General Chat

good one IGP


I remember some years ago trying to help someone on GR to find their family.

I was a bit slow in recognising mis-transcriptions in those days .... but it was on the 3rd evening that I finally decided to take a look at the family with some of the correct names but said to have been born in Somalia. I'd dismissed them on the previous searches



Correct family ......... born on Somerset


The census taker had written Somt, where the t was a superscript.

The transcriber saw it as Som'l


and decided that it was Somalia

Perhaps the family was known locally as "Them Piggin' Corbetts.....!"

I agree with you Shane. It would be boring if things were too easy to find.

One of my names has an ending that changes all the time, so I mainly search with just the first letters and *. But when you have a name beginning Post transcribed with such variations as Bertanni and Voslans, using first names is the only way.

Well, it is interesting to read the other side of the story, and to read of your problems correcting your 'error' Berona. I agree some images are faint and some writing is hopeless. With regard to faintness, some years ago I got a copy of an image of some ancestors. Today I was looking for them on FMP and could not find them. I tried searching by the folio and page number etc and found that nothing on that page had been transcribed. When I looked at the image it was impossibly faint, unlike the version I have (thank goodness). I just hope I dont strike any more like that.

Huia.

FreeBMD rules are that we have to type exactly as we see it, and I do that, even when I see a record duplicated exactly - I do it twice as instructed.
If I make a mistake, I am happy to correct it, but recently I felt it was going a little too far.
Only one christian name is written or typed in full, with the initials only of subsequent names. ie, Mary Elizabeth Jane Jones is typed in the index as "Jones......Mary E.J." I recently was asked to correct my entry for a few names. I couldn't see what was wrong for a while until I finally worked out that I had put a space between the initials and typed it as
Mary E. J. instead of Mary E.J. They were asking me to remove the space between the initials because the original index didn't have a space there.
In order to do this, I had to log into FreeBMD and download the page I had created and save it in Notebook, then load it into the BMD system and delete the space; save it again then re upload it to FreeBMD and also log into the co-ordinator in Canada and re-upload to replace their 'wrong' version also. (Everything goes to two places - UK and to one of the co-ordinators in various parts of the world)..
I felt that was a bit much considering that I did put a 'full stop' (period) after each initial.

IGP, I am suggesting that the first names should be transcribed to Herberl, Alberl etc, since I was told that they have to type what the enumerator (appeared to) write. That particular enumerator did not cross a lot of his ts and there were people born in Manchesler. If the person who checks when an error is reported and studies the page carefully, they will see that the double t in the name Bettany which was transcribed as Bellany does not look like the double l in the middle of the names William Cullen and William Williams immediately above his name, as they have a loop. A bit of comonsense wouldnt hurt, but I guess that is asking too much, even though the first names are correctly typed, but that is more a case of the transcriber not noticing that they were written like Herberl, etc as he/she was expecting it to be Herbert.

Must go, off out for the day.

Huia.

freebmd will eventually have every entry transcribed by two different transcribers, which should cut down on mistranscriptions.

it's not just the writing that is difficult to read, some of the early images are very faint.

I apologise for all the errors I made when I transcribed for them!

By the way, they are always looking for more transcribers, so if anyone has an hour or two a week to spare, why not volunteer.

IGP - my g grandfather's name is wrong on the church register, not just the index. He married in a different county to the one where he lived, couldn't read or write so I assume the Vicar wrote down what he heard!

The trouble is that "Hindsight is a Wonderful Thing"

Assuming that the persons transcribing have absolutely no knowledge whatsoever of the correct spelling of names or places, then give a batch of census returns, especially ones which are not that easy to read in the first place, to 4 different people and they will come up with 4 different sets of transcriptions.

Give the same batch of returns to someone who is fully conversant with spellings and place locations and you will get another set of results.

I think where queries are raised, the site managers should take a balanced judgement on what the entry is supposed to be, taking poor handwriting and even bad spelling into account. Ancestry seem to have the best approach in the way that they add suggested alternatives to the search index.

Now, can anyone tell me how the name Norah came to be transcribed as Beval?

I transcribed for Ancestry - until I realised that I was helping them to make money whilst they didn't even give me a discount on my subscription, so I changed over to FreeBMD.

Whilst with Ancestry, I transcribed the British Criminal Registers and the handwriting was beautiful. So clear and easy for me to read - as the style was that which I had been taught when I first went to school (pre-WW2). I developed my own style as a teenager, then changed it again when my own children learned the cursive writing in the sixties.

However, on many occasions whilst transcribing for Ancestry, they refused to accept my spelling of handwritten names and made me type THEIR interpretation of what it looked like to them.....It made me feel a little insulted because it was all so clear to me - but when you come to think of it - I celebrate my 80th this year so I suppose they don't have anyone that old amongst their 'experts' and they think they know better than I do, as I'm 'only the typist'.

I will only accept the British interpretation of both handwriting and the spelling of British names.

.

Not sure, but I would have thought that FMP and Ancestry, as commercial organisations might have to pay for at least some of their transcriptions. That is one of the reasons they have to charge such high fees, to get their investment back,

I think they also make use of OCR (Optical Character Recognition) software especially where records are type written, such as many of the BMD indices, and this can also introduce errors if the copy provided to them is of poor quality.

My sister in Law's name was wrong on Ancestry, but this was down to their mistranscription. A 'b' was interpreted as an 'e' The page, whilst reasonably clear to a human is just slightly blurred, so I wonder if this was an OCR misread.

Free BMD, I know for certain, is done by volunteers, and, of course Ancestry up to 1915 uses the Free BMD transcription too.

Incidentally I cant see the point of 'correcting' to Herberl, Alberl etc - why? I doesn't serve any purpose.

As far as wrong entries in the BMD records are concerned, in fact, they can be changed, but I suspect it it's a rare occurrence.

I say this because my own marriage details were mistranscribed onto the actual GRO index. Both our surnames were mispelled .....! I reported this to the GRO and after about 6 months they replied saying that indeed they were, and it had now been corrected. This is their response:-

"Re: Correction to marriage entry in the Indexes.

Our investigations are now complete and I can confirm that the error with
your marriage entry is just in our Indexes and not on the actual marriage
entry.

I have informed our Receipt & Indexing Department of the error with the
indexes and they will amend our records accordingly in due course.
Unfortunately the amendment will be to our indexes only and will not be
replicated on any web sites owned by private companies."

So, down to me to persuade Ancestry to change theirs.

Well, I will have to send corrections for the first names that have not been typed as written by the enumerator, with an l instead of a t. I wonder if they will correct them to Herberl, Alberl, etc.

I did find my gt uncles' p.o.b given as Northumberland in the transcription, but it is quite clear on the original image that it is Philadelphia, U.S. of America. Now how does that look like Northumberland? There are certainly a lot of obvious mistakes on the transcriptions. I just wish we could add postems to the transcriptions.

Huia.

PigletsPal - are you sure that FMP & Ancestry's transcribers do it for free? I assumed they were paid.

Freebmd, Freecen and Freereg transcribers do do it for free but I am not sure about other sites.

Freebmd transcribers are told to always "type what you see" and yes, amendments are returned to the original transcriber and if they still see the same, the transcription stands.

Personally, I sometimes am amazed that transcribers can make anything out on some of the census and bmd originals.

Wrong entries on the original census pages and bmd indexes cannot be changed. For instance, my great grandfather's name is wrong in the index (and on his marriage cert) but it will have to stay like that for ever. I have added a postem to the entries against him and his wife on freebmd so that other searchers can see it.

I am glad that some online parish clerks, when transcribing from microfiche, put what can be seen and what other possibilities there are, from their local knowledge.

I was looking at a 1911 census of William Wycherley (head) and Alice Wycherley (wife) a couple of weeks ago on FMP. The 'ley' of William's surname has extended into the 'relationship' column, so looks like 'Wycher' in the 'name' column. However, Alice's surname is fully in the 'name' column. Both have been mistranscribed as 'Wycher', which seems pretty daft to me. Have informed FMP and they agree it's incorrect, but have yet to change it.

I read somewhere that Free BMD send any requests for amendments back to the original transcribers, who then make the decision whether or not to change it on the transcribed index.

Well, IGP, I hope that FMP will change the first names to Herberl, Alberl, Gerlrude, Ernesl, since that is what it looks like on the image! But they obviously werent looking carefully and automatically assumed that the first names were Herbert, Albert, Gertrude, and Ernest.

I just wish there was a way with FMP where we could add a note of what we KNOW to be the correct surname. I just hope any descendents of the aforementioned Herbert(l) and Albert(l) BETTANY/BELLANY/BELLAMY can find them in the 1901 census. But then, I am sure they are as intellegent as I am.

Huia with a high IQ and a big head to go with it!

My understanding is that all transcribers had (and have) to put down what they see, not what they think it should be

I was told that applied to ancestry, findmypast, even freebmd


On ancestry, there is a function to send in an Add Alternative Information and ancestry eventually adds that to the record and thus to the search function ......... sometimes it takes just a day or two, other times it takes weeks.


Incidentally, on ancestry, you can also change birth year and birth place.




sylvia

In answer to the original question, the purpose of the transcription is to provide a comprehensive index which can be interrogated by a computer search engine. Without it it would be impossible to search at all and you would have to go through the census images page by page to find what you are looking for.

As with any transcription exercise, there will be errors. Just an error rate of 1% (which is exceptionally good, as many commercial processes aspire to 5% error rate or better) means that for every million records transcribed, there will be 10,000 mistakes made.....!

I can only speak for Ancestry where I have made around 100 corrections which in all cases have been accepted and added to the search index. I can understand that FMP etc all insist their transcribers copy what they see, otherwise there would be mayhem, especially as a lot of transcriptions are done by people who don't have English as their first language. However, if it is later proved beyond doubt that there is an alternative this should be acted upon. Perhaps if enough people complain and threaten to move their business elsewhere, they might take notice.

Thank you ladies. At least it is possible to put a note on the Ancestry transcriptions, I take it, rather than just asking them to correct it and having to put up with it if they decide I am wrong.

Alcie C is transcribed as Alice L on FMP, so Ancestry has the C correct but not her first name.

It was only through finding her that I was able to find her father and brother. I had found Herbert on 1891, and his daughter Alice Catherine, so I looked for a marriage for her and found her in 1901, with her father and brother who were both mistranscribed. I was just lucky that she and her husband were living with her father and brother. We have to be devious to find them. And I had ticked the 'alternative' box.

Now I have to try devious means to find a few other missing people.

BTW, Herbert was my gt grandfather's cousin and gt grandmother's half brother (might even have been a full brother for all I know).

Huia.