1813 And All That

A comment on Twitter this morning in response to a tweet of mine about George Rose’s Parish Register Act got me thinking.


My original tweet suggested that the post-1813 printed burial registers introduced by Rose’s Act were “somewhat lacking in detail”.

A reply came from another family historian, correctly pointing out that “the printed registers guarantee considerably more detail than ‘buried John Smith’, which is not uncommonly all you get in earlier registers”.

To which I replied that Rose-style burial registers “also (effectively) remove, or at least limit, the possibility of additional details being supplied, such as ‘wife/widow of…’ and ‘son/daughter of…’ or occupational information”.


George Rose by Sir William Beechey.
National Portrait Gallery / Public domain

As I said, it got me thinking. I had this feeling, based on many years of using Church of England burial registers, pre- and post-Rose, that the earlier free-style registers were, by virtue of the lack of any prescribed content, often more informative than the printed ones that followed. After all, the clerks had a blank page in front of them and they could record as much or as little detail as they wanted to.

I notice that I wrote ‘often more informative’ but perhaps I meant ‘sometimes more informative’ – neither of which, I accept, are particularly helpful or scientific terms to use. So, could I quantify this in some way? Well, yes I could. I could carry out an academic study of the registers and write a paper called something like, ‘The impact of Rose’s Parochial Registers Act on the recording of genealogical information in Church of England parish registers’. Which, now I come to think about it, sounds like a pretty good topic for an undergraduate dissertation!

However, as you may have noticed, I’m not an undergraduate (or indeed any type of student) and much as I’d love to spend months carrying out this kind of detailed research, I don’t have, what I believe business types would call the ‘bandwidth’ – bills needing to be paid, and all that…

But, it occurred to me that I could look at a small sample, which might give me a better feeling for what happened when the printed Rose-style registers came into use in 1813. So, with the full and complete knowledge that what I was doing fell considerably short of any acceptable academic standards and that the number of registers I was planning to check was anything but statistically significant, I thought I’d give it a go.

I decided to choose 20 parishes from different parts of the country and compare the information recorded in the burial registers pre- and post-1813. The parishes were chosen entirely at random, simply by going to the bookmarks in my browser and choosing one parish from each of 20 databases. I wanted to have a good spread across the country and I wanted to have a mix of urban and rural parishes. This is the list that I came up with:

  • St Alphege, Greenwich
  • Alfreton, Derbyshire
  • Longbredy, Dorset
  • Ampney St Peter, Gloucestershire
  • St Nicholas, Liverpool
  • St Mary, Eccles, Lancashire
  • Formby, Lancashire
  • Harrington, Northamptonshire
  • Dorchester, Oxfordshire
  • St Mary, Bridgwater, Somerset
  • Godalming, Surrey
  • St Martin, Birmingham
  • Lacock, Wiltshire
  • Batley, Yorkshire
  • St Giles, Cheadle, Staffordshire
  • Wainfleet, Lincolnshire
  • Ratby, Leicestershire
  • St Dunstan, Canterbury. Kent
  • Oakford, Devon
  • St Michael, Macclesfield, Cheshire

The only other rules that I imposed were that the parish had to have been in existence before 1800, that I had to be able to see digital images of the registers and that once I’d selected a particular parish I stuck with it, whatever I saw when I viewed the actual registers.

Next I came up with a list of seven pieces of data which I considered to be the most likely to be recorded in the registers, which, naturally, included the four pieces of information prescribed by Rose.

  • Date of burial
  • Name of deceased
  • Relationship
  • Residence
  • Age
  • Occupation
  • Date of death

When I set out, I wasn’t quite sure how I was going to record the data that I gathered but after a bit of trial and error I realised that while some of the fields could be completed with a binary ‘yes’ or ‘no’, others demanded a degree of subtlety. I didn’t want to over complicate things so I invented some rules, on the hoof as it were (remember, this is NOT an academic paper!).

The first thing I realised was that choosing a simple comparison of the relevant registers for the years 1812 and 1813 wasn’t going to work. There was an obvious and discernible tendency for clerks to begin their 1813 registers full of good intent, recording all sorts of additional detail but after a year or two, or in some cases, a month or two, the effort involved would prove too much and the additional detail would be dropped (we could call this the ‘1855 Scottish Civil Registration Effect’). So I decided instead to compare the registers for 1810 and 1820. This would ensure firstly that the results weren’t influenced by the possible impact of clerks starting to record what they knew was coming before the new system came into effect and also that any cases of an atypical over keen-ness on the part of the clerks in the early, post-1813 registers wouldn’t skew the figures.

I considered four of the fields to fit into the binary category; date of burial, name of deceased, age and date of death. Of course, the reality is that each of these could be subject to some degree of fuzziness. Dates may not always have been fully recorded, first names or surnames could be missing if the clerk wasn’t sure (for example with the burial of an ‘unknown stranger’ or entries reading ‘Widow Smith’ or something similar) and ages might include terms such as ‘Infant’ or ‘Aged Woman’. In these cases, I felt that what was important, and what I should be recording, was evidence of an intent to record names and ages, even if, not every entry in the data sample strictly complied with my ‘yes/no’ approach to recording. It’s worth noting that I can’t envisage any situation in which a register would fail to earn a ‘yes’ for the date of burial and the name of the deceased.

The other three fields had the potential to be more complex but I was eventually able to fit each of them into a ‘yes/no/maybe’ structure and I recorded the data with the following caveats in mind.

Relationship: I recorded this as ‘yes’ if details of children’s parents and the name of a married woman or widow’s husband were routinely recorded, and ‘no’ if these details never appeared. Everything else was marked as ‘in between’.

Residence: This was a tricky one. I only recorded a ‘yes’ where full addresses (i.e. street/house names) were consistently given. Those occasions where some sort of address (i.e. the name of the parish or of a township/hamlet within the parish) was usually given went into the ‘in between’ box. A simple ‘no’ was recorded when an address of some sort was never, or only rarely shown.

Occupation: Here I only recorded a ‘yes’ where occupations were routinely given for adult males and a ‘no’ where occupations were never, or only very rarely given. Burials of clergymen and local bigwigs are almost always denoted in the registers, before and after 1813 but I don’t see this as constituting the recording of occupations as such. In my sample there were very few instances of occupations being recorded and as it happened, I didn’t have to worry about any non-binary options.

Once I’d gathered the data, I converted all the yeses to 1s, the nos to 0s and everything else to 0.5s so that I could analyse the data in a spreadsheet.

With seven pieces of data to assess and 20 parishes for each of two sample years, the total for each year was out of a possible score of 140. This is what I found:

Year Score
1810 67.5
1820 79.5

Regardless of the unscientific approach, it was clear then that there was an observable improvement in the amount of data recorded, post-1813.

However, the detail threw up a few interesting points (which I would love to explore further!). The two areas which showed the biggest improvement from 1810 to 1820 were, perhaps unsurprisingly, the residence and age fields – each, one of Rose’s four prescribed fields. The residence scored just 4 in 1810 and 12.5 in 1820, while the age went from 10 to 20. It’s still interesting to note that 50% of the 1810 registers routinely recorded ages.

When it comes to relationships however, we see things going in the opposite direction. The score was 10.5 for this field in 1810 but it dropped to 6 in 1820. And, although the data really can’t be said to be statistically significant here, occupations were routinely recorded in two of the 1810 registers but only once in 1820. For what it’s worth, the only instance of dates of death being recorded was in the 1810 register of St Nicholas, Liverpool, which also has the honour of being the only register which scored seven out of seven.

1810-St Nicholas, Liverpool burial register

St Nicholas, Liverpool burial register, 1810
Liverpool Record Office; Liverpool, England; Reference Number: 283 NIC/1/9A via, Ancestry.co.uk

Looking at the scores for each of the parishes we find that 14 had a higher score in 1820 than 1810, two stayed the same, while four of them (Liverpool, Harrington, Dorchester and Canterbury) recorded less detail in 1820 than they had in 1810.

So, while the scores show an increase overall in the amount of data recorded, it’s not as simple as the headline figure suggests, and there is clear evidence that certain details which were often recorded before 1813 were less likely to turn up in the Rose-style registers.

One area that I have explored a bit further is the question of urban v. rural areas. If you have any experience of using burial registers you’ll have noticed that you’re more likely to get useful genealogical data from the big urban parish registers and this is borne out by the figures from my sample. I categorised the seven parishes listed below as ‘urban’ and the remainder as ‘rural’:

  • St Alphege, Greenwich
  • St Nicholas, Liverpool
  • St Mary, Eccles, Lancashire
  • St Martin, Birmingham
  • Batley, Yorkshire*
  • St Dunstan, Canterbury. Kent
  • St Michael, Macclesfield, Cheshire

(*Not sure I should have included Batley here – was it urban in 1810/1820? Probably not…)

I added up the scores for each parish in 1810 and in 1820 and came up with an average score for the urban parishes and one for the rural parishes in each year:

Year Urban Rural
1810 4.14 2.96
1820 4.43 3.73

It’s clear that the biggest change is in the rural parishes, so we can theorise that Rose’s Act had the effect of ‘narrowing the gap’ and bringing the amount of detail recorded in rural parishes closer to some sort of equity. The impact in urban areas is less marked with only a slight overall improvement.

So, what have we learnt from all this? The figures suggest that Rose’s Act had a positive impact on the quantity of data recorded in burial registers but we can also see that the printed registers could actually have a detrimental impact, particularly when it came to the not-uncommon pre-1813 practice of recording some sort of relationship information about the deceased.

Perhaps the best way to take this further would be to come up with a useful method of weighting the data. Names, ages and details of relationships are clearly more valuable, genealogically speaking, than fields such as residences, occupations or dates of death (all of which are, of course, important).

What about baptismal registers? What would a similar survey reveal about the impact of Rose’s Act here? Ah well, there’s always next weekend…

© David Annal, Lifelines Research, 4 July 2020

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

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 )

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