Lancashire: Images of register only capture half of information for all records
UK, England, Lancashire—Nonconformist Church Records, 1647–1996 [Part B][MSG4-RZ7]
I am reviewing some of the above records. I am seeing the same indexed images more than once. The batch of images have been taken of a rather wide register, so only one page and (maybe) a half appears in the image. The extent of the coverage in any one image varies somewhat ie 1 page and a third for some images.
None of the records are complete unless one opens the preceding or following image by using the reference images in the box in the right hand side. FS provides guidance for a single record crossing more than one image e.g. starts bottom of one image but complete on the next etc, and those are reasonably clear to manage. In this instance though none of the records are complete in themselves, in any of the images.
- Some indexers attempt to transcribe simply the image that contains the name and date and omit further information or include 'Lancashire' as the location based on the title of the batch at the top of the screen.
- Some indexers attempt to transcribe the image containing the date, age and location, omitting names.
- Some indexers reject both images as having no extractable data.
- Some indexers try to merge the 2 relating images in an attempt to fill in all the boxes for all records in that image. This can work (technically breaches guidance), although on occasions there is an obvious difficulty in lining up the relating information and errors may be evident.
In most cases of attempts to index these images the location box contains Lancashire ie from the title of the batch, rather than either omitting any location name, or inserting Lancashire while failing to see the word "York" in the county column (sometimes present) in the partial reference image.
The image I include below, is one of the better ones, so you are not seeing those occasions where I am wondering what on earth to do. On this occasion I think the transcriber is right to reject their allocation image on the basis (perhaps) that the another indexer will receive the referenced image and do the right thing! But some of these images, as ever, are better than others and really you do need to see both sometimes to translate what information is there or not there.
For those occasions then where the two images really need to be seen together - what to do? As I explain above different indexers are applying different approaches. I am seeing some pages totally rejected, while I am also being faced with the same pages time and again. Given that partial images do often contain some usable data it is technically untrue to reject it out of hand.
On occasions, I have also noticed, by scrolling through the reference box, some of the original images have been repeated as if the person copying the originals was aware of the issue and tried to compensate. This explains why I notice that I am seeing several transcripted versions.
So, which would be FS preferred approach? Index each image (what is visible in that image) on the basis that, however incomplete, each image has some usable data that is retained though with the risk too of duplication etc.? Or, in the case of the images that miss out the names reject those images as having no usable data?
The image below shows the current image I am reviewing. It is not the best example to demonstrate the conundrum I am sometimes facing. Please do not say simply say: 'ok reject the core image and assume someone will do the RHS'. It is not always as clear a divide, in terms of visible and readable data, as the particular image below suggests. I do not wish usable information to go astray.
Comments welcomed. Apologies for length of this query.