Standardized dates and places as they come over from the Source Linker.
This feature is fantastic and I love it!!! Thanks for the years of making it possible.
Adrian Bruce said: I do find it curious when the places don't match the standardised place-names. Shouldn't FS be, what's the phrase, "eating its own dog-food"? (Though if the indexes are old, standard names can change between indexing and now).0
Tom Huber said: First, the source linker brings over the transcribed place, which may (and often does) contain errors. In other words, what was indexed is what is imported by the source linker. That may or may not be equivalent to a standard date or place.0
Elaine Vollmer said: Thanks for the reply. It does make sense why this happens and I'll be more patient with it as I change them from now on.0
Adrian Bruce said: "the source linker brings over the transcribed place, which may (and often does) contain errors."
Agreed Tom. But that brings in a philosophical problem of: Should FS be indexing places (indexing, not transcribing) knowing that the indexed name is, for want of a better term, incorrect? For instance:
England Births and Christenings, 1538-1975
Name: Samuel Pickstock
Residence Place: Longton, Stafford, England
Christening Date: 23 Sep 1888
Christening Place: St. James, Longton, Stafford, England
A simple use of the Standard Places tool (now, but perhaps not at the time of indexing to be fair), shows no place "Longton, Stafford, England" but does show "Longton, Staffordshire, England". Looking at the original page in FindMyPast, I believe that the "error" has crept in because the heading on the page reads "County of Stafford", which is simply a more formal method of naming "Staffordshire".
Now, I'm not pretending that is particularly obvious to a novice indexer from outside the UK, but if other records correctly have "Cheshire" indexing "County of Chester" then why not Staffordshire on this? Using the Standard Place-names tool as a check would have highlighted the issue. But, as I say, a counter-argument might be that this was indexed before the tool was a/v.
It would be interesting to know if cross checks for standard place-names are done in the indexing process.0
David Roderick McLean said: When I index using the web indexing tool it never has a standard place choice come up. I just transcribe what I see in the record.0
gasmodels said: I believe one of the other issues is when was the index created. Some of the older extraction records were created long before "standardization' was a gleam in someones eye. Much has changed over the years as technology has introduced new ideas and ways of doing things. So for example we see records indicating York which would now be standardized to Yorkshire. Another example is the Isle of Man records where the records show, - Isle of Man, England when it should simply be Isle of Man. I do not see anyone going back and updating old records to standard locations - I think the users have to live with updating information. It is no more time consuming that merging duplicate records which many of us have had to do.0
Tom Huber said: While this has drifted off the initial question, to the standardized list and its availability, what is now https://www.familysearch.org/research... started as a BYU lab tool (which I used) at labs.familysearch.org. This was more of a beta site, but it existed from at least 2009, if not before.
The British Isles has always presented a dilemma, sometimes using the place "name" Great Britain and sometimes not, even on that early site.
I am not sure, but I believe that the early familysearch.org indexing tools made use of the labs lookup. I could be wrong, since that has (as far as the familysearch site is concerned) been a long time ago and I remember doing some indexing. However, I do not recall details of the site, only that we could not correct indices, once they were published.0
Tom Huber said: The extraction programs existed prior to familysearch.org, so those records did not have the labs tool available.0
Adrian Bruce said: " I do not see anyone going back and updating old records to standard locations" - well, given that indexes can't be updated in any fashion, we're certainly not going to see the Historical Records being altered.
But I wonder whether there's any joined up thinking in current indexing projects?0
Adrian Bruce said: "When I index using the web indexing tool it never has a standard place choice come up. I just transcribe what I see in the record"
That's interesting David, thanks.
I'm actually torn two ways here - "transcribe what you see" must surely be sacrosanct. But I'm not talking about transcripts, I'm talking about indexes, and the index is actually a concoction, not a transcript. If I take the example above of "Longton, Stafford, England", well, I've seen the original page, so I can say that "England" isn't written down anywhere on the page - in other words, the index is (partly) a concoction. So I suggest that checking the concocted index against the standard place-names would be a useful quality check and could have highlighted that "Longton, Stafford, England" doesn't exist and perhaps leading to the suggestion that "Longton, Staffordshire, England" would be better. After all, lots of other parish registers are indexed as "Staffordshire", even though they will be "County of Stafford" on the page.
On the other hand, if we look at UK census birth-places, those are notorious for recording all sorts of oddities - I really can't expect anyone outside the county to realise that "Widdenbury, Cheshire, England", "Winbury, Cheshire, England" and "Wimbury, Cheshire, England" are actually all "Wybunbury, Cheshire, England".
Hm. It's not easy, I confess....0
Adrian Bruce said: Thanks Tom - so unfair of me to expect that old values match standard place-names.0
Tom, the point that Elaine Vollmer is trying to make is that the Source Linker gives us an opportunity to correct the place in the source to the standard place, but the corrected place is then ignored. Our work is wasted, because our choice is not carried over into the record. It's like FamilySearch is saying, "The source linker is a great place to practice choosing standard places, but we don't trust your first choice and will make you repeat your choice later."
This happens with all places in all sources. I've gotten in the habit, when I've exhausted all the sources for a person, of going back through the record and re-doing all the non-standard places in the record that were brought in via the source linker, but it is a disheartening duplication of effort.
I suspect that because of this DESIGN FAILURE in the source linker, millions of non-standard places are being needlessly added to FamilySearch records every day.1
I'm still correcting dozens of non-standard places every day that I enter myself because of this bug.
I can't understand why it's so difficult to put some priority on fixing this bug, especially given that it works at cross purposes to what FamilySearch is asking participants to do. On the one hand, FamilySearch asks people to standardize places, and even has a program to enlist people do exclusively that, and yet, every day, the Source Linker introduces THOUSANDS of non-standard places into FamilySearch!0