Thursday, January 28, 2016

Standardized Historical and Current Place Names - Post 3: An Elegant Solution

I wrote Standardized Historical and Current Place Names - Post 1 (26 January 2016) and Standardized Historical and Current Place Names - Post 2: Two Potential RootsMagic Solutions (27 January 2016) recently, defining the quandary that many researchers have that use genealogy software or online family trees who want to include historical place names in their family trees, and then offering two options for users of RootsMagic.

1)  For researchers like myself, with 28 years of research and using desktop family tree software, and having over 45,000 persons and over 8,800 place names in my family tree database, to use either of the options outlined in Post 2 is "too much work" at this point in time.  So I will leave the task to someone else in another time.  This is something that I should have done a long time ago, and did not do.  It's my fault that I didn't learn about the historical place name issue earlier and use it in my family tree data entry.

2)  A proposal:  To me, the only rational way to implement a method to incorporate historical place names is by adding a feature to the different software programs that consider the date of an Event, the historical or current place name, and then adjust the place name in the data entry fields depending on the time frame.  

Let me use the current place name of the town of Medfield, Massachusetts as an example.  I created a table, based on information from the FamilySearch Place Research page, for the town of Medfield:

The table above reflects the jurisdictions at the town, county, state and country level from 1635 to the present.  The latitude and longitude of the place name could be added to the table.

If an event (e.g., birth, marriage, death, probate, deed, etc.) happened in Medfield in, say 1685, then the program would enter Medfield, Suffolk, Massachusetts Bay Colony, British Colonial America into the place name for the event.  If I want to see a map of the location, the map of the standardized, current,  location would appear.

Likewise, if an event happened in Medfield in, say 1799, then the program would enter Medfield, Norfolk, Massachusetts, United States into the place name for the event.  Again, the map would show the standardized, current, place name.

3)  The Challenge:  The above is fairly simplistic with respect to Medfield.  A similar table would have to be coded for each town, each county, each state or province, and each country and put into the family tree program.  

There is also the problem of even smaller places than towns, like hamlets and villages, and mid-size localities between towns and counties, like boroughs and townships, where the table needs to account for them.

Then there is the problem of how the historical and current place names are transferred in a GEDCOM export to another program or online family tree.  

Perhaps the FamilySearch Place Research information could be accessed by the family tree program, or online tree site, to offer the user a list to choose from.  Ideally, it would select from the list automatically based on the date of the event.  

If it was implemented in some fashion by desktop and online family trees, a neat "Option" in the program or tree would be to "Select historical place names for all Events" so that the user doesn't have to do them one place name at a time.

4)  I know all of this is idealistic, and pretty "elegant," and i'm sure there are reasons why this can't be done.  But maybe it can be accomplished.  If so, it would be heralded by all researchers as a tool to make our research more accurate and to cover the shame of past sins in entering place names 

"Forgive me, Saint Jacobus, for I have sinned."


Copyright (c) 2016, Randall J. Seaver

Please comment on this post on the website by clicking the URL above and then the "Comments" link at the bottom of each post.  Share it on Twitter, Facebook, Google+ or Pinterest using the icons below.  Or contact me by email at


  1. GRAMPS software does this exactly! I have many British places that have moved counties over the years and they appear under both depending on the date of the event. It's one of the reasons I moved to GRAMPS when FTM was discontinued.

  2. I agree. But gets even more convoluted if trying to get more refined than a town. If you use a 5 position place name so you can include an address or institution name, for example. Street names and numbers change; or disappear all together. Institutions move (hospital where I was born). But these issues happen at all levels. Even gravesites get moved. The town some relatives were born in changed name and country several times in the past 100 years. And the center of town or the country capital changes. So what do you put for a lat/long? Think Jerusalem or Bethlehem. To me, the closest idea is to record a lat/long as accurate as you can. Then use other software to interpret what that means at different points in time. Put geo-political names in the notes.

  3. This comment has been removed by a blog administrator.

  4. I'm glad that you're pursuing this, Randy. It needs attention.

    What shall we do with place names that no longer exist in zip code lists? Some mining towns that once were vibrant communities have dwindled away after a couple of generations, lost their post offices, lost their schools, lost their incorporation status with the state. I understand that the same thing is happening with small farming towns in the Midwest that are losing their populations as corporate farming takes over.

    But they did exist in a particular place that their descendants should be able to learn about, at least in genealogical records, regardless of their current status. Archives have records that originated with those place names on them, and the US Census is one of them. Descendants need to be able to use those names to find those records.

  5. I'm trying out various software programs to replace ancestry. I must look at GRAMPS again. I don't remember why I didn't move it closer to the top of my "like" list. There must have been a reason. But the place name issue is very important to me and I wish it was a more universal option. I know a lot more geography than I ever did before but I can't possibly remember even half the changes through time.

  6. This is one of the features Gramps has and I love it! I've used RootsMagic for years but I find myself using Gramps more lately.
    Rootstrust is another that works with place changes. It's a bit different in that you link places together.

  7. This series has focused primarily on the problems with place names at the big end (e.g. US versus British America) of the place name. However, the tougher problems are at the little end. Many New England towns, for example, adjusted their boundaries with their neighbors as they developed. So a location in one town in the 17th century might be in a different town in the 18th century. There are many of examples. Of course, looking back on it in the 21st century, you're unlikely to know whether your ancestor's home was in the area that was "adjusted" or not. There's no easy way to solve this problem, not even with an automated historical gazetteer. To solve it automatically would require you to enter the *exact* location, not just the town. So, IMHO the little end of place names must be filled in using the place name at the time of the record, not some modern equivalent.
    This problem is a geographical analogue to the problems of old style/new style dates. We've all seen examples of date creep because some well-meaning genealogist converted an old style date to a new style date, and then some later person converted again. That's why most folks prefer to use the slash notation to preserve both.
    For me, the only sensible approach to dates and places (the small end, at least) is to write them exactly as presented in the original record. Of course, it would be nice if genealogical software provided the possibility of including both dates and place as they are known today.

  8. One of the issues not considered is the existence of concurrent places. Thus (and I am using UK examples) we have the religious Parish and the statutory Parish covering simultaneous but different overlapping areas, often with the same name as each other and of one of their constituent villages. There is the various areas of the Registry Districts which changed over time and did not coincide with political boundaries. And there are the Poor Law Unions and the Judicial circuits, etc.

  9. Legacy Family Tree has a partial reality check for locations on events or people.When you enter a county name and state combination, it tells you if the county did not exist for that year. If you enter a county & state for an event with no year, I think it may not check when you add a year later. Cannot recall, but if the county never existed, it will tell you. Now if they could just do that for areas that bop between counties or states!

    There is also a location db manager that helped me clean up inconsistent entries from my 35 years of not so consistent entry. The feature is available in the free version that I used for years.You can sort the locations you have by state or county or city, merge places that are the same (carefully not merging two that are the same place in different counties because of different years!) It corrects all uses of a location to your standardized entry.

    I believe the county check only works in the US and the database does assume that every country uses city, county, state, country. There's no parish per se. You can put in multiple levels separated by commas, but they will be assigned to the US style defaults when you sort locations. Sweden gets messy, as i didnt originally know how many jurisdictions there are. You can geocode locations.

    Thanks for the article & comments. Didn't know about the Gramps feature - will have to explore that one.

  10. Your idea would also be great for those of us researching Central European countries whose boundaries flexed with every new king or conquest. Q: Did my immigrant ancestor come from Poland or Prussa? A: Yes.