A couple of weeks ago, I was searching for some baptisms on Ancestry’s London Parish Register Database, when I noticed something odd. I’d just done a basic search for anyone with the surname ROGERS and the parents’ names William and Mary, within five years either side of 1800 – and it didn’t produce any results.
It’s the sort of search that I do several times a day and I was surprised that it had drawn a blank. Particularly as, in this case, I knew there were some records that should have come up – records that I had previously found on this same database. After checking that I hadn’t included any rogue data in the search I tried a few other similar searches and got the same (lack of) results. I was confused.
So I tried a different approach. I moved the focus of the search from Birth to ‘Any Event’ and restricted the Record Type to Baptism.
And this is what I got…
So, what was going on?
I tried some searches in some other databases and the post-1812 London Baptisms Database quickly threw up something enlightening. Searching on a Birth date of 1820 +/- 5 years, I got just three results:
But the same search using ‘Any Event’ produced 36 hits – including the three above.
So what was different about those three?
You may have worked it out by now – I just about had! The three St Giles in the Fields entries included the actual dates of birth as well as the date of baptism. So what the Ancestry database was doing when I was searching using the date of birth was filtering out the 33 results which only recorded the date of baptism and didn’t show a date of birth. It’s almost as if it was saying ‘we don’t know when those 33 were born so we’re not going to show them in a list of births…’
I tweeted about this a couple of times, tagging Ancestry in the hope that they might see it and respond. They haven’t so far. Why am I not surprised…?
What I was more concerned about was that several people who did respond didn’t seem to grasp the enormity of this. 280 characters isn’t really enough to get the subtleties of a point like this across, so I decided a short blog post might do the trick instead.
It’s around 20 years now since the first major commercial databases appeared on the scene and changed the way that we do things – mostly for the better! One of the firm principles established early on was that you could search across a variety of collections using a single interface. This allowed you to search on a number of different data items at the same time, but the cornerstone of it all was that you could search the records looking for details of an ancestor’s birth, marriage or death.
Of course, in most cases, before the introduction of civil registration anyway, what we’re actually looking at – the records that the databases are guiding us towards – are records of baptisms, marriages and burials. But that’s fine. Research has shown that the gap between birth and baptism gradually increased from the 16th century when parents were expected to baptise their children on ‘the Sunday, or other Holy day next after the child be borne’ to the early-19th century when 75% of children were baptised within 64 days of birth. (For obvious reasons, burials have always taken place with days of death so the death v burial question isn’t really an issue as far as searching for records is concerned.)
So even towards the end of our pre-1837 period of interest, the vast majority of children were being baptised before they were a couple of months old and it’s not unreasonable to conclude from these figures that well over 90% would have been baptised within a year. That’s certainly the impression I have from more than 40 years of looking at these documents.
We can see then that baptisms of adults and older children were the exception and it’s therefore quite acceptable for the commercial websites to have equated birth with baptism when it came to developing their databases.
Naturally, we need to allow for the possibility that our ancestor had a non-infant baptism. We’ve all come across examples of people being baptised as adults and there’s no doubt that these can be more of a challenge to find – particularly for the inexperienced researcher. But to treat baptism as a separate event for search purposes or, even worse, to make the date of baptism the primary focus of the search, really doesn’t help anyone.
Ancestry have only changed this recently and my hope (and my suspicion too) is that it’s happened accidentally; that someone has disconnected the link that tells the system that a baptism on 13 February 1792 should be treated for search purposes as a birth about 1792, and that the switch can quickly be flicked back into its original position.
It certainly hasn’t happened on all of Ancestry’s parish register databases. The West Yorkshire Database for example seems to produce the same results whether you use Birth or Any Event (restricted to baptism).
Our research should be much more than just collecting names and dates but without the key facts – the borns, marrieds and dieds – we’re always going to struggle to tell the stories. We need the basic structure – the timelines – to allow us to see how our ancestors’ lives chimed with what was going on around them.
We rely on the databases to help us to identify these facts and when something like this happens it just makes it that much harder. How many thousands of unsuccessful searches have been carried out over the past few weeks? How many times have people looked for a record of their London ancestor’s birth and found nothing – even though a record of their baptism is right there on the system? And how long will it be until Ancestry do something about it…
© David Annal, Lifelines Research, 17 July 2021