You Had One Job…

I’ll try to keep this brief.

I’ve blogged a lot in the past few years about the problems with many of the databases on the major commercial genealogical websites and in particular the various county-wide English and Welsh parish register collections. I’ve already looked in depth at the issue surrounding missing registers and the difficulties that this can cause for the websites’ paying customers, so I’m not going to go over that ground again.

Instead, I want to focus on another connected issue that all-too-regularly rears its ugly head; I’m talking about the situation where incorrect and often misleading descriptions are attached to a collection of documents. I came across a good (or rather, a bad) example of this phenomenon today while carrying out some research into my wife’s Cheshire ancestors.

I was particularly interested in viewing the marriage record of Daniel Smith and Martha Coundye (or Cundy) who I knew had married in Prestbury, Cheshire in 1804. My initial attempts at tracking down the relevant entry met with, shall we say, mixed results. I knew that Findmypast was the place to go for Cheshire parish register material and I was quickly able to find the entry in the Cheshire Diocese Of Chester Bishop’s Transcripts Marriages 1576-1906 database but I couldn’t find anything relevant in the main Cheshire Diocese Of Chester Parish Marriages 1538-1910 database.

Search results from a search for the marriage of Daniel Smith and Martha in 1804 in Prestbury, Cheshire

I really needed to see the original entry in the parish register; not only is the Bishop’s Transcript (BT) a copy of the original (and therefore lacking any original signatures) but, in this case, the BT was heavily truncated, providing us with none of that additional information which can so often make all the difference when it comes to research in this period.

Marriage of Daniel Smith & Martha Coundye as recorded in the Prestbury Bishop’s Transcripts, 9 January 1804.
Cheshire Diocese Of Chester Bishop’s Transcripts Marriages 1576-1906, Findmypast

I checked the parish list which accompanies the database and this told me that the coverage for Prestbury extended from 1560 to 1910 and that 36,566 records were included in the database. (I guess that this means that there are 18,283 marriages as each marriage record generates two database entries.)

However, I was aware (from bitter experience) that these dates are, by necessity, merely start and end dates and that they can mask all manner of gaps in the records themselves. As I’ve blogged elsewhere, there are any number of reasons why there might be gaps in parish register collections so, again, I won’t go into that here.

My default policy in cases like this is to pop over to the relevant county archives’ catalogue to see what their coverage looks like for the parish in question. Cheshire Archives is one of those commendable record offices which allows you to download a comprehensive list of their parish registers, providing you with all the information you need: covering dates, archival references and microfilm references.

Turning to the entry for St Peter, Prestbury, I was quickly able to see that there were no obvious gaps in the holdings, and that there was a register (reference P 338/8504/18) covering marriages from 1803 to 1810.

Detail from Cheshire parish register list. Cheshire Archives & Local Studies

Cheshire Archives clearly had the register I wanted, so why was I unable to find the record on the Findmypast database?

It took me the best part of an hour to sort out what was going on and rather than give you a blow-by-blow account of the exhausting process I went through, I’ve summarised my findings in the following table:

Summary list of marriage registers for St Peter, Prestbury together with descriptions on Findmypast

So, as far as the Findmypast Cheshire Diocese Of Chester Parish Marriages 1538-1910 database is concerned no marriages at all took place at Prestbury between 1748 and 1828. In fact, the records are all there but about two thirds of them have been incorrectly assigned to the parish of Saltersford cum Kettleshulme while the remainder haven’t been assigned to any parish at all. So, if you’re searching the database for marriages in Prestbury in this period, you won’t find them.

I really don’t think it should be too hard to do something like this and to do it properly. Each of the sections of digital microfilm is preceded by the familiar title page which describes exactly what’s on the pages that follow; in the case of the register that I was looking for, Prestbury Parish Church Marriages from 1803 to 1810.

Description page at start of Prestbury marriage register (1803-1810)

I just can’t understand how it’s possible to fail to assign a placename to the marriages in this clearly-marked Prestbury parish register (likewise in four others) as Findmypast have evidently done. Even harder to understand is how the many thousands of marriages which are recorded in the Prestbury registers between 1754 and 1786 and between 1812 and 1828 have ended up being assigned to the tiny parish of Saltersford cum Kettleshulme.

The job of the commercial genealogical website here is a simple one; to take the material provided by the archive and to make it available to the paying customer via a suitable name index. It seems to me that creating the name index is the difficult part: making sure that all of the material you were provided with is actually accessible and that it’s fully and properly described is (surely!) relatively straightforward.

In the case of the Prestbury marriage registers we’re dealing with nearly 15,000 marriages over an 80-year period which are either incorrectly described or effectively not described at all.

If this was an isolated incident, it would be a simple enough task to report the problem and trust that it would quickly be sorted. But experience tells me that it’s anything but. These collections are full of errors like this and they all go towards making our jobs as researchers that little bit harder.

It’s just not good enough…

© David Annal, Lifelines Research, 10 October 2021

This entry was posted in Document Sources, research, Soapbox, Uncategorized and tagged , , , , , , , , . Bookmark the permalink.

8 Responses to You Had One Job…

  1. Thanks for drawing my attention to the short comings! I have been struggling with my Buckley and Milletts from Prestbury with FMP. Now I understand.

    Liked by 1 person

  2. brucefuimus says:

    FindMyPast had nothing to do with it. When Chester RO was assembling the stuff for its “supplier” to put online (not sure if “supplier” is the correct word because I don’t know what the contract / agreement says), it used the images from FamilySearch (logically enough since FS had already filmed virtually all the registers in scope) and contracted FS to produce the indexes to those images. These indexes were produced by FS volunteers.

    The matching images and indexes were then supplied to FMP, who loaded them up.

    Now, it is a “well known” fact that FS make a habit of messing up their indexes if there is more than one place on a film. I’ve no idea what percentage of stuff goes wrong, but one typical issue is that the place name from the first title page gets applied to all items on that film. On one Cheshire film, there is a Wilmslow register at the start of the film and a Wybunbury register later on. Sure enough, the Wybunbury events have been indexed as Wilmslow (warning – it might actually be the other way round, or even different places, but you get the picture.)

    In this case though, it does appear to be something different – film 2093547 is described as
    Parish registers for Prestbury, 1560-1990,
    Church of England. Parish Church of Prestbury (Cheshire); Church of England. Chapelry of Chelford (Cheshire); Church of England. Chapelry of Saltersford-cum-Kettleshulme (Cheshire); Cheshire Record Office

    So Saltersford is there, somewhere on that film, but it’s not the classic pattern because the first title page on the film is indeed Prestbury. Still, I’m sure that the issue is *probably* very similar.

    Having spent time in the FamilySearch message boards, the situation is even more complex than you might think. The actual fingers-on-keyboard indexers from FS are innocent because any FS index is a combination of the indexes of the images and meta-data for the film. (“Meta-data” = data about the data on the film). Seems like the place-name is often recorded in the index – but in some cases there is only one set of meta-data for the whole film and therefore only one place-name for the whole film. So the error is in the design of the meta-data processing by FS.

    So in summary, FS supplied the images and corresponding indexes but on a number of occasions, because of defective design by FS, an incorrect place-name is applied to some images on the film. The error was not made by a commercial company but a volunteer led / free-to-access organisation.

    FMP did as they were told / advised / contracted to do and loaded the images and their corresponding indexes.

    Liked by 2 people

    • brucefuimus says:

      Two methods to help anyone with a similar issue in the future.

      1) FMP created a PR Browse facility for Cheshire – see collection “Cheshire Vital Records Browse”. Using that, you can locate the PR and BT books for marriages at Prestbury. (Books, not individual records).

      Not sure what the selection criteria is for the date
      – 1804 +/- zero didn’t find anything
      – 1804 +/- 2y found the Prestbury PRs for 1803-1810 and Prestbury BTs for 1805-1820.

      Note 1: I believe that the selection of Prestbury *works* for selecting the books for the browse because FMP indexed this browse collection themselves. But they only indexed the start of each book.
      Note 2: This is just a browse of the images for the book but given that (in this case) we know the date from the BT, finding the PR is quick and easy, given that the original PRs are in date order.

      2) Alternatively, in this instance, an enquiry on FMP for Daniel Smith marriages in 1804 in Cheshire (no town, but county *only*) gives 7 results – one is the BT, the other is the PR – but the PR is indexed by an *empty* place-name, so you can see the relevant page but that image may not help with knowing which church it was. If so, you need to navigate back along the film strip to find what the church is.

      Method 1 only works if there is a Browse collection. Method 2 works if the data is there.

      Like

    • Interesting – and thank you.

      I would, however, take issue with your statement that ‘FindMyPast had nothing to do with it.’

      Findmypast had EVERYTHING to do with it. Clearly, I don’t have access to the terms of their contract with Cheshire Archives but Findmypast’s ‘contract’ with me (the paying customer) is (implicitly) to provide access to the parish registers held by their ‘Partner’ (i.e. Cheshire Archives).

      The method they choose to achieve this is not my concern. As a paying customer, I expect the records to be properly described. I don’t think that’s an unreasonable expectation.

      Cheshire Archives have chosen to make their partnership with Findmypast their preferred method for users to access the records and they have entered into a contract with Findmypast (a commercial company) to enable this. The fact that the commercial company have (apparently) used work created by volunteers to provide an index to the records raises a number of separate issues but it is unquestionably Findmypast’s responsibility to ensure (on behalf of their partner) that the records are properly described – as they are in the Cheshire Archives’ own catalogue.

      Findmypast have a responsibility to their paying customers which they are failing to meet.

      Like

      • brucefuimus says:

        “Findmypast have a responsibility to their paying customers which they are failing to meet”
        Actually, that’s a good point.

        I would, therefore, say that FMP had nothing to do with the original misindexing error – they were neither directly nor indirectly responsible for that process.

        What they are very much responsible for, is fixing errors that are brought to their attention.

        In the case of indexing errors like this, none of FindMyPast, Ancestry nor FamilySearch have a good record in correcting *bulk* issues, which are usually to do with a failure to assign a specific and correct place. If I raise a bulk indexing error with them, I inevitably get directed to the index correcting facility which, of course, is wholly impractical to use for a bulk update. I have to return their response emphasising the bulk nature, the impracticality of their suggestion and suggesting that they forward the issue to their technical guys. The problem is then that I hear absolutely nothing.

        Someone has to raise these issues with them – their own data quality checking can only do a sample. FMP have responded in the past – we had an issue with the images of the extended format of the registers for St. Helen at Witton, Cheshire. These are single line entries that extend over 2 pages. The FS supplied index pointed only to the lefthand page and (with no browse facility then), FMP could not access the righthand pages. Eventually, after many compliants, FMP did respond and fix the issue, but it was like getting blood from a stone.

        Liked by 1 person

  3. Glenys SYKES says:

    Please tell me that when you reported this problem, FMP assured you that they will be correcting this -I was going to say incredible error but, alas, it’s all too credible in my experience – without delay?

    Like

  4. DM_Walsh says:

    The ratio of hard-working techies to genealogists at FMP might be high. “Getting stuff live, and keeping it live”. I am sure I too have been Puzzled by Prestbury. Flipside: I like the way FMP now allow us humble users to report transcraption errors.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s