Familypedia
Forums: Index > Watercooler > Standardization of place names and categories


  • Okay, before this gets out of control we need to address something. We have pages and categories for places in which the names are not consistent country-by-county and even within countries. I am going to gather previous discussions, combine them, and address more issues.

Place/category names problems[]

In regards to place names for articles and categories. Please ignore whether or not the actual articles or categories exist.

Regarding articles, we have the basic continents:

  • North America (Born in..., Died in..., Married in..., etc.)
  • South America (Born in..., Died in..., Married in..., etc.)
  • Australia (Born in..., Died in..., Married in..., etc.)
  • Europe (Born in..., Died in..., Married in..., etc.)
  • Asia (Born in..., Died in..., Married in..., etc.)
  • Africa (Born in..., Died in..., Married in..., etc.)
  • Antarctica (Born in..., Died in..., Married in..., etc.)

We also have country articles and categories:

  • United States (Born in..., Died in..., Married in..., etc.)
  • Australia (Born in..., Died in..., Married in..., etc.)
  • England (Born in..., Died in..., Married in..., etc.)
  • New Zealand (Born in..., Died in..., Married in..., etc.)

We also have several divisions of country articles and categories:

  • New York (Born in..., Died in..., Married in..., etc.)
  • Pennsylvania (Born in..., Died in..., Married in..., etc.)
  • Georgia (U.S. states) (Born in..., Died in..., Married in..., etc.)
  • New South Wales (Born in..., Died in..., Married in..., etc.)
  • North Yorkshire (Born in..., Died in..., Married in..., etc.)
  • Västmanland (Born in..., Died in..., Married in..., etc.)
  • Gotland (Born in..., Died in..., Married in..., etc.)
  • Gelderland (Born in..., Died in..., Married in..., etc.)
  • Baden-Württemberg (Born in..., Died in..., Married in..., etc.)

and the thousand of others...

The problem is inconsistancy:

  • Chautauqua County, New York - specifies county and state (does not specify the country the United States.)
  • Meddo - specifies hamlet (does not specify the muncipality of Winterswijk, the province of Gelderland, and the country of the Netherlands.)

These are just two of the many examples I could give. Also, a too simplified of a name would not account for the many duplication of places names. Examples:

  • Georgia is a country as well as a U.S. state
  • Winterswijk is a town within the muncipality of Winterswijk.

Solution[]

Use country name in place name

We'll use our examples from before:

Instead of using these:

  • New York (Born in..., Died in..., Married in..., etc.)
  • Pennsylvania (Born in..., Died in..., Married in..., etc.)
  • Georgia (U.S. states) (Born in..., Died in..., Married in..., etc.)
  • New South Wales (Born in..., Died in..., Married in..., etc.)
  • North Yorkshire (Born in..., Died in..., Married in..., etc.)
  • Västmanland (Born in..., Died in..., Married in..., etc.)
  • Gotland (Born in..., Died in..., Married in..., etc.)
  • Gelderland (Born in..., Died in..., Married in..., etc.)
  • Baden-Württemberg (Born in..., Died in..., Married in..., etc.)

We should use these:

  • New York, United States (Born in..., Died in..., Married in..., etc.)
  • Pennsylvania, United States (Born in..., Died in..., Married in..., etc.)
  • Georgia, United States (Born in..., Died in..., Married in..., etc.)
  • New South Wales, Australia (Born in..., Died in..., Married in..., etc.)
  • North Yorkshire, England (Born in..., Died in..., Married in..., etc.)
  • Västmanland, Sweden (Born in..., Died in..., Married in..., etc.)
  • Gotland, Sweden (Born in..., Died in..., Married in..., etc.)
  • Gelderland, Netherlands (Born in..., Died in..., Married in..., etc.)
  • Baden-Württemberg, Germany (Born in..., Died in..., Married in..., etc.)

Then we can go on to use:

  • Chautauqua County, New York, United States (Born in..., Died in..., Married in..., etc.)
  • Erie County, Pennsylvania, United States (Born in..., Died in..., Married in..., etc.)
  • Bacon County, Georgia, United States (Born in..., Died in..., Married in..., etc.)
  • Newcastle, New South Wales, Australia (Born in..., Died in..., Married in..., etc.)
  • Hambleton, North Yorkshire, England (Born in..., Died in..., Married in..., etc.)
  • Västmanland County, Västmanland, Sweden (Born in..., Died in..., Married in..., etc.)
  • Visby, Gotland, Sweden (Born in..., Died in..., Married in..., etc.)
  • Winterswijk, Gelderland, Netherlands (Born in..., Died in..., Married in..., etc.)
  • Heidenheim, Baden-Württemberg, Germany (Born in..., Died in..., Married in..., etc.)

There. That solves the inconsitancy problem and the duplicate place names problem. Too long of a name? No problem. We have redirects. We can make disambiguations.

Info page parameters[]

New York, Pennsylvania, Georgia, New South Wales, and Baden-Württemberg are states, North Yorkshire is a county, Gotland and Gelderland are provinces. Thus, all countries are subdivided differently. There are countries, states, provinces, districts, muncipalities, counties, towns, cities, villages, hamlets, townships, etc.

I once proposed this and I think it needs to be the standard.

So, the problem we currently have is we only use town, county, and state, in which not all countries are divided.

|Birth town       = 
|Birth county     = 
|Birth state      = 

Thus, the better info page parameters are:

|Birth country subdiv3 = 
|Birth country subdiv2 = 
|Birth country subdiv1 = 
|Birth country         = 

That way it can apply to all places generally. If the place actually uses "County" or something like that in it's place, it will be put there like such:

|Birth country subdiv3 = Ellery
|Birth country subdiv2 = Chautauqua County
|Birth country subdiv1 = New York
|Birth country         = United States

Which gets us down to the township. And perhaps a more detailed division:

|Birth country subdiv4 = Byrnsdale
|Birth country subdiv3 = Jay
|Birth country subdiv2 = Elk County
|Birth country subdiv1 = Pennsylvania
|Birth country         = United States

Which subcategorizes at the township level. Of course, townships may infact have so little of a population, that dividing them up may be unnecessary.

Forums/location templates/info boxes[]

If all the above are applied, I can make one forum template, one location template, and one infobox template in which we can apply to any location and any subdivision.

Conclusion[]

If all my proposals are followed, we would be better organized, consistant, and less confused.

What I meant earlier by out of control is that we need to address this at 4,300 info pages and not 100,000 info pages as well as an enormous amount of place pages. -AMK152(talkcontribs) 16:54, 6 April 2009 (UTC)


Standardisation necessary and possible[]

I'm all for standardizing what can be standardized. However, countries are structured in very different ways. What is logical in one place, is awkward in another, and impossible in a third place. I would think we want to standardise per country, and then aggregate from country to continent. Rtol 17:59, 6 April 2009 (UTC)

That's why I used subdiv. Every country that is subdivided can use those categories. If a country is not subdivided, just use the country (like Vatican City). -AMK152(talkcontribs) 19:35, 6 April 2009 (UTC)

Many thanks to AMK for putting all of that together. The idea sounds fine in principle. However, I'm at least partly inclined to favour Rtol's response here. Or to say "Why stop at the country level and not go on to include continent name?" (because not everybody knows where Oman is). Maybe AMK can produce a table of country subdivisions (doubtless copying most of the data from Wikipedia, e.g. Wikipedia:Category:First-level administrative country subdivisions, and maybe doing only twenty countries initially)) showing how they name their first-level subdivisions, second-level subdivisions, and so on, with footnotes showing what overlaps they might have (e.g. the city of Chicago is in parts of two counties and New York City encompasses five). Then (or first) AMK could make his proposed "one forum template, one location template, and one infobox template ... which we can apply to any location and any subdivision" and show us examples of what those "one size fits all" templates actually look like and do. And category templates along the lines of {{Born in UK}} and {{Ctycat-LA}}. — Robin Patterson (Talk) 06:31, 7 April 2009 (UTC)

Studying this a bit more, we have language problems already and now we are introducing further problems of understanding. I know that Zwaag is village, but how do I know it is subdiv4 and not subdiv5 or subdiv3? This would only work if the info boxes show aliases rather than the logical structure. Rtol 09:11, 7 April 2009 (UTC)
is this usable? Naams- en datumconventies

Name and Date Conventions |color=rgb(216, 191, 216) |content=

  • Datum: dag, maand, jaar. bijv.: 19 mrt 1874.
  • Plaats (voor Nederland): dorp, gemeente, provincie, Netherlands
  • Plaats (algemeen): town/city, county/district, state/province, country.

Fred Bergman 09:43, 7 April 2009 (UTC)

Subdividing countries[]

When figuring out how to divide a country, first ask yourself: What is the country divided into? For example, states in the United States would be subdiv1.

If subdiv1 is further divided, take what you have for subdiv1 and ask yourself: What is subdiv1 divided into? Using the state of New York, it is divided into counties. This would be subdiv2.

If subdiv2 is further divided, take what you have for subdiv2 and ask yourself: What is subdiv2 divided into? Using the New York county of Chautauqua, it is divided into two cities and several townships. This would be subdiv3.

and so on until you get to towns, villages, hamlets, cities, etc.

Examples:

United States[]

Country: United States
subdiv1: Pennsylvania (a U.S. state)
subdiv2: Elk (a county of Pennsylvania)
subdiv3: Jay (township of Elk County)
subdiv4: Byrnedale (a village within the township of Jay)

The Netherlands[]

Country: The Netherlands
subdiv1: Gelderland (a province of the Netherlands)
subdiv2: Winterswijk (a munincipality of Gelderland)
subdiv3: Meddo (a hamlet/village of Winterswijk)

Sweden[]

Country: Sweden
subdiv1: Västmanland County (a county of Sweden)
subdiv2: Norberg Municipality (a munincipality of Västmanland)
subdiv3: Norberg (a town within Norberg Municipality)

-AMK152(talkcontribs) 22:23, 7 April 2009 (UTC)

It's not as simple as that, even in the US. As I've said before, settlements cross boundaries, and not every genealogical record pinpoints the locality accurately enough for determination. Are you sure "Chautauqua ... is divided into two cities and several townships" and not just into townships?

  • We cannot ignore New York City merely because it covers five units of "subdiv2"; where would you put it - maybe on a parallel "settlements" hierarchy?
  • Four of New Zealand's local government regions (subdiv1) have no separate councils but are administered by the council of the city or district that they contain; how do we get a resident of New York to realise that "Motueka, Tasman, New Zealand" should be "Motueka, Tasman, Tasman, New Zealand" and/or not to leave a blank in the "region" line of a template?
  • Belgium is divided into "three communities, three regions, and four language areas", each set covering the whole country in a slightly different way and having relative autonomy in defined respects.
  • Much of Australia has overlaps that are similar though less complex: counties (not covering every state) are basically only for defining land ownership, though some have had governing powers, whereas a Local Government Area covers all or most of each of the most populous states with little or no respect for county boundaries.

Robin Patterson (Talk) 05:50, 8 April 2009 (UTC)

Divide by continents?[]

I would agree to this. -AMK152(talkcontribs) 22:23, 7 April 2009 (UTC)

I would not. Russia and Egypt are in two each, and Turkey is marginal. Impossible to standardise. — Robin Patterson (Talk) 05:04, 8 April 2009 (UTC)

"One size fits all"[]

I just came up with what I believe to be a great idea. Since we have information regarding a country in its infobox on the article, subdivisions and what it is a division of, categories, and we categorize the categories with chains of categories.

Why not apply the info pages principle to places as well? Just have one page to store a place's information, and automatically generate it wherever we need it?

Now, time to deal with the "border changes" issue: Examples:

  • Someone may have been born in what is now Chautauqua County in 1807, when it was Genesee County. In 1808, Chautauqua County did not actually exist, so how could one be born in it?
  • Someone born in Pennsylvania in 1753 was not actually born in the United States. Pennsylvania did not become part of the United States until it ratified the COnstitution in 1787.
  • Someone born on Manhattan Island in 1633 was born in New Netherland, not New York. The English never took New York until 1664.
  • Someone who was born in Massachusetts in 1633 was not born in the United States.

To solve this, ee can have successor and predessor fields to show the evolution of border changes. Between each border change, there will be a history. Example using New York:

Tribal homelands (? to 1614)
example: TRIBENAME/REGION, North America - we will put the history of the place from settlement to Dutch colonization.
-->
New Netherland (1614 to 1664 and 1673 to 1674)
example: New Netherland, Dutch Empire, North America - will will put the history of the place under Dutch rule.
-->
Province of New York (1664 to 1673 and 1673 to 1776)
example: New York, British Empire, North America - will will but the history of the place under British rule.
-->
It broke into:
New York, North America (1776 to 1788)
Republic of Vermont, North America (1776 to 1791)
example: New York, North America - we will put the history of New York during the Revolutionary years.
example: Republic of Vermont, North America - will will put the history of Vermont during the Revolutionary years.
-->
Then the ratification of the Constitution:
New York, United States, North America (1788 to present)
Vermont, United States, North America (1791 to present)
example: New York, United States, North America - history of New York since ratification of the Constitution.
example: Vermont, United States, North America - history of Vermont since ratification of the Constitution.

This would solve the "change of borders problem" which brings up the question "where did this happen?"

This solves the border change problem and allows us to divide up the history of each place.

-AMK152(talkcontribs) 22:47, 7 April 2009 (UTC)

I think this is an excellant idea. However, I suspect that it will not actually work. Some of my pages are already hitting the #ifexists limit, showing the dreaded Category:Pages with too many expensive parser function calls. Can we implement this without lots more #ifexists calls? Can we get the limit parameter increased for this wiki? Thurstan 23:22, 7 April 2009 (UTC)

I'm not talking about putting more #ifexist calls on person pages. I'm talking about place pages, which, as a far as I know, do not have #ifexist calls. -AMK152(talkcontribs) 23:44, 7 April 2009 (UTC)
Sorry, since you said "info pages principle", I was expecting that you were talking about automating the "Born in XXX" categories from "|Birth place = XXX". If you are not, then to me you are talking about a side-issue, and I have nothing to add. Thurstan 02:55, 8 April 2009 (UTC)

There are two issues here. First, can one come up with a generalised indicator of place that works everywhere and everywhen? The answer is yes: The place of birth etc should be given as longitude and latitude, and then the year determines the name of the country etc. Second, it should be usable by novices, who just want to upload their grandparents, only have a place name, and do not want to sort through elaborate instructions in what may be a foreign language. The solution, I guess, is that someone creates a giant lookup table that replaces Winterswijk with 51° 58′ 0″ N, 6° 43′ 0″ E, and another lookup table that places this in the County of Guelders in the 11th century, in the Duchy of Guelders in 15th century, and the Province of Gelderland in the 20th century. This can be done, but it is a lot of work. Rtol 05:47, 8 April 2009 (UTC)

Sorry. I was wrong. Instead of a giant lookup table, one could also work with links to individual rows in that table -- and this is essentially AMK152's idea of infopages for places. Rtol 05:57, 8 April 2009 (UTC)

Status of Info Pages: April 8/9, 2009[]

I have just spent the past 2 hours working on the Info Page for places. It has turned out better than I thought. Check out Chautauqua County, New York, United States and its info page.

The parameters I have put in so far:

  • Subdivisions from continents
  • Area information
  • Population/Density
  • Website
  • Predecessor/Successor (the historical change of bounderies)

What I plan to do:

  • Format Predecessor/Successor parameters so that it goes across the entire template
  • Add parameters for what that particular geographic location is divided into (similar to showinfo children for people pages) Example using United States: Child1 = Alabama, Child2 = Alaska Child3 = Arizona, etc. but I will replace Child with subdivision or something like that.
  • Create a category template to generate categories from the place's info page
  • Other parameters?

What other parameters should be included in the template? I don't know where everyone stands on what to include and what not to include. (Elevation? Government? Time zone?) -AMK152(talkcontribs) 05:12, 9 April 2009 (UTC)

Excellent start. The info page strikes me as complete. There will be duplication, though, as towns move from one country into the next. Do we want many pages for Strassbourg (in the Frankish Empire, in Lotharingen, in the Holy Roman Empire, in Burgundy, in Germany, in France, in Germany, and in France again ...)? Or do we say Strassbourg, France 1945-, Germany 1940-1945, France 1918-1945, Germany 1872-1918 etc?
The page itself should have more links. See e.g. Zwaag Rtol 05:43, 9 April 2009 (UTC)
Straasbourg's history will thus be divded between those pages. There will only be one page for each France and Germany. Just like my example using the Province of New York, which will have one page for being part of the Dutch Empire and one for the British Empire. (The Dutch took New York from the British briefly from 1673-1674). -AMK152(talkcontribs) 05:56, 9 April 2009 (UTC)
This will be messy for some places, but it would work. By the way, the Dutch RETOOK New York from the Brits from 1673-1674. Manhatten was claimed for the Dutch in 1609. New York was founded in 1614 (as New Amsterdam), and stayed in Dutch possession until 1664. Rtol 18:17, 9 April 2009 (UTC)

Subdivision parameters[]

Okay, I'm not sure what to name the subdivisions. Should it be:

|Country = 
|Countrysubdiv1 = 
|Countrysubdiv2 = 

or just

|Country = 
|Subdiv1 = 
|Subdiv2 = 

? -AMK152(talkcontribs) 02:36, 10 April 2009 (UTC)

I vote for the latter. Rtol 04:41, 10 April 2009 (UTC)
I'd shorten it to "subd1" etc if it was going to need typing now and then. In the context of place-names nobody would confuse that with subdeacon, subdominant, subduction, subdue, or subdural. — Robin Patterson (Talk) 01:44, 18 April 2009 (UTC)

Parameters - from a historical perspective[]

  • Okay, we need to be absolute certain of what the parameters are going to be before we go ahead and use the template. From a historical perspective, we need to determine what is necessary and what is not for the infobox. Here are my suggestions:
  • Subdivisions from Continent to country to subdiv etc. - Needed
  • Year of settlement/foundation/incorporate/conquered/etc. along with successor and predecessor information - Needed.
  • Population - changes too often. I suggest we just use the popualtion parameter only on places currently existing.
  • Long/Lat - Needed
  • Government - perhaps only for current? For historical places are we going to omit this or display the first and last leaders?
  • Elevation/Highest/Lowest point - not necessary
  • Nicknames, symbols (state bird, flower, etc.) - not necessary
  • Capital - Needed
  • Largest City - I would suggest using this
  • Flag, seal, and map - yes, if they exist.
  • Website - reserve for external links
  • Wikipedia link - link to Wikipedia in the infobox, just like we do with the surname infobox.
  • These are my suggestions. I am going to work on applying them. Further discussion and further implementation of what was discussed will bring us closer to a final product. -AMK152(talkcontribs) 22:41, 10 April 2009 (UTC)
I come at Wikia Genealogy with a slightly different opinion than others here. I think we should complement Wikipedia, rather than replace or duplicate it. So, to me the link to Wikipedia is essential, but stuff like government can be found there. So why burden this site with the need to keep it up to date? On the other hand, our place articles must have links to the people that live there, although most of that can be done at the page rather on the info page. Ditto for links to genealogical records. A link to Wikimedia would be welcome. Rtol 06:42, 11 April 2009 (UTC)
I'm in partial agreement with Rtol there. Genealogy doesn't depend on government. It isn't even particularly relevant to "capital" or to "largest city" (whatever that may mean in a particular area)) or to flag or seal. — Robin Patterson (Talk) 02:50, 18 April 2009 (UTC)
Nor do we want "Year of settlement/foundation/incorporate/..." except as part of describing the page as a time-division - and it might have conquering as the end year of that time-slice. A more likely end to a time-slice would be the year in which the next county or other division was carved off to leave a smaller unit. — Robin Patterson (Talk) 02:50, 18 April 2009 (UTC)
"# Subdivisions from Continent to country to subdiv etc." - yes, but it's not quite that simple. As I mentioned earlier, some (probably most) countries have more than one. Notably modern Belgium. So the parallel strands of division will have to work like marriages, where in Belgium "communities", "regions", and "language areas" would each have its own "spouse" and set of "children". — Robin Patterson (Talk) 02:50, 18 April 2009 (UTC)
Yes! I was thinking about that when I was doign Chautauqua County, which has cities, villages, and townships. Spouses. Never even thought of that. Great idea. -AMK152(talkcontribs) 02:57, 18 April 2009 (UTC)

Updated parameters list - April 17, 2009[]

So based on the discussion above, here is what we have:

  • Subdivisions from Continent to country to subdiv etc.
  • Year of settlement/foundation/incorporate/conquered/etc. along with successor and predecessor information
  • Population - changes too often. I suggest we just use the popualtion parameter only on places currently existing. - Should we include this at all?
  • Long/Lat - Should we include this? After all, it is a location.
  • Government/Elevation/Highest point/Lowest point/Nicknames, symbols (state bird, flower, etc.) - don't use
  • Capital - Could be useful as a historical perspective... capitals of ancient empires? I'm neutral on this one.
  • Largest City - kind of goes along with population
  • Flag, seal - questioned
  • Map - Needed
  • Website - reserve for external links
  • Wikipedia link - link to Wikipedia in the infobox, just like we do with the surname infobox.
  • Subdivision types of that place be broken like "spouses" and each subdivision a "child" of that "spouse."

-AMK152(talkcontribs) 03:02, 18 April 2009 (UTC)

Template info categories[]

I adjusted the sub-templates for birth, death and marriage place so that the annoying ", " does not show up for those that have a birth town but no birth state. Rtol 18:17, 9 April 2009 (UTC)

Not the world's ultimate classification system[]

We should be driven by actual usage scenarios, not theoretical ideas of most perfect classification scheme.

What I see currently is a need to correctly indicate location. A gedcom file says Saint Petersburg. (Russia or Florida?) With property Property:locality of country, and Property:locality of state, our tools can know this information and suggest one or the other, with confirmation from the user. -~ Phlox 21:51, 26 May 2009 (UTC)

Place name anachronisms[]

I am intrigued by AMK152's suggestions above to incorporate historical changes of political jurisdiction, so that our ancestors can be placed in their contemporary environment. For example, it would be nice to locate my ancestors from 1640's Woburn, Massachusetts in the Massachusetts Bay Colony, rather than in Middlesex County, Massachusetts, United States (none of which yet existed). At a smaller scale, I have another New England ancestor who lived near the edge of a township, and after his death the land was annexed into a neighboring town, which is still the modern geographic referent for his property. It would be nice to be able to present the political geography as our ancestors experienced it, as well as the modern directions to the site.

Unfortunately, the example given for linking places through time (Chautauqua County, New York, United States) is now seriously broken, relying as it does on apparently outdated info technology. Is anybody else wrestling with this issue? Are there reasons (other than the need to create additional place pages) to keep historical jurisdictions out of the SMW forms? One practical question might be how, within the current hierarchical scheme, to designate the Massachusetts Bay Colony as a subdivision of the United Kingdom yet locate it within the continent of North America. But are there deeper philosophical issues here? How are people writing about European ancestors dealing with the constant flux of national boundaries? Bruce Kendall 03:58, January 18, 2011 (UTC)


We have passed beyond the "info page" system that was being considered for most of the above discussion, up until mid-2009.
Our current standards (hinted at on Familypedia:Model_place_page but possibly needing revision there) follow Wikipedia in most cases, for ease of page creation and linking. Wikipedia's page names use current boundaries, but there are pages for historic boundaries (places like Sussex and Huntingdonshire and the Massachusetts Bay Colony), which we and Wikipedia link to.
It would be lovely if every historical document about an event said exactly where the event happened, but life's not like that (and even now with GPS being available, though subject to errors caused by land movement, we can't be totally precise). Sometimes all that is known from documents about a person's parent is that he or she was born in Russia before a certain date. There is no way that we can pin that down to a division or even a continent or a year. Even Bruce's ancestor who lived outside a town in an area that was later included could have had problems. If he or someone was known to have died there but the date wasn't known, how could you say where the person died using the "historically correct" place name, the "contemporary environment"?
This is where the "Other" box of our "place" subforms comes in. Where the current standard locations (nation, "subdiv1", county (or equivalent), and locality) are "known" with some confidence we use them (for good reasons, including the ability to use templates such as {{show county people}}, which you may see working at Devon). Where there's doubt, we can use the "Other" box and the "Notes" section, not amenable to computer analysis in any exact way but visible on a person's page.
If Familypedia finally documents every person who was (actually or possibly) born in the 18th century, we will still be unable (even with huge advances in computing power) to say with any confidence how many people were born in West Sussex in that century, but we will have a rough idea of "at least how many" actually were, and how many less well-defined people could have been. That's all we can hope for. — Robin Patterson (Talk) 14:23, January 23, 2011 (UTC)