Should transcribers have local knowledge?

[Sinann]

There is transcribing what you see and knowing what you are looking at.
There are loads of the name Fox transcribed as Fose in a set of records, the problem is not that it looks like se because it does and often looks like sc, that's what it's meant to look like, the problem is the transcriber didn't know what an x looked like.
They should at least know how the writing they are transcribed had its letters formed.

A person familiar with the names of a country would have at least questioned Fose and (hopefully) discovered how an x was written.

[c-side]

There is transcribing what you see and knowing what you are looking at.

That's it exactly 

Christine

[Guy Etchells]

snip

As a transcriber you are supposed to write exactly what is written and I do 99 per cent of time, but obviously when you are transcribing for databases I am  certainly not going to type in Bristol Enland ( even if England is misspelled like that on the original document) I will type up Bristol England, or it will B8gger up the set.

snip


Kind regards:)

There you have given one of the main reasons for the errors on the Ancestry site.

It reminds me of the fiasco when the 1881 Census was transcribed in a joint effort between Family History Societies and the LDS.

Many errors were present in the 1881 Census (transcribed by members of UK family History societies, checked, before being sent to Salt Lake City to be keyed in by members of the LDS).

In this case the church members did not handle the actual census facsimiles, but a transcription in block capitals produced by UK volunteers.
The errors were introduced at the keying in stage as was proved by comparing the transcription with the final product.

There was also geographical confusion by those who keyed in the data which put Sunderland (N.E. England) into Scotland, because someone thought it must be Sutherland (NE Scotland).

This therefore supports the local transcriber theory, however on the National Burial Register [NBI] (1st edition) there were huge chunks of transcription errors produced by local volunteers in the UK.
I had previously transcribed the burial register for Bottesford, Leicestershire and had made it freely available online. When the NBI came out I was aghast at the differences between my transcripts and the NBI transcripts and dug out my fiche of the original records to check to see if I was in error.
Having re-assessed my transcripts I then asked a friend who was well versed in old handwriting to double check as I know that it is easy to read what you want a word to appear as. Sure enough she confirmed my transcripts were correct and the NBI was in error. (I should add the compilers of the NBI accepted my transcripts were correct and amended the Bottesford entries in the later version of the database.)

Cheers
Guy

PS Yes Majm that is the meaning of the word idiot. However I was using it in a self effacing way (banter) rather than in the derogatory sense.

[StanleysChesterton]

One place I have an interest in is one that's pronounced unusually, locally .... and spelt in lots of ways in transcripts, by vicars, book writers and now FH sites.

It's hard to type the pronunciation, but locally it's pronounced with a "silent L": Co(L)-dee-cot.

Official modern day spelling is Caldecote.

Online/transcriptions and over time it's changed, I've seen: Caldicote, Caldecote, Caldicot, Caldicott, Coldicot, Caldcoat, Coldcoat .... and every other variation/combination you can think of. 

Even the online sites' auto-correct/'pick a place' offers varying choices of spelling and sometimes you're unsure whether choosing this place is correct at all because ...  it's also a place name that's got a lot of alternate locations too - so by choosing it from a list you could be declaring you're after an entirely different place with the same name. There's another place of the same name just 10 miles away - and 5-6 others dotted around.

I think this is a less-common situation of what should be typed if you KNOW the spelling, but neither the original person who said where they lived, nor the transcriber, knew how it was spelt.  If somebody from a small village moves 100 miles away and the census-taker says "where are you from?", they could spend ages on the doorstep with the householder saying it again in their strong accent and the census-taker saying "go again.... er, and again....slower...."  Clearly it should be typed as it was written down, then transcribed, but I like the way the "A" site enables users to submit alternative spellings etc which would help to sort these out (when discovered).

[MJW]

Interesting discussion.

I transcribe for an OPC (online parish clerk) website.  We transcribe what we see - this is easy to say but not always easy to do!!  Sometimes you can look at a word that's difficult to read several times and see it differently each time!

I think that local knowledge can sometimes help as long as long as this supports - but does not replace - "transcribe what you see".  As previously highlighted, having knowledge of old handwriting styles is really useful.

We don't correct mistakes, spelling errors etc. or make assumptions about what the minister intended to say - and for many names in the past there were no standard or correct spellings.  However, to help users of the OPC website, I have the option, for example, to:-
- insert [sic] after a word  (e.g. I would use this if a word was clearly written but looked odd)
- insert ? after a word to show that I wasn't certain what was written
- insert an explanatory note (e.g. surname not recorded, abode name covered by ink blot & unreadable)

Regards
 Malcolm

[geno500]

Yes they should,but in the race to get the next census online first the big companies use non local people,the 1911 census was farmed out to India,now they may well have a knowledge of the language but to transcribe a census was crazy,I would not dare to transcribe a New Zealand one with place names with all those R's in the names the same with other countries I might share the language but local knowledge is whats needed

[LizzieL]

One place I have an interest in is one that's pronounced unusually, locally .... and spelt in lots of ways in transcripts, by vicars, book writers and now FH sites.

It's hard to type the pronunciation, but locally it's pronounced with a "silent L": Co(L)-dee-cot.

Official modern day spelling is Caldecote.

If you mean the Upper and Lower C straddling A1 just north of Biggleswade, I've heard it pronounced Cordicot, so as not to confuse with Codicote south of Hitchin, I suppose.

Not so far away, ones that have given me grief, Chishill (prononced like the woodworking tool) sometimes spelt Chishall, but not to be confused with the nearby village of Chrishall.

I have found that the further away a person is living from village of birth, not only more likely to mis-spell, but more likely to put nearest largest town instead. This could be because enumerator had never heard of it and neither party could spell it, so the enumerator asked "where's it near?" and when told a town he recognised was happy to record that instead of having a long conversation about a small obscure village.

[weste]

As i transcriber i am told to type what i see. There is a notes column where suspect things can be mentioned e.g putting male and the person is in fact female that is found occasionally on a baptism entry on the original entry. You are told to check other sources but i always put a note. The main problem is when they give a baptism date before the birth date on the original record. The software system will not allow the entry to put in that way round so you either have to put the dates in the other way round or you can put *.Problem is unless you get a year in somewhere that entry may be difficult to find. Its very useful to do some transcribing as you learn more about problems with data in respect to your own research. I prefer to areas i'm interested in and can get to check the original myself because you transcribe from different media cd, transcripts, photo register images for example.

[LizzieL]

One of my relatives that I have been searching for for a long time turned up this morning. He was born in Bury Sussex, but the census record said Berry (and had been transcribed as Berry). Perhaps a local transcriber might have picked this up. His surname was mis-spelt, but obviously the transcriber wouldn't have known this. His age should have been 62, but it looked like a blob of water had been spilt on the 6, so all that could be seen was a number with curves. One site transcribed it as 82 and another as 32. But the age had been originally been put in the female column in error and crossed out. The 6 was visible under the crossing out, so that should have given the transcriber a clue.