Genealogy Chat

Now, the Free BMD records on Ancestry were done by OCR, and this was from microfilm, not the originals.This is how they were able the facility up and running so quickly, but there was a downside, as many people found Where pages are indexed wrongly, it is often down to a poor image on the microfilm.

Some of the mistranscriptions you just have to laugh at, but at the time ...when you've been searching for months (maybe years in some cases) it's not particularly amusing. A lot of relief though. Thing is, I can understand how *some* mistakes are made with the old style handwriting, but there are other times when it's very clear. It's a challenge of course, and we'd be worried if they were all easy, but sometimes you just want them to work right. Just today I found a Marks that was Indexed as Marts. I like the male and female mix ups though, that really amuses me...mostly because when I'm adding siblings or children to my tree I continually forget to change from the default male... and have males called Fanny or Elizabeth *grins* I usually catch it pretty quick though.

As time has gone by I've become very imaginative with my searches, trying to think how names can possibly be mis-transcribed. The surname Garratt has many usual variants, many of which can be found using wild cards... but Parrot!?

I sympathise with your name, Kathlyn. I have some in my tree and I can never find them - I usually end up doing three separate searches instead of one - you get Mawdsley, Mawdesley, Maudsley etc and I am most surprised that my search for 'Mordsley' came up with nothing. A lot of the confusion is in the style of the old lettering. My Siggs family have come up as 'Liggs' (the old style S has been mistaken for an L) or 'Seggs' or even 'Sigge' - curiously the National Archives census site has them all down as Siggs, it's Ancestry that come up with these creative ones. I've had Ursula coming up as Mirala, her daughter Ursula V coming up as Wenta (with Wenta, I can see how it could be read as that name but I know it's Ursula) - the only way I found these two was by looking for their partners. Somebody mentioned it before, but I do try to submit the name I know to be correct to Ancestry - don't know whether they've updated lately, but it might help someone. And as you said in the original post, Liz, I've been doing some transcribing for FreeBMD (haven't had time lately with all my uni work) and I'm still on the first page - it keeps getting sent back to me because there are errors and it's so frustrating.

One of the census transcriptions was done by someone not living on planet Earth, cannot remember which one (1871 ?), but people living in Somerset were indexed as born in Somalia - same abbreviation apparently. Di

I was unable to find my grandparents on the 1901 census. I knew the area they lived in, etc, etc. They were John and kate Maudesley,but no luck. I appealed for help on this site and some dear person found them for me. They had been miss-transcribed as Mauderley. The reason I did not find them was becasue it was at the very start of my rersearch and i was unaware that you could use a wild card to help your search. Kathlyn

The worst mistranscription I found with my lot was on the 1891 census... I couldn't find a particular family even when trying an assortment of different spellings. The family were Albert E Winbow Matilda and their children Albert Matilda Charles. They were transcribed as Elizabeth Hevenor male! Gladiola and their children Eller E male Mosella female Elizabeth male! To be fair to the poor transcriber,the image was very,very faint!...we had such a laugh when we finally found them! Sandra. xxx

Andrew All the bmd's on Ancestry are completely FREE Reg

It's the indexing of Ancestry's BMD images that annoys me the most. They always list the first and last names on each page, even when the last name is a late registration written in at the bottom and therefore out of alphabetical sequence. This means that often the search engine fails to find a page because the search relies on all the names being in alphabetical order. For an expensive site like Ancestry this is just sloppy.

Wasn't it the 1881 census which was transcribed by prisoners in a USA jail? OC

Found my great uncle Walter Murray as Waller Muncey on Ancestry, on looking up original saw was Walter Murray, only through going through, putting his first name and year and where born, did I see this and then went into original. Had this happen before where kind members on Tips found them for me. Also the Registrars can hear wrong, looking for ages for my 2 x great grandmother's family name which I thought was Hazel , which was on my great grandfather's birth cert. Only through a kind Essex FHS member, finding out it was Layzell, was I able to go into the correct family. Three years I was trying to find this family.

I very much doubt if OCR was used for census returns. It is not clever at reading bad typing and even the best OCR needs to 'learn' handwriting (and is still not clever). With census UK returns being handwritten these were hand transcribed by 'furriners'. Who else would have Bradford, Yorkshire, Wales, England? CB

Same for the Pallots marriages on ancestry - check out the actual image of the note if one partner sounds right but the other is totally way out, because the names transcribed can be very wrong.

I was told that the census were transcribed by OCR (optical character recognition) not by humans, which could account for some of the glaring errors I have found. I searched for months for James ECCLES and finally found him under James E CALES, there was a small space between the first two letters, I had to search hundreds of James born in 1875 in Manchester to find him, but I knew he was there, and I had found his son with the grandparents. As soon as I saw it, I knew it was the right one. The enumerator was worse - he left Edmund Tanner out of the middle of his family entry - and stuck in some unrelated visitor instead - visitors are supposed to go at the end of the family, not in the middle. I hope you sent in corrections for the names you found Diane

Lol.... my gt gt gt grandfather.... [i have the certs] was born William Landale Walker.... married as William Randall Walker.... and.. was buried as William Randolph Walker.... cannot even blame the transcribers ! Sue xxx

Some mistakes I can sort of understand how they're made, but there are some that are just laziness (or maybe tiredness) on the part of the transcriber. I'm gone through and added alternate names to all four of them and added a comment to James, but there's not a lot else can be done. Can't even do that on the bmd when they've said they're not sure about what an image says but it's quite clear to others.

Done a similar thing myself - only on IGI, which is free, but very badly transcribed. I knew my Gx3 grandfather was called Edward Burt, I had to find his marriage. They had an Edward Burt, right time place etc marrying a Mary Boulet Evans, 10 years his senior. Having found this, I went to the records office to check it out on the parish records. Her name was Mary DOUTCH Evans (very important as Doutch was the name of her father), and she was the same age as Edward!! I could have spent ages looking for a non-existant person in the wrong time - 10 years before she was born!! maggie

I'm looking at one line from my family, Hiscocks. Their name has been mistranscribed several ways but as I was expecting variations I was fortunate enough to find them. I have however spent quiet a while scratching my head over why James Hiscocks (Bc1811) was living with a bunch of Fishers on 1841, one of whom had the same first name as his wife (Rhoda, the age was even right). When I couldn't find a marriage between a James Hiscocks (of any variation) and a Rhoda, I thought maybe she'd been married before and widowed. So I started looking on IGI for a first marriage of a Rhoda to a male Fisher, only to find the marriage of James and Rhoda HARDEN in 1832. I set about trying to find out who Rhoda Fisher was and where Rhoda Hiscocks was on 1841 if not with her husband. Lots of head scratching later I decided to go back to take another look at the entry. Only this time I decided to look at the original image... which I should have done before. What did I find? Rhoda and their daughters were all Hiscocks. The transcriber had continued the FIRST surname from the previous scan and not the LAST of which James Hiscocks was... his girls were all over the page surnameless so instead of carrying on the last surname the transcriber had some how messed up. How is such a mistake made? Moral of the story and TIP - Always check the original image even if you think there will be no more info on there. It was such a silly oversight on my part and has wasted a lot of time and effort.

The bain of my life >.< I know it's a tough job, I do transcribing for freeBMD as a volunteer. A lot of effort is put in to making sure that we list the correct information. So it really annoys me that Ancestry, a site we pay to use, has such poor transcribing on the censuses. (explanation and tip below)