standardized Swedish place names
Could you please standardize the parishes below. There are cities of the same name that will come up, but I need the larger parish juristiction for church records.
Foss Parish, Göteborg och Bohus, Sweden [see MP55-7GS]
Svarteborg Parish, Göteborg och Bohus, Sweden [see LDQT-J45]
의견
-
Both Parishes do have a current place ID. They can be found at:
Foss Parish: https://www.familysearch.org/research/places/?focusedId=617972
Svarteborg Parish: https://www.familysearch.org/research/places/?focusedId=11834320
0 -
Those locations don't have the word PARISH in them. The problem with just using the parish name without the word 'parish' is that many people mix up the name of the town, like Foss, with the next juristiction up 'Foss Parish'. The same with Svarteborg and with Bro and Håby. All have cities of the same name in the parish of the same name. This causes problems and I would like to be able to select the name with Parish in it to be clearer.
Indexers made a mess of this when indexing the records for Bro PARISH, all of the records have been placed under Bro, Munkendal. Bro being a city in the parish of Munkendal. When these records are connected as a source they all indicate the wrong parish of Munkendal. I know to change the location when I link in the sources, but others don't know that they are all wrong.
1 -
"Those locations don't have the word PARISH in them. The problem with just using the parish name without the word 'parish' is that many people mix up the name of the town, like Foss, with the next jurisdiction up 'Foss Parish'."
I totally agree but, being pessimistic about it, at first glance, FS seems to have set its face against having the type as part of the name. I can't actually find any clear statement on this, nor can I find any justification.
Oddly enough, for further reasons that I don't understand, the prejudice against typed names does appear to have been broken for (at least) Germany. See https://www.familysearch.org/en/fieldops/fs-places-familysearch-places-country-specific-guidelines#europe where it says
Especially for German administrative districts (Kreise), we have been asked to include type words in display names. This applies at the district level and only to German display names. Display Names for English should simply be the noun form of the city, for which the administrative unit is named. For example, the district of Wesel in Nordrhein-Westfalen will have a German display name of Kreis Wesel and an English display name of Wesel.
So they'll do it for Germany, but only for the names in German. No idea why...
But there is perhaps a further reason to be optimistic - some of my local names in England (sensibly) cover churches, hospitals, etc. The ones that stick out for me do include type-words in the name - e.g. "Holy Trinity Church, Blacon ..." or "Countess of Chester Hospital, Chester, ..."
So come on, guys, if you can do it for some, why not all?
0 -
Hi. I'm on the Authorities Team, and I can speak to this.
Our official guideline for this is listed here, at the bottom of the page: https://www.familysearch.org/en/fieldops/familysearch-places-principles-guidelines#place-names
Place types in display names:
Guideline: For higher levels of jurisdiction or administrative units, full names of places often include a place's type and more description (e.g. Davis County, Commonwealth of Virginia, City of Chicago), while Short names represent the place as it is normally called (Davis, Virginia, Chicago). Display names should not include the place type unless it is commonly used to refer to the place.
Explanation: Display names should describe a place the way people would normally refer to it. Most cultures do not normally include a designation of type (city, county, municipality, island, etc.) when referring to a place.
Exceptions: Some cities do, in fact, include the word "city" in their normal name; for example, Salt Lake City and New York City, but not St. Louis City or Las Vegas City. At present, many townships in the US include the designation "township" in their display name. Display names of cemeteries, churches, schools and hospitals generally include type designators. In addition, certain cultures (i.e. Irish, Chinese, Korean and Japanese) commonly include type designators when referring to places. These should be detailed in country-specific guidelines for editors.
The main purpose of this measure is to reduce bulk and repetition. When display names get too long, the systems that use those display names end up with errors.
The exceptions we have made (such as the German names in Germany that you identified) are specifically requested by record custodians such as archives or by other FamilySearch teams in-country. We do it on a pretty limited basis.
We do recognize the challenge of having multiple places with the same name. We are trying to solve that problem with record indexing by having the place type show up clearly in the user interface and/or limiting the types of places that can be chosen for a given record collection, but the wheels of engineering do turn slowly. Having data on those kind of errors (e.g. how often records are mis-indexed) can help us push for those changes or review policies in a given region.
Let me know if you have any other questions.
1 -
@amandamills1 - thanks for responding to us. The link about Principles & Guidelines is not one I'd seen before so I've bookmarked that.
I must applaud the idea of "having the place type show up clearly in the user interface". From my own personal knowledge of UK placenames, I'm instinctively dubious about "limiting the types of places that can be chosen for a given record collection". I'm not sure if real life is sufficiently disciplined to make it feasible. I know I've seen birthplaces in English censuses where I don't know whether someone was born in XXXX town or XXXX parish because all it says on the original is XXXX. But see what the designers can come up with.
In the meantime, I can perhaps reference your principles and refer to "multiple places with the same name" when I run into an issue created by that sort of problem. For example, one of my ancestral towns is Nantwich, Cheshire, England. The standard place list has 4 types of place called Nantwich but in all honesty, I don't think that the Poor Law Union and the Hundred of those names is going to cause us any problems as I don't think they'll be referenced very often. As for the other two - we'll see.
Again, thanks for responding Amanda.
0 -
Yes, that pattern was created in the UK more than a decade ago -- I believe to support a British maps project, though I'm not sure. It does mean that some places have 4 entries, which isn't ideal. Places supports more than 30 services, so their goals sometimes conflict.
I do know a team currently working on updating the search algorithm, though, so I will make a note to them about this issue and see if they can lower the priority of place types like Poor Law Union so the town shows up first. I know they are also working on some filtering functions, so I'm hoping we'll be able to see those functions get over to the person page and tree views eventually.
1 -
@amandamills1 - that was probably the England Jurisdictions 1851 set of maps. It layers (optionally) parishes, civil registration districts, etc, on current maps.
Interesting idea about lowering the search priority of place types like Poor Law Unions. That seems a good idea because if we're looking at collections of data about people, I find it difficult to believe that they would have data encoded with a poor law union or a hundred without having a more specific data item such as settlement or parish. Even someone in a workhouse (which suggests the appearance of a possible poor law union) will probably have the workhouse named by the town that it's in (as well as the poor law union).
0