Genealogy Chat

I have come across many transcription issues, especially with census data. I have had to resort to trawling through the census data for a location to find some relatives. Fortunately most of my family were rural based so searching a whole village is not as bad as the prospect of a town. The reply above about the use of OCR explains how this can happen. I found the transcripts for a whole village were really poor in one case, clearly the OCR had problems with the handwriting concerned.

Julie

Inky - Re:
"Other Churches possibly would have written that as Feb1706/7 Thus making clear that the marriage was in Feb1707 and in the Ecclesiastical year of 1706/7."

I don't think so - it wasn't just the ecclesiastical year that began on Lady Day=Mar 25 - everyone's year began then, at least in England - it got changed in 1752, so then the year began 1/1/1752. Scotland changed in 1600, and one occasionally sees Scottish documents that are dated with both the Scottish year and the English year, so a C of S record could conceivably date the marriage as Feb 1706/07 - that's how I've seen it written, but it's not very common.

The problem with dates can be a lot worse if your ancestors came from other countries as the change from the Julian to the Gregorian calendar took place over centuries.

Each country changed separately except for the main Catholic countries. Some did not change until the twentieth century.

Also many had separate ideas about when the year started.

Ah! -that explains a lot.

I have been aware of the year change from history lessons at school. But not everyone will be.

The Cheshire Parish registers are also being indexed and on a web site by the Liverpool University, but there must be about 300 possible registers, and they are only half way through. If you are lucky though they are the most comprehensive of the Parish records on line for Cheshire.

I shall just have to keep searching for my elusive burial.

There are not just alphabetical errors in transcriptions. As you are looking at quite old records, you are probably aware that the Church year used to begin/end on Lady Day. Dependent on how the Church record is written, that can cause confusion.

As an example of that, I have a marriage in Feb1707 which was transcribed as Feb1706. On checking the actual record for that Church the year was shown as 1706 from January through to March. April onwards was 1707. Other Churches possibly would have written that as Feb1706/7 Thus making clear that the marriage was in Feb1707 and in the Ecclesiastical year of 1706/7.

Some time ago I assisted in the FREECEN project. There were always at least two people involved in a transcription prior to the data going live. I am sure that not all transcription activities are as thorough as “FREE…”

But it is not only humans that make mistakes. Following is from an email sent to me by the Ancestry Team in July2007. Their email was in response to my whinge about transcription errors:-

Because these Historical Records have far too much text to be indexed manually, indexes are built using advanced Optical Character Recognition (OCR) technology to automatically recognize text within an image. We believe we are using the very best OCR technology available, but it is still less accurate than human eyes and brains. These limitations make it impossible to achieve the same level of accuracy found in other indexes. In cases where the original microfilm text is unclear, the OCR technology tries to guess words intelligently. It seems to have failed in this case.

Not at all, it took some time for the penny to drop when I saw my first hand written will. Beautifully written, neat and clear, but with these odd words dotted about.(1760).
I suppose Companies just hand out films? for people to make of them what they can in transcribing?
So now that I am aware of it I can be more careful, but people must miss connections if they have not seen these different uses of letters etc.

Thought I would see how many Mafsies and Jefsops there were coming up on Ancestry. On one page of burials I correct 6 mistranscriptions (not all for those names) and have found somebody with a Jefsop family in their tree, so not only transcribers who don't know about old handwriting!

One of the JESSOP names I found actually had Jessop or Jefsop in the transcription, so it is perhaps the responsibility of the transcriber to know these anomalies - especially as in some geographical areas some names etc are more common.

I think 300 mis- transcriptions in one county is a bit off! I haven't even looked outside Cheshire for these two names.

I would urge everyone to report these errors, The Family Search Org were very supportive. The other sites - Ancestry/FMP, have not yet responded. and I don't know where they get their information from, so may be it goes round and round..

As you have noted, a mane like Massie often has the first s written as a long s, very similar to a script f (except the foot curves to left), and is occasionally transcribed as Mafsie. Rarely, I have seen the Mafsie form translated as Mapie - the long s followed by a regular (but squiggly) s can be run together and read as a p with a high upstroke. It's even worse when they live in Tyrie, and one keeps finding this mistranscribed as Fyvie (which is another Aberdeenshire parish, but 30km to the south.

My ‘T/F’ problem was for a marriage in 1825.

On Ancestry the transcription of Pallot’s Marriage Index showed the surname as starting with an F. So too did the early typed marriage listing of that Church, which was from the London Metropolitan Archives. The image of Pallot’s was not very clear.

But the LMA also had films of both Banns and Marriages for that period. So on a visit to London I went to the LMA to view them.

This is how I described it on an email to another researcher.

Banns.
Called on Sunday 06, 13 and 20 February 1825.
This all happens to be in one handwriting. The 'T' of the surname and three 'F's' of February are similar, both looking like a stylistic '7'. The difference is that the T is a finite letter whereas the F is lightly crossed much like what I would refer to as a 'French 7'.

Marriage Register
The ‘T’ is quite clear and supported by the written date 'Twenty Second' (month) and (year) 'Twenty Five

So have you found this changes names? To me its knowing what the alternatives might be!
Years ago I am sure you could search on a site using the Asterisk as a 'wild card' but I don't seem to be able to do this recently.
Or - is there another method? It isn't a question of sounds is it, more interpretation .

It's not just lower case 's'.
Upper case 'T' can be confused/transcribed as upper case 'F'

The name was spelt correctly for those days, but when it was a double 'S' it was written like an unbarred F.
If you were spelling it out to someone you would have said 'S' not "medial S" just as in the words "Possession" and Permission " Administrators and Thousand " and a host of other words printed and written in this style.
They didn't actually speak with a lisp! sounding an F for an S.

If you look at any old will or document for 1750+' ---- this long 'S' which looks like an F doesn't have a full bar across it.
Of course in handwriting it might look like that. Nevertheless, it still means hundreds of names are mis indexed , and a lot of people will not be aware of this..

But the index will be correct if that was the spelling back then.

Also applies to "ss" when at the end of surnames


Occupations such as dressmaker would have shown as drefsmaker etc

It has come to may attention that I have missed 200 entries for a name, because the index has shown them incorrectly.
Names with a double ' s' in the middle and sometimes singly, have been transcribed as F. this being the long 'S' used in medial printing, and used much later in handwriting- up to 1875 I find..
Some indexes carry a note about this, but others don't.
So names like Massey, Hussy, Jessop, will be shown as Mafsey, Hufsey, & Jefsop, which the search procedure may not show.
Just a warning to try alternatives in these cases.