Are (near)duplicate names for different English counties a risk?
The county of "Hampshire, England" (Place Id 10604925) has a string of names, some of which I've never seen but never mind. One that concerns me is the first name:
Hāmtūnscīr, ang, Full Name
There's also:
Hamtunscīr, Latin (la), Full Name
The county of "Northamptonshire, England" (Place Id 10600728) has a string of names, the first one of which is:
Hāmtūnescīr, ang, Full Name
There's also:
Hamtunscire, xno, Variant Name
The 4 names are different - there's an extra "e" in the middle of the 3rd name, and similarly at the end of the 4th name. Also, the various accents (diacritics?) tend to be different.
Given that there has to be a degree of flexibility in the matching algorithms for placenames (otherwise "York" would never match "york", for instance), does the closeness of these names for Hampshire and Northamptonshire pose a risk to matching in FS place-matching algorithms?
Note that the (near) duplication of the names does have an origin in fact - according to the Wikipedia entry https://en.wikipedia.org/wiki/Northamptonshire that county was first recorded as "Hamtunscire" and the "North" was sensibly added later to distinguish it from the similar name for what became Hampshire.
If the near identity of the 2 names for different counties does pose a problem, I would advocate removing the names "Hāmtūnescīr" and "Hamtunscire" from the record for Northamptonshire. Yes, they are strictly correct but I would suggest that they are genealogically irrelevant - they are Anglo-Saxon in origin (that's what the "scir" bit is - it's now "shire" in English) and they will be seen, I believe, in the Anglo-Saxon Chronicle, the Domesday Book, early medieval charters, etc. Not likely to be found in FamilySearch - certainly not in the original Anglo-Saxon…
Comments
-
When I just tested entering "Hāmtūnscīr" as a placename, I got offered "Hampshire, England, United Kingdom" as first in the list and "Northamptonshire, England, United Kingdom" as 3rd in the list.
Given that I've never been able to convince myself that the placename auto-standardisation 'bot always takes the first in the list, this does suggest a risk to me.
1 -
The goal of the alternate names is for when a person searches based on what the records may list the place as. We prefer to have the other spellings, even if they are wrong.
0 -
@russellwaltgeorge1 - thanks for the explanation. I have sympathy with the "even if they are wrong" angle, having once been trapped into wondering why "Middlesborough" wasn't in the list of standard places in England. The clue is that (at that time) it was in - but only under its correct spelling of Middlesbrough. A red face from me when I finally twigged I was mis-spelling the name.
I asked for the spelling mistaken version to be entered as an alternative name, which appears to have been done.
The problem that I worry about is when the background auto-standardisation 'bot is running. If it has to deal with "Hāmtūnscīr" as a placename, will it always pick "Hampshire" or might it pick "Northamptonshire"? If it's always the first, fine, but if it's unpredictable, I worry this might be a risk.
Two things (1) I don't have a solution to mind and (2) that particular placename is Anglo-Saxon so is unlikely to appear. But I do worry that other similar names might be lurking that I haven't thought of… Maybe we shouldn't create them without checking for potential confusions?
0 -
Good point. However, we tend to add various spellings as alternate names, to help others find the place. Yes, this makes it a little more challenging.
0