When I am documenting records, do I use the name of a city or country as it is called today or as it was in the year an ancestor was there?
I believe it's good practice to use the place names that were in use at the time the event occurred, as geographic data changes: county boundaries can shift, counties can disappear, towns can get renamed, countries can get invaded, etc. Keeping the information on place and date together will make it easier to retrace steps later. If you would like, you can add a note saying that now (say when!) the town is in such-and-such county/province/state/whatever.
The historical name may be reused and become the current one again. The current name will be historical one day. A town may disappear, be merged with another, etcetera.
The right name to use is the name at the time of the event.
By the way, in regions where a town has two names, e.g. Belgian towns that have both a Dutch and French name, either name is correct, but it is best for your own and your reader's sanity to be consistent in your choices.
When you publish a genealogy, you may want to include a section on place names, perhaps illustrated with an old and new map, for readers unfamiliar with the place name history of the region of your research.
If I may add a touch of humour to the site, already, questions such as this remind me of the old Communist era joke:
A sociological survey:
“Where were you born?” St. Petersburg.
“Where did you go to school?” Petrograd.
“Where do you live now?” Leningrad.
“Where would you like to live?” St. Petersburg
In my opinion, the name of the place recorded should be name of the place at the time of the recorded event... otherwise it loses the historical context.
An alternative would be the city of Gdansk/Danzig - where the name reflects the nationality of the population (in theory, if not necessarily in practice)
I record the place (and address if available) as it was specified in the source which I have concluded is the most reliable for the event -- that it, the place at the time the event was recorded.
If there are multiple versions of the contemporaneous place name, I record the alternatives in a note associated with the event, with an explanation for the choice I've made as the "main" place and (if appropriate) a brief explanation of why variants exist.
I also record the present-day equivalent (with the date at which the note was made) and I've recently started to record as well the Latitude and Longitude.
My reasoning for all this is to enable others viewing my research to see how I've derived the conclusions that I've recorded from the available sources AND to locate the place unambiguously on modern maps.
The cheat's answer to your question is BOTH. For different purposes, both the former and the current names will have value. At the heart of the issue is how should you record these alternative descriptions.
Does your software allow you to assign a latitude and longitude to a place? I am able to enter as text the name of the place as it was at the time of the event and rely upon the associated lat and long to find the current name through any one of several on-line mapping tools.
That means that if your ancestor stayed in one place while the national borders shifted around her, several place entries indicate which government entity might hold records made at different times but all will map to one location on earth.
Addition to explain management of risk of unwarranted specificity.
If I have evidence for a precise location such as the church where a christening occurred or the street address listed in a census, then that is the lat&long data recorded. Where all I can justify is the name of a town or region, then I enter the coordinates for an administrative centre (such as a Town Hall) with at least a conceptual connection to the records but the label of the general location.
When better evidence becomes available I can transfer the link from an existing event to a more precisely defined place. The general place descriptor remains in the data file for future use even if there is no event attached to it at present.
I record all of the place names available, in addition to coordinates when I can:
- I use a system (Gramps) that allows me to maintain a single record for each "Place", where I can record copious amounts of information about that place. Each record that refers to a place does not contain all of the place information, just a reference to the Place. Where appropriate I make a note in the referring record so that historical context can be captured.
- Recording all of the place names allows me to see that several generations of a family have lived in the same location even though the town/county/district has been merged/renamed around them.
- Using coordinates permits an unambiguous, absolute reference to a single location on Earth -- and simplifies the preparation of a map. (To respond to a comment on another answer by Gene Golovchinsky: I try to avoid over-specificity in coordinates so as not to pinpoint a precise location. Specificity is usually obvious -- a reference to a city is different than a reference to a specific street, or even address in a given city: Anytown, USA vs Main St, Anytown, USA vs 123 Main St, Anytown, USA.)
For example, let's say Marie St. Pierre was born in 1830 (a fictional, if plausible, name/date) in Berthier-en-Haut, Québec. Her grandparents might have been born in la Seigneurie de Berthier -- which was the same place. Her living descendants might be living in Berthierville -- which is the same place. In my records, I have a reference to a single Place (Berthierville), which has the alternative place names listed, as well as coordinates 46°05′N 73°10′W. Marie's record would contain a reference to Berthierville with a note saying that the name at the time was Berthier-en-Haut. In a 1980s record, I would have a pointer to Berthierville with a note containing the city & county -- both pieces of information being important because there have mergers and changes to administrative districts.
As another example, I have ancestors who lived in Leuven, Belgium, which is spelled Louvain in French. I have both names recorded so that if I come across a new record that refers to Louvain instead of Leuven I will know that they are the same place.
Places (just as with people) may have alternative names. This could be due to a total rename at some point, alternative spellings, colloquialisms & informal names, or names in different languages. These "accepted variations" should be part of the Place entity recorded in your data. That way, there is only one definition and history for each place in your data.
However, records will certainly have "unaccepted variations", most commonly where a name was misspelled or mistranscribed. Another variation in when erroneous data was deliberately given at the time. This is part of the evidence associated with a cited source.
The point I want to make is that the historical name(s), the hierarchical association of the Place (e.g. which district or county is it in?), and its general history, should all be associated with the specific Place entity in the data. Other data can then make a reference to that entity. The name as it was written in a record may be recorded verbatim if there is something unusual or ambiguous about it but this is best judged on a case-by-case basis. The majority of my places are references to associated Place entities, and I only record the specific text if it is questionable in any way.
The decision about what name to use will depend on how you intend using that name. So if you intend to run a feed out of your database and into some mapping software (and your software only allows one name), then as said above, contemporary names like "X, Y, Austro_Hungarian Empire" won't be of much use to you and you'll (probably) need to use current names. Depends on the software and database, obviously.
To me, a place-name needs to satisfy 3 criteria:
- It works in the software;
- It defines the place accurately (how accurately is another matter) to any plausible reader, given the context;
- It reads well.
So using a name contemporary with the event can really make a story read well because the name is so different that it stands out and you know you're "not in Kansas anymore." And it can be said to be more accurate as it's closer to the source-text.
But if the contemporary name is meaningless (or ambiguous) to the reader, then it fails the 2nd purpose if unqualified by any explanation. So for instance (quickly consulting Wikipedia), if I were to designate a place as "Springfield, Missouri Territory, USA" that would probably give the wrong impression to anyone from my side of the Atlantic, who would equate Missouri Territory with the state of Missouri. So if you use such a contemporary name, then you need to put in somewhere in your text as presented to the reader that (a) you've used contemporary names and (b) for Missouri Territory see this map that you've taken from Wikimedia Commons (and credited, of course).
Bear in mind with the first bullet ("It works in the software"), that most software will see "Widnes, Cheshire, England" as a different place from "Widnes, Lancashire, England". Which they're not.
Oh - and if you're writing for a Journal, they may have their own standards.
So fundamentally, you need to examine those three bullets and make your own decision - but be consistent about it!
The problem with using "current" names as the only name, is that "current" changes over time. Is the Crimea part of the Ukraine or of Russia. If you attempt to use "current" you're dooming yourself to a never ending job of updating locations.
I prefer an approach, supported by the tool I use, Genbox, that uses the name at the time of the event which links to a database location where all the names are indicated.
born L'vov, (...) Russia Bar Mitzva Lemberg, (...) Austria Married L'vov, (...) Soviet Union Died L'viv, (...) Ukraine
[Where the (...) indicate the equivalent of town/city, township, county, province based on the location on the date in question]
I have the option to add the current name also, which makes the events look like
born L'vov, (...) Russia (Lately, L'viv, (...) Ukraine) Bar Mitzva Lemberg, (...) Austria (Lately, L'viv, (...) Ukraine) Married L'vov, (...) USSR (Lately, L'viv, (...) Ukraine) Died L'viv, (...) Ukraine
This allows the information to be correct and not require an update with Russia conquers the Ukraine.
I use modern names solely because it is a limitation of most genealogical software, which does not allow the indexing of one place name to another place name. I am also mindful of the trend in genealogical software to start to incorporate mapping integration with online map providers, and Google Maps has no idea where "Nadworna, Galicia, Austro-Hungarian Empire" is, much less an ethnic designation like the Jewish Kehilla [community] district "Nadworna, Nadworna Kehilla, Austro-Hungarian Empire".
So, in my software, I record towns with their modern names, such as "Nadvirna, Ivano-Frankivsk oblast, Ukraine". I only add the raion (county) name in there too if it is an exceptionally small location or if there is more than one town with that name in the same oblast (or voivodeship, for Poland).
Personally I use today's name but also place a note with the old name.
We're looking for long answers that provide some explanation and context. Don't just give a one-line answer; explain why your answer is right, ideally with citations. Answers that don't include explanations may be removed.
Let me make a suggestion which reflects my experience in military reporting -- There, when mentioning a location of an event, the military report includes both the name of the community as it is known, and the map coordinates for the place. The name of the country or city might change, but the coordinates of the town usually won't (unless it was destroyed and relocated ). You can find the coordinates on Google Maps or Wikipedia. Often you'll find alternative names of the community as well on these sites. Certainly, if you are doing a book, this would make great sense. But I think it is too much work if you already have a substantial family tree; I can't see going back through actual tree and figuring out the coordinates for everyone. My tree has 13,400 names; it's not going to happen.