Why have the indexed versions of England & Wales Marriage Registrations been replaced?
Why has the original collection been replaced by an inferior one, I wonder?
From the record shown below one would think Oliver Shaw definitely married Elizabeth Margaret Brady, but the confusion appears to relate to a new version of these records acquired from Find My Past in March 2024. Other records prove that he married an Emma Siggins. She, along with a William Frederick Kunze, are all indexed under the same reference, as can be seen on the FreeBMD website.
It appears the collection FamilySearch is now choosing to use indexes just (any) two names (instead of multiple names) indexed against the same reference. Elizabeth Margaret Brady actually married William Frederick Kunze - the other missing name from the new version on the E&W Marriage Registrations collection.
Please revert to the previous version, which showed all names indexed under the same reference, instead of this one, which appears to always to display (any) two names. Sometimes I have found the two names to even represent individuals of the **** - so the selection of the "Possible Spouse" now appears to be a random one (from the original list).
Opening the Collection Information detail shows the new, unhelpful version appears to have been introduced on 6 March 2024:
Okay, FamilySearch will probably place the onus on Find My Past for providing this incorrect information, but I just just discovered their own version shows the record perfectly well:
As suggested previously (concerning census records displayed courtesy of FMP), if the versions that appear on FamilySearch are not down to processing errors at the FamilySearch end, there really needs to be greater liaison with Find My Past so the details are not distorted once they appear on the FS website.
Answers
-
Just to illustrate another example, showing that the current version of the collection is not really fit for purpose. The event took place somewhat before **** marriage was legalised here!
0 -
I can't figure out what word it is that the nannybot is bleeping out from your posts, @Paul W.
And that is a truly messed-up index. As you said, it is not fit for purpose: you cannot find any actual records using it.
1 -
It was a case of now-you-see-it, now-you-don't and now-you-see-it (again) - which is how it appeared when I last viewed it: the uncensored version, that is!
The missing word / phrase is "same-s*x" (marriage). I know a workaround is to put "same gender", but not the same thing: legally, or otherwise.
3 -
OK, in what universe is it considered to be vulgar or anything else bleep-worthily objectionable to point out that the sexes of the person and of the "possible spouse" are the same?
Trying to find a silver lining: can we capitalize on the Community software's objections to get the mangled index fixed on the main site?
2 -
I reported this behaviour in a different question without the knowledge that it was already recognized here. However, I noticed that the full content of all possible spouses was correctly presented when the source was viewed when attached to a person. The problem described above appears to be in the code that creates the presentation of the search results record which is presenting only the first, alphabetically, of the possible spouses and chopping off the other possible spouses.
A comparison of the two blocks of code should be enough for the engineers to isolate the cause.
1 -
However, I noticed that the full content of all possible spouses was correctly presented when the source was viewed when attached to a person
Unfortunately, this does not always appear to be the case, as illustrated in this example, relating to a search on Patrick Wright:
FreeBMD version of indexed record, bearing out the top name is always being selected as the possible spouse:
Here is the FamilySearch version of the record:
I have temporarily attached this to an ID (not matching - I will remove this later), but (upon opening the source) the Possible Spouse still just shows the name of William Bird: no alternatives, as found in FreeBMD, and used to be the case here:
Hopefully, this issue is being investigated by the engineers. As with other issues that still remain unresolved (e.g., disappearing middle names from records, and the removal of the prime individual's name in source titles), this matter is not helping us at all in our efforts to confirm identities and relationships.
3 -
I've not been impressed by the accuracy of records at findmypast, their error rate is significant otherwise I'd not have noticed it. The records at freebmd.org.uk are double keyed so errors are recognised but sometimes are not correctable because the original indexes are unclear. Even the records at gro.gov.uk have a significant error rate. I find it so ridiculously wasteful that these records seem to have been transcribed three times.
0 -
Whilst I agree with you that the other websites have plenty errors and omissions, too (I make regular reports to gro.gov.uk about their birth / death records) these are usually corrected fairly promptly, in my experience.
I would prefer the focus here to be on FamilySearch reversing whatever action their engineers took that is now making tracing correct marriage partners a far more difficult task.
In circumstances like this, it transpires to be a good thing that these records have been indexed / transcribed multiple times. Without the (admittedly far from perfect) version at FreeBMD researchers would be totally baffled by the suggestion there were apparently a huge amount of same-s*x marriages in England & Wales, far before this became a legal possibility.
1 -
Not just that these errors make the marriage appear to be same se*. It also make it appear that one or more persons are polygamous.
0 -
I've not encountered the problems with spouse names, if that's a new issue and affecting multiple users then I agree, but familysearch doesn't seem to like reverting after they've made changes, I find the new logon procedure unworkable as do many other users. I've also never forgiven them for removing access to the Scottish records. In the end we can't really complain when there's no charge.
0 -
@Re Searching said
"... It also make it appear that one or more persons are polygamous."
Not really. Where there is a list it should say "Possible Spouses" - and the FamilySearch screenshots do include that word "Possible".
The FreeBMD list is just a list of names on a page of the marriage register - 8 names for the four marriages on that page.
2 -
It's never been possible to determine the spouse from the marriage indexes alone. Even if there is just one man and one woman on the page, they are still possibles, there could be another couple omitted/mistranscribed. The relationship can sometimes be positively determined from census records or from the mother's maiden names of the children, but unfortunately paying for a copy of the certificate is the surest way.
It is not possible for familysearch to detect or reject male-male or female-female marriage errors, the s*x is not recorded in the index.
0 -
Understood. However, I was making the observation about the current state of affairs in which the output of search results for marriages presents only a single possible spouse. This leads to separate results for individuals that imply that they all married the same person, i.e. the first one to appear in alphabetical order in the list of all possible spouses. Subject to the above clarification by @Paul W.
1 -
I've had a look at an interesting example, Mary Eliza Tinckham's marriage in Kensington in 1888: https://www.familysearch.org/ark:/61903/1:1:2DDB-NL3
No 'others' visible:
But if you look at the underlying data for Mary Eliza's persona, visible here: https://www.familysearch.org/platform/records/personas/2DDB-NL3, these three 'Others' pop up (summarised):
Meanwhile freebmd (which again gets the gold star here) has:
while FMP has
It's pretty clear that there are 2 problems here:
a)
FMP has failed to index Mr Alloway correctly, putting him on page 236, so FS doesn't pick him up as an 'Other' for Mary Eliza, instead dutifully showing him as on page 236:
b) FS has a user interface problem in that the 'Others', to the extent FMP has provided them, are missing completely from the normal Record display.
I do suspect that the FS problems reported on this thread are largely if not entirely user interface issues only.
The citation for Mary Eliza on FS is dated 2014, by the way, which again makes me think the data is fine:
"England and Wales Marriage Registration Index, 1837-2005," database, FamilySearch (https://familysearch.org/ark:/61903/1:1:2DDB-NL3 : 13 December 2014), Mary Eliza Tinckham, 1888; from "England & Wales Marriages, 1837-2005," database, findmypast (http://www.findmypast.com : 2012); citing 1888, quarter 3, vol. 1A, p. 235, Kensington, London, England, General Register Office, Southport, England.
0 -
I've had a look at an interesting example, Mary Eliza Tinckham's marriage in Kensington in 1888:
No 'others' visible:
But if you look at the underlying data for Mary Eliza's persona, visible here: , these three 'Others' pop up (summarised):
Meanwhile freebmd (which again gets the gold star here) has:
while FMP has
It's pretty clear that there are 2 problems here:
a)
FMP has failed to index Mr Alloway (and almost certainly others) correctly, putting him on page 236, so FS doesn't pick him up as an 'Other' for Mary Eliza, instead dutifully showing him as on page 236:
b) FS has a user interface problem in that the 'Others', to the extent FMP has provided them, are missing completely from the normal Record display.
I do suspect that the FS problems reported on this thread are largely if not entirely user interface issues only.
The citation for Mary Eliza on FS is dated 2014, by the way, which again makes me think the data is fine:
"England and Wales Marriage Registration Index, 1837-2005," database, FamilySearch ( : 13 December 2014), Mary Eliza Tinckham, 1888; from "England & Wales Marriages, 1837-2005," database, findmypast (http://www.findmypast.com : 2012); citing 1888, quarter 3, vol. 1A, p. 235, Kensington, London, England, General Register Office, Southport, England.
(probably a duplicate following the usual edited-comment-disappears problem) https://www.familysearch.org/ark%3A/61903/1%3A1%3A2DDB-NL3 https://www.familysearch.org/platform/records/personas/2DDB-NL3 https://familysearch.org/ark:/61903/1:1:2DDB-NL3
2 -
P.S. Mary Eliza definitely married James Edmund Alloway: https://www.familysearch.org/ark:/61903/1:1:3PPL-7N2
0 -
Excellent analysis @MandyShaw1 - I had always assumed that FMP had started with the FreeBMD index but the fact that JE Alloway is right on FreeBMD and wrong on FMP suggests otherwise.
Ancestry, by the way, which I am sure started with the FreeBMD index for stuff up to 1915, matches FreeBMD and is correct.
2 -
The freebmd and FMP versions come from completely different indexing initiatives-I was involved with the latter in its early days and know some of the history.
2 -
The current format of sources within the above collection are likely to cause serious damage within Family Tree if not rolled back to how they appeared previously.
As an example, see https://www.familysearch.org/ark:/61903/1:1:2DQV-B9G. Most users would immediately believe this confirms the spouse of George Andrews was Catherine Gostelow, but there are multiple names indexed under the reference provided (as shown on other websites), so it would be likely (in this and all other similar instances) that the person listed at the top of the list would necessarily have married the person listed immediately below. (Often the name below is that of a person of the same s*x, which would not be legally possible in England & Wales until fairly recently.)
As long as this new format (only one "Possible Spouse" shown from the original entry with its multiple suggestions) there are going to be a lot of errors made by inexperienced Family Tree users in attaching a completely incorrect name for a spouse.
Would a moderator kindly refer this to the appropriate team, so a reversion to the original source presentation is made before too much harm is done.
If you have access to Find My Past, see https://www.findmypast.com/transcript?id=BMD%2FM%2F1856%2F3%2FAM%2F000030%2F037, which shows the three alternative suggestions of George's spouse.
(So sad that we are still unable to post screenshots to illustrate, when it appears possible for moderators to still do so.)
0 -
There is another thread containing more detail on this:
1 -
Thank you for finding that. I thought I probably had raised the issue previously, but:
(1) I admit I find it difficult to make searches in Community for topics previously raised, and
(2) Knowing how literally inexperienced users take FamilySearch suggestions, I feel this really needs to be addressed as a matter of urgency rather than the problem still remaining without (at my previous post) any moderator acknowledgment that the matter has been escalated.
I'm pleased to see (in April) we were both able to use screenshots, which clearly illustrate the problem with the sources being offered in their current format.
2 -
Thank you for your feedback. We have gathered all the information and this has been escalated.
1 -
Many thanks for that, Ashlee.
0 -
Would be most grateful on any update on this, as two months have elapsed since your escalation of this important issue.
As I carry out work daily on individuals who spent all / most of their lives in England, I see this issue virtually every day. What concerns me is that many users are accepting that (from my original example of 1 April) the person Oliver Shaw married was Elizabeth Margaret Brady, so - without being able to see the other options - confidently add this detail to his Family Tree profile. Not so bad if there were just one or two of these instances, but the whole collection of England and Wales Marriage Registrations has been effectively corrupted by some action during the last few months. This might be the result of how an update to the collection has been sent across by Find My Past, but (again as illustrated above) neither Find My Past itself (or the FreeBMD website) are displaying these records with just one option (from a list) as the spouse. It is just a matter of fortune if that one suggested partner happens to be correct, so - as I can't stress enough - much unnecessary damage is probably taking place in connection with relationships (of English & Welsh individuals) that are currently being added (daily) in Family Tree, by unsuspecting users.
The "fix" for this is probably quite straightforward (as with another unrelated issue, it should just involve reverting to the former way these records / sources appeared). So, it would be great if the "appropriate team" (possibly in liaison with Find My Past) could deal with this as a matter of urgency, in order to help stop further damage to Family Tree. Many thanks.
3 -
@Paul W When I analysed one of these in detail (above) I could see all the data in the underlying Record json, it just wasn't appearing on the screen. I can't, therefore, see that this can possibly be down either to FMP or to the feed from FMP to FS.
@Ashlee C. While a fix is developed, how do users not involved with this thread know to avoid attaching these FS Records as sources without cross-checking the information? Could there not be some sort of warning shown?
3 -
Thank you for your ongoing monitoring of progress on this matter. Hopefully it will be resolved before too many relationship errors are added to Family Tree.
0
