Genealogy Chat

Is this a crazy idea? bit of a read - sorry! Purely out of curiosity (light headed after midnight!) I have been browsing the National Archives 1901 census site for unusual entries and so have been searching on increasingly ‘improbable’ combinations of (first two) letters to see how far I could get before I ran out of hits. The result was astonishing as you can see from the samples: William Lmp 48 Suffolk Campsea Ash Suffolk Wickham Market Market Gardener Joseph Lkin 43 Salop Prees Salop Prees Farmer Hardy Mn 46 London London Liberty Of Saffron Hill Mary Mghluyale 28 Lanc Liverpool Barrow Barrow In Furness William Nforbey 39 Evenley Norttamptonshire London Paddington News Agent James Vgee 57 Suffolk Framlingham East Suffolk Framlingham Bootmaker Journeyman Margaret Tvis 34 Deptford London St Paul Deptford I apologise if there really is a Lmp family from Suffolk! I've found hundreds of equally, or even more, bizarre transcriptions. The problem from our point of view is that many of these are so way out that there is almost no chance of finding them in a surname search no matter how imaginative we are with different spellings. So these mistranscriptions often never get seen, never get compared with the original image and therefore never get reported or corrected. There must be tens of thousands of gross mistranscriptions like these on 1901, which are mostly inaccessible to researchers until a better transcription is available. Now I know how enthusiastic and determined we all are about finding our relatives and also how generous people are in helping others. If that energy and commitment could be brought to bear on the problems besetting the 1901 census it wouldn’t take long to iron out most of the problems. The big obstacle is that it would cost a small fortune. So I have an idea for getting these hidden errors out into the daylight and correcting them without great expense to the National Archive and without resorting to a new transcription. I am thinking of putting the idea to the NA team. I would like comments and criticisms before I do. This is what might happen: 1. NA compares an accurate database of current UK surnames (eg electoral role?) with all the surname entries on 1901. All those that do not match current surnames are flagged up and listed. This list will include some correct transcriptions of surnames that have died out and lots of foreign names that have since moved abroad again. But it will include ALL mistranscriptions that are not real names. (All the mistranscriptions that are not flagged up will spell other real names and most of those will be easy-to-guess variants of the correct name, so easier to find.) 2. NA then makes this list available for anyone to look at (both the transcription and the image) free of charge, and invites comments and corrections in the usual way. We get thousands of juicy images to view and transcribe for free. NA get the more difficult bits of their census raked over, and corrections proposed, by some of the most dedicated, motivated and experienced people around. 3. The suggested corrections (if any) are listed alongside the mistranscription in a database and can be sorted and read so that anyone looking for an elusive relative can look at the possible candidates as soon as someone has suggested that name as a correction. This immediate (free) access would be an added incentive for people to get stuck in and correct all these weird transcriptions. 4. There will be a lot of other errors picked up for other fields and these can be notified at the same time (but would not appear on the database – just surnames). 5. Over time NA would work through the database of corrections and update the census and delete each name in turn from the free list. The cost to NA would be the cost of programming the extra bits on the site and administering a big increase in reported errors. NA could take their time updating the census as the free list will be available for research until they do. NA would not lose much income from the free images since they are the images no-one can find currently (so do not pay to see), in fact the reverse could happen because as they become searchable they will produce income. The benefit to NA of improving the accuracy of the census would be felt in increased usage (income) and better public image and user satisfaction. The benefit to us users is obvious. Sorry about the long message. What do you think? You might say well the NA should do something like this anyway! John

John! Wow!! How Interesting!! Knew there were transcribing errors but not in those considerable numbers! Your ideas sound good to me!! Go for it !!

John Brilly idea! Even if it only gets rid of my Ololodon - who should be Holden... Yep, go for it! Marjorie

John Sounds good to me - might find what I'm looking for Ann

Thanks everyone for taking the time to read this. I frightened myself when I saw how much I'd written. No not a scientist Katazyna - just a bit of a mad inventor. I'll send them the idea and see what happens. I expect it'll take a while to get a reply but I'll update when I do. Thanks again. Good hunting. John

A good idea, yes. But, I being the cynical sort, find myself wondering whether the NA would be interested. After all, they already get our cash so why should they worry about incorrect transcripts.

Hello Carol. You might be right and they won't be interested. But they are civil servants so the profit motive is less important perhaps than doing the job properly. I think the biggest barrier could be getting someone to commit the cash and take the decision. Certainly from the point of view of an archivist it ought to appeal as a way of improving the service. We'll see I guess. Marjorie how on earth did they arrive at Ololodon? Have you asked for it to be changed yet? I'm still waiting to get some changes for Oudney currently shown as Ondrey, Oudrey and Oudsey. They seem to update every 3 months or so which is a bit frustrating. By the way anyone looking for new ideas for name variations to try, my bit of digging has shown lots of errors where there is an extra letter keyed in accidentally. So try adding letters next to those on the keyboard eg Copper becomes Cxopper. I've just seen the family Cfififex! Oh, from Russia so maybe it's right? Must stop searching for these, it's addictive. John

I very much doubt TNA will take you up, and not because your idea is crazy or because they don't care. The major reason is budget. Money allocated to them is finite. Archives come low down on government priorities because cash to archives isn't going to make them popular; cash to hospitals is. Contrary to popular opinion, TNA does NOT make vast profits. When people write that censuses make oh-so-much money for the government and for TNA, they're just wrong! TNA would have to pay for the electoral roll and pay extra if they intended to use it commercially. You appear to be suggesting building a website, or at least extending the ones they have. TNA don't build their own websites and I assure you that to get a half-decent website you won't get much change from £50,000. You can buy a house with what some companies charge. Annual webhosting fees to host, maintain and repair the website will also be affected. All their websites have to comply with disabled access guidlines, too. Now that Ancestry have their own transcription, is that extra investment actually going to be worth it? There are probably a thousand things TNA would rather devote their staff, resources and equipment to. Some of those records are 1000 years old and are irreplaceable, and they don't take care of themselves! It's not that genealogy is viewed with contempt or that TNA don't care about the errors (I hope you reported them, by the way). It's just that they have so many demands placed on them and they have to account for every one of them and battle to get funding.

Hello Twinkle. Your sober analysis is probably correct and I realise that this proposal is a long shot. You have made me think that maybe they'd go for a low tech, low cost option. We'll see. Thanks for your comments. By the way I haven't seen the Ancestry version, is it fully searchable too? John