern-day "Plymouth, Plymouth, Massachusetts, United States." Any viewers of this data would need to find for themselves what the place was called at the particular time of the event. As far as I know, the standard practice by genealogists in situations like this is to put the name of the place at the time the event occurred, which could be translated into a modern place if the reader wants. So in short, the most common thing to do (as far as I know) in offline genealogy is to put the date of the event and the name of the place at the time of the event. Here, we end up putting the date of the event and the modern place name, with no indication of its name at the time without further research. --Joeljkp 18:26, 6 March 2007 (MST)

This is a really good question. There are several issues:

  1. Genealogists say that you should record the name of the place as it was when the event occurred.
  2. We'd like to link the place to a category that includes the sources (and maybe people and families) that pertain the place, and that's easier to do if all names for the place point to the same category.
  3. Later when we try to determine automatically whether two wiki pages are likely for the same person, it will help if all names the place was known by are linked somehow.

Here are a couple of thoughts:

--Dallan 22:03, 6 March 2007 (MST)

Thanks for the response. There are a few issues that need to be considered before we can decide the right way to do things.
  1. The person/family page for the event should display the name at the time the event occurred.
  2. Sometimes, historical place names are all that is known, and they don't match well with modern place names. For example, if I knew only knew John Doe was born in Plymouth Colony, I wouldn't know if he was born in modern-day Plymouth County, Bristol County, or Barnstable County, so redirecting to any of these would be misleading.
  3. Places should be linked together on their place pages (a person viewing the Plymouth Colony page should be linked to the county pages above)
  4. It makes no sense to have categories for historical places separated from modern places, since that divides things into blocks of time (like having a category for Kingdom of England and another for United Kingdom).
Maybe this needs to be taken case-by-case, and maybe I need to do more research on the Plymouth Colony example. For Person:Thomas Rogers (1), all I know from my one source is that he died in the winter of 1620/1621 that took half the pilgrims. If I find out (as is likely) that he died in New Plymouth (the colonial town), this can pretty easily be linked somehow to the modern town of Place:Plymouth, Plymouth, Massachusetts, United States, either with a piped link or using a redirect. However, for a person who died somewhere in the whole of Plymouth Colony (which was a rather large area), it becomes more complicated, as this was a historical name for several modern-day counties.
I hope laying this out will help you decide a good course of action. --Joeljkp 07:10, 7 March 2007 (MST)
Thanks Joel. I hadn't considered the case where the historical place doesn't correspond to a modern-day place. I believe that historical places that don't correspond roughly to the same "piece of dirt" (area) as a modern-day place deserve their own place pages. You could then set up "see-also" links between the modern-day place(s) covered by the historical place and the historical place (there is a separate field on the place page for see-also links). Under this approach the historical place would have a separate category than the modern-day places it covered, which is less than ideal, but I'm not sure what else to do when a historical place covers or overlaps multiple modern-day places. How does that sound?--Dallan 20:31, 8 March 2007 (MST)
Yes, that sounds fine. Thanks for the help. --Joeljkp 21:58, 8 March 2007 (MST)
Thank-you for bringing up the issue. Are you going to create the place for Plymouth Colony?--Dallan 22:21, 8 March 2007 (MST)
I already did :-) : Plymouth Colony, Kingdom of England. It brought up a few other issues, though, about naming of countries and such. We can discuss here or take it somewhere else. The jist is that according to Wikipedia, the place we know as the UK changed names several times between 1649 and 1660, from Kingdom of England to Commonwealth of England, to The Protectorate of England, Scotland, and Wales, back to Kingdom of England. This just brings up the issue of a given place being named different things over time, and encompassing different areas, etc. This becomes a problem if we name pages with the country name in them (officially anyway, it was "Plymouth Colony, Kingdom of England" part of the time and "Plymouth Colony, Commonwealth of England" other times). Either we need to pick a common system for all these pages, or we'll end up with x copies of every place in the UK that happened to exist when the country changed names. I had thought maybe naming pages just by the most local name, then automatically tacking on the other names based on the dates associated with the reference, but I don't know how feasible that would be. --Joeljkp 08:06, 9 March 2007 (MST)
I'm not sure how to tack on the name based upon the time period, because each page needs to have a single unique title. If the area covered by the Kingdom of England didn't change significantly between 1649 and 1660, perhaps we should just use alternate names for Commonwealth of England and Protectorate of England, Scotland, and Wales. Also, if "Plymouth Colony, Kingdom of England" and "Plymouth Colony, Commonwealth of England" cover the same piece of dirt, perhaps "Plymouth Colony, Commonwealth of England" could be a redirect to "Plymouth Colony, Kingdom of England". I think that establishing pages for all of the historical place names is going to take some concentrated effort on a country by country basis.--Dallan 09:41, 10 March 2007 (MST)

Disambiguation Pages

I think disambiguation pages would be a great addition, especially for places. Many sources specify only a town name, and linking that name to a specific place would mean inferring details that may not be sound. Instead of placing a birth at, say, Duxbury, you could place it at Duxbury. This latter page would be a disambiguation page stating something to the effect of "Duxbury is a place name that can either refer to a town in England or a town in Massachusetts," etc.

Can we create these now by creating normal place pages and just not filling in the specific details?

--Joeljkp 22:24, 23 March 2007 (MDT)

I'll change the place pages so that you can leave all of the data fields empty. This should allow you to create disambiguation pages. I'm wondering though - there are a lot of potential disambiguation pages - instead of creating all of the disambiguation pages, would we be better off showing all possible matching place pages when people follow a "broken" place link, like an auto-generated disambiguation page? So when someone followed the link Place:Duxbury, they would see a page listing links to all existing places with Duxbury in the title (like what you see in the auto-completion drop-down), followed by a link to create a new Duxbury page.--Dallan 12:37, 26 March 2007 (MDT)
That might work. The only reason we'd need a human-generated one would be if there needed to be notes or something about a specific name, like if it's common for some reason to mention a place by an ambiguous name, but that a vast majority of such references are referring to one of the possible options, and a user would want to be notified of this (e.g. "Smithfield refers to an ancient French town that was destroyed in 1259 AD or a modern city in the UK. The vast majority of references to Smithfield refer to the modern city"). I don't know, the examples in front of me don't really need a disambiguation page anymore (I found more details in my source). --Joeljkp 13:59, 26 March 2007 (MDT)
I'll add auto-generated disambiguation pages for places to the "todo" list. And tomorrow I'll remove the requirement for each place page to have a preferred name so that people can create disambiguation pages by hand whenever they want to provide additional information as you suggest.--Dallan 23:00, 27 March 2007 (MDT)

More on representing historical places

the following was moved from User talk:Dallan

Dallan --

User:DebbyV tells me there may be something in the works for places that used to exist but don't any more. I keep running across them as an issue that puzzles me enough that I haven't marked them as patrolled.

She has a couple of new places in Franklin County, Illinois, in this category:

Place:Webb's Hill, Franklin, Illinois, United States

N! Place:Abe, Franklin, Illinois, United States

(let me know if I'm not cutting and pasting these right)

She's listed these places as formerly contained in Franklin County, which makes sense except that that designation is also used for places that changed counties due to boundary changes, rather than going out of existence altogether. But if we say that Abe is now in Franklin County (because she could take me there, and then we'd be in Franklin County, it's just that no one lives there any more), then it seems funny to call it an "inhabited place" when it's no longer, well, inhabited. She and I have discussed this a bit.

Similarly, User:MMartineau has added a new place

N! Place:Utah Territory, United States

which strikes me as a good idea (without having thought through the implications), but it gets tricky because now Provo is listed as a formerly contained place, and when I click over to Provo, it is said to have been in Utah, Utah, United States since 1849, and is also said to have formerly been in Utah Territory 1850-1896.

I tend to be careless about this sort of thing in my own work (listing 17th-century Puritans as living in the state of Massachusetts rather than the Massachusetts Bay Colony, for instance), and I haven't thought through the implications of various ways of handling this, especially situations where a historical place used to contain a bunch of currently existing places. I can try, but hopefully you already have!

--Hh219 12:49, 3 May 2007 (MDT)

Handling historical places well without making the place index incredibly complex is a really hard problem. I've been thinking about this for a while and don't have a solution that I'm completely happy with, but here's the best I've been able to come up with so far:
  • A lot of people want to list places according to the jurisdictional hierarchy that was in effect when the record was created. Our jurisdictional hierarchies are encoded in the place title. So that means we need to have separate place pages when the jurisdictional hierarchy changes. So if a city used to be located in county A but is now located in county B, we should have one place entitled "city name, A, state, united states" and another place "city name, B, state, united states". These two places should have "See-Also" links pointing to each other to let researchers know that both places really reference the same "piece of dirt".
  • I'm thinking that we should eventually drop the distinction between the located-in and previously-located-in fields on the place pages, and combine them into a single field with a date range. The list of contained places would look at the date range to determine whether each contained place was still current or not. In the previous example, the page for "A, state, united states" would list "city name, A, state, united states" as historical because it is no longer located in A, while the page for "B, state, united states" would list "city name, B, state, united states" as current.
  • Hopefully we won't get carried away with coming up with all possible combinations of jurisdictional hierarchies over time. It would be easy to explode the place index beyond usefulness if this happened. In particular, I'm not sure whether the distinction between Utah Territory and Utah (state) is significant enough to warrant creating a bunch of "X, Utah Territory, United States" places with "See-Also" links to the corresponding "X, Utah, United States" places.
  • For jurisdictional changes that aren't significant enough to warrant creating new place pages, we could still note both located-in places in the place page. So for example Utah County could be located-in Utah (state) from 1896-present, and also located-in Utah Territory from 1850-1895.
  • Perhaps this particular example is bad because the difference between Utah Territory and Utah (state) is significant enough to warrant creating additional place pages. I don't know. I hope that we can limit the number of pages we create for the same "piece of dirt" so that the place index remains easy to navigate.
  • Some places in Getty are listed as "former inhabited place". This type (or former village, former town, etc.) might make sense for places that used to be inhabited but no longer are.
--Dallan 14:17, 3 May 2007 (MDT)

The following are real questions, I'm not arguing one way or the other.
  • Would it simplify matters if there was only one working place page for Provo (under the current hierarchy, say) and the others just referred over to it? Then it would be possible to enter all data in a historically correct way without having several different Provo place pages to fuss about. The index itself might look pretty complicated, but would that affect users?
That's an interesting idea. The historical place pages could be redirects to the current (or most recent, if the place doesn't exist) place page. Both the real and redirected place pages could appear in the drop-down list for auto-completion. And the current place page could list all of the historical jurisdictions it used to fall under.
  • What would constitute an explosion of the place index to the point of unusability? As a searcher, I wouldn't mind seeing the redirects, or the list of earlier-date places-contained-in on Provo's place page -- it would be kind of a quick bare-bones history lesson. But if there had to be an actual separate functioning place page for Provo, Utah County, Utah Territory, United States, and another for Provo, Utah County, Utah, United States, that looks like bad news.
If we have the historical jurisdictions represented just as redirects and as earlier located-in places on Provo's page, then I think unusability depends upon how we represent Provo in searches and in auto-completion.
  • For search results: do we omit redirects from the result list so that you see only one result for Provo in Utah even though it may have occupied different counties over time, or do we include redirects in the result list but also display the page that they redirect to, so that you might see several Provo's but you can tell that they all point to the same place?
  • For auto-completion, I think we have to include redirects, but we should also include the current place that the redirect points to. So if Provo was in several counties, you'd see separate entries in the auto-completion list for Provo, but you could also see that they all pointed to the same current Place page for Provo.
  • On the other side, what would constitute a significant change of jurisdictions? My first thought is, a different location for records. Part of what is now western Pennsylvania was once a county in Virginia; Licking County, Ohio, was at one time a honkin' big township in Fairfield County, Ohio. In both cases one may need to search in what now seem unlikely places for information because of this history. (Obviously this should be elaborated under "research tips" regardless of how we wind up doing places.) Other changes, like from the colony of Connecticut to the state of Connecticut, don't seem to rise to this level of significance. Or do they?
This is my thinking as well.
  • If you've ever seen the Atlas of Historical County Boundaries for any given state, this is a major issue. Within a given present-day county, different towns could have different hierarchies in the past.
  • There even are some places in the US that were disputed territory for quite some period of time, like Vermont, which was claimed by New York. I'm pretty sure that some Vermont property records are still in Albany.
  • Would it work to adopt a burden-of-proof standard on some of these matters, such as creating an additional historical hierarchy only if someone could point out a significant difference it would make?
--Hh219 15:27, 3 May 2007 (MDT)
If we represented historical jurisdictions by redirects and additional located-in places (with historical date ranges) on the current place page, then having a place listed under multiple historical jurisdictions isn't nearly as bad as if each historical jurisdiction were represented by a separate place page. In this case I don't think we'd need to have as high a standard for what constitutes a "significant difference".
--Dallan 22:41, 3 May 2007 (MDT)
I like "Former inhabited place" much better than "Previously located in". And the new format for the places in the sidebar looks wonderful.--Debbie V. 05:58, 4 May 2007 (MDT)
Thanks. Then let's go with "Former inhabited place" for the "Type" of a city/village/town that no longer exists. And for now, record the place that it used to be located in in the "previously located in" field, so that you can enter a date range and so that the former place is marked as "historical" in the located-in place's list of contained place. Once we combine the located-in and previously-located-in fields, we'll have date ranges on everything, and places will be marked as historical if the date range has an end date.--Dallan 13:44, 4 May 2007 (MDT)
I think I might have mentioned this in an email, but there is a problem with trying to attach an end date to a place. Most places just fade away slowly. I am thinking the idea of distinguishing between current places and former places might not be that important. Especially considering most of our work is focused in the past and distant past. The status of a place could be a discussion point on the Place page itself. --Debbie V. 15:42, 5 May 2007 (MDT)
I think having an end date is more important when a place moves from one located-in jurisdiction (e.g., county) to another. But having an exact end date for a place that just faded away does seem a lot less feasible, and less important.--Dallan 22:57, 5 May 2007 (MDT)


Would it be possible to include a Google-type map on the Places pages, similar to the ones on the pedigree charts ? In some of my research I have been led to wonder how close in distance two counties were for example, and I had to go to the Wikipedia articles to find out. Probably not a priority but would be nice to have. --CTfrog 08:15, 16 July 2007 (MDT)

Yes, it's on the ToDo list already in fact. If you've ever looked at the "diff" screen for a place page (which you probably have), you've probably noticed that we copy the latitude and longitude of the contained places up to the containing place page. That's so that we can add a google map of the place and all of its contained places to the place page. It should be in before the end of the year.--Dallan 21:20, 19 July 2007 (MDT)

Questions about places. [4 August 2007]

I know you have pulled the pages from Wikipedia, but do they update? I fixed the Seymour, Indiana one at Wikipedia but it has not updated here. Will it update or can we fix them? Also, I am the Jackson County Coordinator on the INGenWeb and try to get photos when I can of places that are of interest in the county. I have the county courthouse for Jackson that I would like to put here and others. What I want to know is do we treat Places the same way we would our pages? Just edit?

FamilyTwigs - Twigs 09:58, 3 August 2007 (MDT) -

Don't quote me on this, but I believe I read somewhere on here that updates from Wikipedia occur once or twice a year. I also believe that any changes to a place page on WR will get overwritten unless the text is outside the wikipedia template. --Ronni 09:28, 4 August 2007 (MDT)
Here ya go Twigs: Help:Place_pages#Adding_research_tips_and_histories --Ronni 09:32, 4 August 2007 (MDT)

Don't know yet what I am going to do. I hadn't explored this part of the site till now. Thanks, Ronni! - Twigs 10:04, 4 August 2007 (MDT) -

What about TOWNSHIP'S in the United States? [14 October 2007]

Many of us have in our family a whole bunch of farmer's and before the time of hospital's these farmer's had their children right there on the farm. Thus I have folk's who's birth was (for example) in Loudon township, Seneca County, Ohio. I do not want that to appear as "Loudon, Seneca, Ohio, United States" (this gives a misrepresentation of a birth in a city called Loudon, which does not exist) and I do not want it just to show up as "Seneca, Ohio, United States" as I have researched to show a more narrowed down area and don't want that lost. Can I set up my places to look like this: Loudon (township), Seneca, Ohio, United States ? --Msscarlet1957 13:11, 12 October 2007 (EDT)

You certainly can. We don't have a lot of townships currently in the place index yet, but we do have several thousand, and many of them are of the form you mention: Loudon (township), Seneca, Ohio, United States. In fact, I agree that should be the preferred form for township titles.--Dallan 01:50, 14 October 2007 (EDT)

Cemeteries as Places? [5 November 2007]

I've recently jumped in on we relate. In the process of tidying up my GEDCOM upload, I realized that having cemeterys as specifically named places adds some real value. They obviously have lat/long values. There's a lot of source information that may be specific to a particular cemetery and would be handy on the place page, info on how to query if the town has stuff like that, etc. So I just started creating them as I find useful. I'm trying to follow the naming conventions as I understand them, for example "Leach-Wardwell Cemetery, Penobscot, Hancock, Maine, United States".

Still, I don't want to fly in the face of convention, so I'm soliciting guidance on the matter. Thoughts?--Jrm03063 17:31, 3 November 2007 (EDT)

Yes, you are absolutely right. It's Cemetery, town, county, country for the US anyway. Different countries have different hierarchy. But, its alway starts with the cemetery followed by the political subdivisions, smallest to greatest. Thanks for participating.:) --sq 21:06, 5 November 2007 (EST)

I've subsequently seen that folks are creating cemeteries as specifically named places, but I would like to know if my interpretation of the place conventions is generally acceptable...--Jrm03063 17:04, 5 November 2007 (EST)

I think that creating place pages for cemeteries makes sense. Here's a good example. There's been some previous discussion of pro's and con's at WeRelate:Cemeteries as Places; I personally think that the pro's outweigh the con's. At some point I plan to import all of the cemeteries from the GNIS database, but not until next year.--Dallan 23:36, 5 November 2007 (EST)

Longitude and Latitude from Google Earth [5 November 2007]

I've learned of a way to extract coordinates from a google maps session. It seems more precise and cheaper than picking up a GPS unit, going places, and saving numbers. I initially found it documented at [1].

One place where I used this procedure is Lorraine American Cemetery.

The description suggests that the coordinates are only correct if the map is centered by selecting an address, and that shifting the map (click and pull) to a different location will not work. I don't know about that, but here's what I do:

  • Position the map any way you like - set address, move, whatever
  • When you've zoomed in to the maximum on the location of interest, use the satellite view to see actual landmarks (you may have to move out a little, as maximum resolution does not always have a satelite view).
  • Right click and re-center to select a spot. For a cemetery, I pick the front entrance, since it seems more apt to give useful driving directions.
  • Enter in your browser bar/url field:

--Jrm03063 17:01, 5 November 2007 (EST)

This is great. Thanks. --sq 21:06, 5 November 2007 (EST)

I'll comment again on my own topic. An even easier approach is to use http://mapper.acme.com, which uses google earth and gives you a constantly refreshed lat/long value on the display.--Jrm03063 17:15, 5 November 2007 (EST)

This is very cool! I've added it to the place help page.--Dallan 23:36, 5 November 2007 (EST)