Unconfigured Ad Widget

Collapse

Announcement

Collapse
No announcement yet.

Hazards of exact transcriptions

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Hazards of exact transcriptions

    Just to alert you, in case you hadn't thought of this...

    Find My Past, quite correctly, insists on exact transcriptions of what is on a document (so far as it can be read). This is fine for personal names, and for placenames, such as villages, towns etc. It becomes more of a problem when it's for a county name.

    For the last century or so, county names have been pretty much fixed: there have been changes to boundaries, and counties have been invented, disbanded, re-invented, but the actual spellings of counties' names have been defined by the laws set up to create them.

    I recently pointed out (with no great expectation of the policy being over-ridden) that it might be helpful to show a 1911 place of birth entry as "Bristol, Somerset", rather than "Bristol, Summerset", but FMP has stuck to its regular policy.

    This is to alert you to the fact that, if you are using FMP (and, maybe, other sites) and cannot find your ancestor, you may have to experiment with alternative spellings for county names!

    Christine
    Researching: BENNETT (Leics/Birmingham-ish) - incl. Leonard BENNETT in Detroit & Florida ; WARR/WOR, STRATFORD & GARDNER/GARNAR (Oxon); CHRISTMAS, RUSSELL, PAFOOT/PAFFORD (Hants); BIGWOOD, HAYLER/HAILOR (Sussex); LANCASTER (Beds, Berks, Wilts) - plus - COCKS (Spitalfields, Liverpool, Plymouth); RUSE/ROWSE, TREMEER, WADLIN(G)/WADLETON (Devonport, E Cornwall); GOULD (S Devon); CHAPMAN, HALL/HOLE, HORN (N Devon); BARRON, SCANTLEBURY (Mevagissey)...

  • #2
    And place names. This is just one example I have found because they were my family & I knew they lived at Roudham, Norfolk c1840-1878 but I could not find them in 1871, admittedly their name had been mistranscribed but because it was such a tiny place I did a search just left the form blank & just put in the place name & nothing came up. Eventually I found them at Rendham? This is what it says just above the image--- Norfolk > Rendham > District 8 > but the name Roudham can be seen quite clearly at the top of the census form.
    Does nobody check these place names?
    Moggie

    Comment


    • #3
      Bad transcriptions are always exasperating, especially when they're so obvious, but - if the entry is shown to be incorrectly transcribed - the sites will usually change the database. In the situation I described above, the transcription was quite correct, in the sense that the householder had written "Summerset". What I was concerned about was that, if someone enters "Somerset" in the place of birth search window, then that "hit" won't ever come up, because the transcription won't be changed because it matches the document.

      Place names evolve and vary - I've got an example (one of my favourites!) where the PoB is given as "Demport", because that is - for sure - what the enumerator heard. If you've heard a strong Devon accent, then you may guess that the PoB is formally known as "Devonport"! I wouldn't expect that to be changed, because of the variability of placenames (how certain are you that there is nowhere called "Demport"?), but Counties have fixed name-spellings - certainly in the era of censuses.

      Christine
      Researching: BENNETT (Leics/Birmingham-ish) - incl. Leonard BENNETT in Detroit & Florida ; WARR/WOR, STRATFORD & GARDNER/GARNAR (Oxon); CHRISTMAS, RUSSELL, PAFOOT/PAFFORD (Hants); BIGWOOD, HAYLER/HAILOR (Sussex); LANCASTER (Beds, Berks, Wilts) - plus - COCKS (Spitalfields, Liverpool, Plymouth); RUSE/ROWSE, TREMEER, WADLIN(G)/WADLETON (Devonport, E Cornwall); GOULD (S Devon); CHAPMAN, HALL/HOLE, HORN (N Devon); BARRON, SCANTLEBURY (Mevagissey)...

      Comment


      • #4
        As I think I've mentioned before, it's a real nuisance if the surname is wrong on the GRO index. I have a Lidyard who appears as Sidyard in the GRO index and therefore also on FreeBMD and Ancestry. The certificate from the local office has the correct spelling.
        Uncle John - Passed away March 2020

        Comment


        • #5
          And even having the number 1 before all the names, making them impossible to find, unless you know the address!! I would not have minded but the number 1 is not on the original census. I have tried to get FMP to change obvious county mistakes, but they won't budge despite my challenges and despite the original looking different to what they have transcribed. In desperation I have suggested that they try to come up with a solution to overcome some of the obvious county problems but they never replied to that suggestion. I think the only answer does appear to use ever more imaginative ideas. My census finds over recent months have been very challenging!!


          Janet
          Last edited by Janet; 20-08-11, 20:35.

          Comment


          • #6
            I did get them to back down on one occasion - but they were wrong on the basis of their own policy, so that wasn't such a big thing.

            Perhaps, next time they have a questionnaire (I missed the recent one, because I was too busy) we can all plead for some system to allow glaring errors like that to have some kind of searchable alternative - even if they persist in keeping what's on teh original (which is a good reliable starting point, in one sense) perhaps they could do something like:
            Summerset OR Somerset
            I think that that would cover the problem.

            Christine
            Researching: BENNETT (Leics/Birmingham-ish) - incl. Leonard BENNETT in Detroit & Florida ; WARR/WOR, STRATFORD & GARDNER/GARNAR (Oxon); CHRISTMAS, RUSSELL, PAFOOT/PAFFORD (Hants); BIGWOOD, HAYLER/HAILOR (Sussex); LANCASTER (Beds, Berks, Wilts) - plus - COCKS (Spitalfields, Liverpool, Plymouth); RUSE/ROWSE, TREMEER, WADLIN(G)/WADLETON (Devonport, E Cornwall); GOULD (S Devon); CHAPMAN, HALL/HOLE, HORN (N Devon); BARRON, SCANTLEBURY (Mevagissey)...

            Comment


            • #7
              It does seem rather blinkered of them not to adopt alternative options...I might jump ship from Ancestry if they were a little cheaper and more flexible.

              Comment


              • #8
                This thread shows me that researchers are becoming less skilled and rely on being spoon fed information.

                I would far rather the transcript recorded what was written rather than someone include a false record because they know the area.
                It is such "knowledge" that leads to huge errors.

                Transcripts should record the original, the researcher then weighs the evidence supplied and draws their own conclusion based of facts not guesswork of a transcriber.
                Cheers
                Guy
                Guy passed away October 2022

                Comment


                • #9
                  But Guy they do not always record the original. Looking for a family in Poplar about a month ago I could find them nowhere. I was lucky though because I knew the address that they were probably at, and so keyed in the address and there they were all 6 of them each transcribed with the number 1 in front of each name. A glance at the original saw no number 1 in front of any name. I have corrected it with FMP and they accepted my findings. Anyone else looking for that same family will now find it, but if I had not bothered to correct it as I have done about 100 others, then others will not find these families because most of the ones I find are quite obscure. Some have been counties which I have corrected and many have been accepted but not all. Some of us are very experienced researchers in our own counties and know what we are talkng about. And as an experienced researcher I have looked at the transcription, know it's wrong, look at the original and what I see is what I know to be correct, and these are the ones that I object to FMP not changing when I have suggested it, and there have been a few of these.

                  If the transcribers wish to be precise then why not check one census against another. I just find it incredible that an 1891 census for people living at one address in a county, in 1901 can be living at the same address but in a different county, town or village! These are the ones I am challenging and these are the ones not always accepted.

                  Janet
                  Last edited by Janet; 21-08-11, 11:44.

                  Comment


                  • #10
                    Thank you Guy. As a transcriber for LanOPC, my brief is to enter the exact record and not to interpret what I think is/was the record.
                    Having grown up in Liverpool, I prefer to specialise in its Parishes where I know most of the areas and often come across spelling mistakes with place names and so tempting to correct them but , "type only what you see !" always applies.
                    I think sometimes, the subject of correct spelling is subjective in this time period. e.g. Does Philip have one or two 'l's and at some point in time, Liverpool was Leverpool.
                    The biggest problem I have is bad handwriting badly faded after all these years so for those less tolerant of transcribers.....Have a heart, we doing our best.

                    Comment


                    • #11
                      You misunderstand my intention.

                      I fully recognise that there are great "risks" in second-guessing personal-name spellings, settlement-name spellings, occupation spellings... I was suggesting this only for county-names, and that is because there is a fixed and restricted list of county-names for the period in question. The names didn't change during the period 1841-1911 (that is, so far as I know - I'm prepared to learn differently, given the evidence), even if the boundaries did. I certainly would not suggest, for instance, "correcting" a county name which was different from what was expected as a result of a boundary change.

                      Please note, too, that I am not suggesting that the exact transcription is removed. I am suggesting that it should be possible to include a parallel searchable entry with the conventional spelling of the county-name. The National Archives has such a facility, I suspect, on the basis of my own experience of "hit" lists. Presumably TNA's version is a kind of "fuzzy" search. ... FMP allows a name-variants search for people; perhaps they could incorporate a "variants" facility for county-names? That would solve the problem.

                      And please bear in mind that I did find that "hit", even though it had been recorded exactly according to the original.

                      I also understand, from trying to read script for myself, that it can be extremely difficult to interpret, sometimes: upper-case L and S, for example can easily be confused; and "rn" / "m". In our church there is a modern-printed church anthem book where the phrase "transforming clay" has been rendered as "transforming day"... so it still happens.

                      We are fortunate to have such easy online access to so many records, and I am grateful for all those who have spent time making them available. (I hope people appreciate the transcriptions that I have made available, myself.)

                      Christine
                      Researching: BENNETT (Leics/Birmingham-ish) - incl. Leonard BENNETT in Detroit & Florida ; WARR/WOR, STRATFORD & GARDNER/GARNAR (Oxon); CHRISTMAS, RUSSELL, PAFOOT/PAFFORD (Hants); BIGWOOD, HAYLER/HAILOR (Sussex); LANCASTER (Beds, Berks, Wilts) - plus - COCKS (Spitalfields, Liverpool, Plymouth); RUSE/ROWSE, TREMEER, WADLIN(G)/WADLETON (Devonport, E Cornwall); GOULD (S Devon); CHAPMAN, HALL/HOLE, HORN (N Devon); BARRON, SCANTLEBURY (Mevagissey)...

                      Comment


                      • #12
                        Totally agreee with everything you have said, Christine. I also hope that others have gained from the transcriptions that I have already corrected on FMP.

                        Janet

                        Comment


                        • #13
                          Thank you, Janet.

                          I have, as I anticipated, received a message saying that they won't change the transcription. That's fair enough. I have suggested that they might consider extending the "variations" facility to other fields, to get around the problem.

                          Christine
                          Researching: BENNETT (Leics/Birmingham-ish) - incl. Leonard BENNETT in Detroit & Florida ; WARR/WOR, STRATFORD & GARDNER/GARNAR (Oxon); CHRISTMAS, RUSSELL, PAFOOT/PAFFORD (Hants); BIGWOOD, HAYLER/HAILOR (Sussex); LANCASTER (Beds, Berks, Wilts) - plus - COCKS (Spitalfields, Liverpool, Plymouth); RUSE/ROWSE, TREMEER, WADLIN(G)/WADLETON (Devonport, E Cornwall); GOULD (S Devon); CHAPMAN, HALL/HOLE, HORN (N Devon); BARRON, SCANTLEBURY (Mevagissey)...

                          Comment


                          • #14
                            Christine, you're so helpful on this 'Site' and far more knowledgable than me...................my remarks were intended as a general comment rather than aimed at a personality.
                            I often read curses of Transcribers on this and other Sites by Researchers who have no idea of the difficulties that can arise and I think they should have a 'go' at transcribing themselves.

                            Comment


                            • #15
                              Thanks, Alan.

                              It's always worth making sure that people do understand the problems of transcription - it doesn't just happen! Or, perhaps, if it does "just happen" (by OCR) it almost always needs the human eye to sort out the errors!

                              Christine
                              Researching: BENNETT (Leics/Birmingham-ish) - incl. Leonard BENNETT in Detroit & Florida ; WARR/WOR, STRATFORD & GARDNER/GARNAR (Oxon); CHRISTMAS, RUSSELL, PAFOOT/PAFFORD (Hants); BIGWOOD, HAYLER/HAILOR (Sussex); LANCASTER (Beds, Berks, Wilts) - plus - COCKS (Spitalfields, Liverpool, Plymouth); RUSE/ROWSE, TREMEER, WADLIN(G)/WADLETON (Devonport, E Cornwall); GOULD (S Devon); CHAPMAN, HALL/HOLE, HORN (N Devon); BARRON, SCANTLEBURY (Mevagissey)...

                              Comment


                              • #16
                                typical, something like always comes up after ive done the surveys for findmypast on improving their records and the site!

                                Comment


                                • #17
                                  I have a little list of things I plan to tell them next time I actually catch a survey. I was busy and delayed until the deadline was past for the recent one.

                                  In fact the things that bugs me most (which also happens in the recent upgrade here) is that I put the cursor in the cell I want (or the place I want), only to find - usually after I've started typing - that the system thinks I should start with the given-name cell (or the beginning of the window). On FMP that means that I find I've overwritten a name with a date, or typed a MMN, or a spouse's surname, over the given name of the person I'm trying to match. It really bugs me!

                                  Christine
                                  Researching: BENNETT (Leics/Birmingham-ish) - incl. Leonard BENNETT in Detroit & Florida ; WARR/WOR, STRATFORD & GARDNER/GARNAR (Oxon); CHRISTMAS, RUSSELL, PAFOOT/PAFFORD (Hants); BIGWOOD, HAYLER/HAILOR (Sussex); LANCASTER (Beds, Berks, Wilts) - plus - COCKS (Spitalfields, Liverpool, Plymouth); RUSE/ROWSE, TREMEER, WADLIN(G)/WADLETON (Devonport, E Cornwall); GOULD (S Devon); CHAPMAN, HALL/HOLE, HORN (N Devon); BARRON, SCANTLEBURY (Mevagissey)...

                                  Comment


                                  • #18
                                    Coming in a bit late here, but...

                                    To my mind, there are two main types of mistranscription.

                                    The first is the genuinely unreadable original. I've seen a few of those and marvel that anyone can make anything out of them, let alone an accurate attempt.

                                    The second is the perfectly readable original which has been mangled, either by a typing error or by someone who has no working knowledge of the English , but who is spelling the word out letter by letter as THEY SEE IT. I feel that any paying site should correct these errors promptly.

                                    I was baffled for years at the temporary disappearance of a family. Eventually I found them, their birth place having been transcribed as LANARKshire - the original said LANACshire, a scriptural error by the enumerator for LANCAshire.

                                    OC

                                    Comment

                                    Working...
                                    X