To view this media, you will require Adobe Flash 9 or higher and must have Javascript enabled.

Duration 00:50:48

Brick walls and lost ancestors

Even the luckiest of family historians will come up against brick walls from time to time. Dave Annal shows that everyone left a paper trail and that it’s almost always possible to find it. He introduces a number of strategies to help you get around the brick walls and dead ends in your research, paying particular attention to getting the most out of online databases and advanced techniques such as ‘family reconstruction’.

Dave Annal worked for The National Archives for ten years, at the Family Records Centre and at Kew. He now runs Lifelines Research where he is frequently called upon by clients to help solve their trickier problems. Dave is the author of a several family history books, notably the second edition of his bestselling beginners’ guide Easy Family History. He is a resident expert on Your Family History magazine and regular speaker on the family history circuit.


I’m going to talk this afternoon about brick walls and lost ancestors and how to go about solving family history problems. I don’t, I hasten to add, have a magic wand. So unfortunately if that’s what you were hoping for, I’m not going to be able to solve all your problems just like that. But hopefully what I’ll do is I’ll give you some pointers. I’ll give you some ideas, make some suggestions in some areas that you might not have thought of or just some approaches to the problem that you might not have thought of.

I’ll just introduce you to this little group of people here. This is my grandfather here, and this is the person that everyone has in a family, that is, the great aunt that everyone says, when you start asking questions, ‘oh, if only you’d asked her before she died’ – which is very helpful, isn’t it? Yes. But apparently she knew everything about the family, and she was the person to speak to. But when I knew her – I think she died when I was about ten or 11 – I wasn’t interested in family history. I didn’t ask all the questions then, so unfortunately a lot of information died with her.

I’ll explain about that little place later on. I’ll just mention while I’m here and have the opportunity to do so, just mention my latest book. This is the new edition of Easy Family History which is a beginners guide but hopefully it’s also useful for people who have a fair amount of family history research. The first edition was produced in 2007, sold about 10,000 copies. This one has just come out this year, hot off the press. They have plenty of copies in the bookshop downstairs. If anyone wanted to buy one, I would be very happy to sign it for them later on if that’s what you’d like to do.

Right. Family history is easy these days, isn’t it? You just log on to a website, type in your name – bang, there you go. That’s it. It’s all done, isn’t it? Family history just pops up on the screen in front of you, all the names, everything. Yeah? No, no, no. Well, that’s what they might have you believe, but of course this isn’t true. And if you’ve been researching your family history for ten years or ten weeks, you are almost certain already to have come up against a problem. You’re almost certain to have experienced that dreadful moment when the search that you were expecting to bear fruit bears absolutely nothing.

In the old days that might have been going through the GRO’s indexes to births one after the other and coming to the right page, and there you go. You open up the page, and there the entry isn’t. Or you might have been going through microfilms of census returns and eventually you come to the village, and there you go. And you look, go all the way through the village where you’re expecting to find them, and they’re not there.

In the digital age it’s more likely that your unsuccessful search will lead you to the dreaded ‘no results’ screen on your computer, probably the most popular page in family history because everyone seems to look at it all the time. So either way you’re stopped in your stride. You feel like you’ve stepped on as stair that isn’t there.

You might try a few other ideas and quickly solve the problem. But what if you don’t? What if nothing that you try works out? What if months or even years later you still don’t have the answer to your problem? What if the person you’re looking for seems to have appeared from thin air, seems to have arrived on earth as a fully formed adult with no past history? I’m sure we’ve all experienced that.

Welcome to the world of family history brick walls. Now I say I hope in the next hour or so to convince you that with a bit of effort, with a bit of luck, and with a lot of lateral thinking, it’s almost always possible to work your way around a brick wall, problems are usually solvable and that almost everyone left a paper trail, however obscure that paper trail might be. I also hope to convince you that brick walls are what makes our hobby so fascinating. I mean we wouldn’t really want it all handed to us on a plate. We wouldn’t want to just type in the name and get everything because it’s the thrill of the chase, isn’t it?

Of course you want to solve the problem, but part of the fun is the intellectual exercise of trying to solve it, trying to battle the documents and come up with the person you’re looking for. I’ve got a family history brick wall I’ve been working on for about 20 years. And it’s a problem that I come back to frequently. Every time a new database is released I, in expectation, type in the details and try all sorts of searches. And so far nothing has come up.

But it’s probably the part of my family history research that I’m most interested in. So so far I haven’t solved it, and ‘so far’ are the key words, really, because a brick wall really is a problem that you haven’t solved yet. That’s all it is.

So the first thing to think about is: is your problem, is what you’re stuck with actually a brick wall? Or is it just a little annoyance? You can’t find a census return. You can’t find someone’s birth certificate, but it’s all right because you’ve got the parents’ marriage so you can work back. It’s annoying that you’ve got this little gap in your research, but it’s not really a brick wall. You’ve managed to work past it. So, for example, you can’t find someone in a census return. It’s disappointing because ideally you want to find your people in every single census return because you never know what it might tell you. So yes, you want to find them. But it’s not really a brick wall.

So I really think it’s very important to understand – (that’s all right, ladies, please come in, thank you) – I think it’s really important when you’re searching family history to have an understanding of your sources to understand what it is you’re looking at. And in the digital age when we search for a name and up comes a page on the screen, it’s very difficult to lose the feeling of context, not to understand what it is we’re looking at. We’re looking at a page. And there, if we’re lucky, there is our ancestor right in the middle of it.

But how did that name get there? What happened? Why was it produced? Why did people take census returns? What was it all about? So just for example, I don’t have time to go in detail this approach, but you really need to understand, particularly when we’re looking at census returns, that what you’re looking at when you’re searching an online census database is, at best, third hand information.

The information that we’re looking at was originally created by the householder who had been given a schedule to fill in at the time of the census itself. So that was the very first thing. The enumerator or an army of enumerators around the country would deliver the forms. The householder, if they were literate, would complete them. If not, they probably knew someone down the road who was literate or they had an aunt or an uncle, a niece or a nephew who could read and write. They’d be able to get the information down somewhere. If all else failed, the enumerator would write down the information for them themselves.

So there we’ve got the first piece of information. It’s written by the householder on the schedule. But then the enumerator goes and collects all the schedules, copies the information from the schedules into the summary books. And that’s what we’re looking at online is the summary books, not the original householder schedules. So you can imagine there is lots of room there for errors to crop up.

So the householder has written his name. He’s written occupation and he’s written his place of birth. And the enumerator, in most cases, is able to read that, and that’s fine. Writes it down, copies it accurately, everything is OK.

But what if he couldn’t read the surname? It wasn’t a local name that he was familiar with. Or he couldn’t read the place of birth. It was a village in a county the other end of the country; he didn’t know it. Well, he’s got two choices. He can go back to the householder and ask them what did this say. What does this say here? And I’m sure that lots of conscientious enumerators did exactly that. But I’m equally sure that a lot of them thought: I’ll just write down what it looks like. Who is ever going to look at this? They’re not going to know. Doesn’t matter. So it’s not really important.

So they’ll write down their best possible guess of what the word was. There’s going to be surnames and places of birth normally that are the problem. First names, there’s a relatively small set of first names to choose from. Surnames, lots of them. Places of birth, lots of them. Counties, they’ve probably read that accurately.

So there we go. So there are already lots of areas for the information to have been corrupted. Now of course, 100-150 years later, the information has been transcribed by commercial companies, and they have created databases that we’ve searched. Now again with the best will in the world the people who have done this work have done generally a good job. But again they are going to have made mistakes. We all know. We’ve all looked at census returns. We’ve seen things and we think how could anyone have possibly have read X as Y, for example.

So inevitably, however good these people are, however conscientious they are, more mistakes are going to be introduced at the stage of the transcription. So we’ve got the householder schedule. We’ve got the enumerator’s attempt at transcribing that. And then we’ve got the modern transcriber’s attempt.

And this is of course assuming that the householder got the information right themselves in the first place. And there is any number of reasons why they may not have known the true information, or they may not have cared. They may deliberately have lied. So lots and lots of ways that that information that we’re searching is not going to be accurate. And if we take it too literally, if we put in everything and we say we know this person was born in this year, we know this person was born in this place, and put in that information strictly, we’re going to miss loads and loads of our ancestors.

So in most cases the failure to find someone in the census return is going to be explained by a simple mis-transcription. That’s inevitably it. And the best way to approach this is to apply the ‘less is more’ technique. We’ve got this search screen here. Actually, I tell you what I’m going to do. I’ll show it to you on here. I’ll show it to you on this one here. This is the old search screen which you can still access on Ancestry and it’s, I think, a much better search than the new one they introduced a couple of years ago.

But we’ve got all these boxes, and the temptation is to fill them all in, isn’t it? For the inexperienced researcher you think: I know her first name. I know her second name. I know where she was born. I know where she was living. I’m going to put all that information in. But if you do that and just one of the pieces of information that you’ve put in is wrong, you’re not going to get the person that you’re looking for.

So here is a case that I worked on years and years ago where they were looking for a family whose surname was Blissett. Tried lots of searches. Tried various things. They knew in particular that they had a daughter called Ada, and they knew that she was born about 1869 in St Giles in the Fields in London. So, having tried lots of different searches, I came to this one where all I put in was the first name, Ada, the year of birth plus or minus a year, and in the parish or place of birth just the word Giles – so anything that says St Giles spelt fully Saint or just St Giles or anything would come up.

So I’ve left the key information, the surname, out because the surname is the bit that’s most likely to have been mis-transcribed. And running that search, I got those results there. Now remember the person we’re looking for is Ada Blissett. Now the eagle-eyed amongst you might already have spotted the person we’re looking for: Ada Blipell – that famous old English name Blipell that we all know!

This is the person we’re looking for. And what we’re looking for is the shape of the name because remember what I said about mis-transcription. It’s not talking about a variant of the name. That’s not a genuine variant of the name Blissett. It’s just a mis-transcription, a misreading of it. So you leave out the surname and then you look at the list of results, and you hope that one of them looks like the right shape. You hope that one of them jumps out at you. And in that case you look at a name and it’s obviously not a real name. It’s not a surname in English use.

But before we’re too critical of the transcriber, let’s just have a look at how that surname was actually written there. It’s the double s. It’s the old cursive double s, that’s right. And it’s been read as a p. And I’ve got to say I can’t criticise them for putting double l at the end because they’ve not crossed the ts: that has not been crossed. So I think Blissell would have been a very good transcription of that surname. Blissett would actually have been an incorrect transcription. Although it was correct, it would have been wrong, if you see what I mean.

So there we go. If you have a look at it, as you say, it’s the double s that has been transcribed as a p and then the double l. But it’s the same shape as the name we’re looking for. And that’s the crucial thing. We’re looking for shapes of names. So that was just a basic simple one. You need to understand that ‘less is more’ technique. Do not be tempted to fill in the boxes. The more boxes you fill in, the less chance you have of finding the person you’re looking for. And it seems counterintuitive, that. It doesn’t seem to make sense, but that is the logic for it because you’re applying a filter. The more you filter out, the less you will get at the end. So try a variety of things.

Right. Let’s have a look at a case study. This is something that I actually helped someone with about – it must be over ten years ago. It must be because it was before the 1911 census was out. I think it might have been just after the 1901 census came out. We were looking for someone called James Edward Sharpe, and they knew quite a bit of information about him because they’d found a record of his baptism which told us that he was born on 3 April 1903 – this was family knowledge that he was born on that date in Hull – and that his parents were James Sharpe and Edith, but they didn’t know Edith’s surname.

And of course on the record of baptism from the parish church it wouldn’t give that information. It would just give her first name, Edith. He had no siblings. This was known by the family that there were no siblings. It was just James. And so the baptism was found at the local parish church.

Now we can find James quite easily in the 1911 census. Let’s just go back a few pages. Right, there we are. There’s James in the 1911 census. Now at the time that I did this search, we didn’t have this. We didn’t have the 1911 census. But there is James and Edith, the parents, and there is James Edward, the son, aged eight. So that ties in with the birth on 1 April 1803 [1903] because this was taken on 2 April. So it was the day after his eighth birthday. And it confirms he was born in Hull and apparently so were his parents both born in Hull as well.

It also tells us that James and Edith had been married for 11 years. That would all look pretty straightforward. If you found that information you’d think, there we go. I’ve got everything I need to find James’ birth and to find his parents’ marriage. The problem is there is no birth recorded for a James Edward Sharpe, and there is no marriage of a James Sharpe and an Edith, whether we have the e on the end of Sharpe or not, by the way, just if you’re spotting that.

So a bit of a problem. No record of James Edward Sharpe’s birth. Now you might think, well, it wasn’t registered. That might be your conclusion. And one thing I will hope you to get out of today if nothing else is that you will leave this room not thinking of that as a very likely answer. It is low on your list of solutions, that a birth wasn’t registered – low, OK? Particularly as recently – and I use that in a family history sort of term – as recently as 1903.

So we’ve got a bit of a problem. We could actually – I was talking about whether it’s a brick wall or not. In terms of the Sharpe family it’s not because James Sharpe, the father, you can find his birth pretty quickly, pretty easily, and you could work back with the Sharpe family. But what about Edith? We’ve got a real problem with Edith. We don’t know her name. Normally if you can’t find the birth of one of the children, you look for the birth of one of the other children [and] you find the mother’s maiden name. Or you look for a marriage [and] you find the mother’s maiden name. But here there are no siblings and there is no marriage of the parents.

So there is a brick wall. How do you crack the Edith Sharpe problem? How do you work back with that? Not easily, but it was possible. We did manage to crack it. The first thing I did, or one of the first things, I did a search on FreeBMD [] for anyone called James Edward who was born in Hull or Scullcoat, which is part of Hull as well – this is a bit of local knowledge – around 1903. And in fact because we could be pretty sure he was born on 1 April 1903, you can really limit that search to the June quarter of 1903. If the birth was registered, it should be registered in the June quarter of 1903.

And if you do that search, you just get two records. So I’ve left the surname out, and all you have is James Edward Colvert and James Edward Williamson. Now if you’re flush with money, just order both those certificates and see whether one of them could be the right one. You could do that. But you can also be a little bit clever because if you do a search in the 1891 census, we know – this is them in the 1901 census; this is James and Edith before James Edward is born – and they’re living in Bank Court in Hull as James, aged 25, Edith, aged 26. Edith, rather than saying Hull this time, Edith is said to be born in Beverley.

So what I did, I went to the 1891 census and I said show me anyone called Edith who was born in Beverley round about 1875 plus or minus a year. And there were 22 of them. That’s quite a lot, but remember the two names that we had from the FreeBMD search? Colvert and Williamson were the two names. Well, there’s no Edith Colvert, but there is an Edith Williamson, born in Beverley, around about the right time. And if you look for that Edith in 1901, you don’t find her. There is no – she’s either married or died. But what if that Edith is our Edith and the fact that her child was born in 1903 and registered under the name Williamson [was] because the parents weren’t actually married?

So that would give me, then, the confidence to buy that birth certificate of James Edward Williamson just to see what does it say. Can we prove that it’s the right one? And you’ve guessed already. We can prove it’s the right one because there he is, 3 April – sorry, 3 April 1903, 2 Banks Court, Blackfriars, James Edward, son of Edith Williamson of no occupation, no father’s name. It’s him. There’s no doubt that is the person we’re looking for.

Now why when they got the child baptised they claimed to be married, and yet when they went to get the birth registered she was honest and said no, I’m not married – can’t put the father’s name there – I don’t know. You’d have thought it would be the other way round, that in the church, with God looking down on you, you might be a little bit more truthful, but in the register office you could tell lies not worrying about being struck down. Don’t know. But there we go. That’s him. That’s the same address they were living at in the 1901 census. It’s the exact date of birth that we had. And it matches the census information.

So by leaving the surname out we can find something like that. Now surnames are what we cling to because surnames are what we, for most of us, probably got us started on family history. Where did that name come from? But surnames are the things that are actually most transient for our ancestors. Their first names, they’re probably going to go through their life with the same first name. But there is any number of reasons why they might use more than one surname during their life.

And there is a case in point. So the parents weren’t married, but they lived as if they were married. And so he’s registered as Williamson, but he’s always known as Sharpe, and the family always knew that he was Sharpe. There was nothing to suggest that he wasn’t. And they were unlikely to have told children, grandchildren, nieces and nephews etc.

So I’m now going to introduce you to the most confusing case I ever worked on in my life, this one. Sounded quite straightforward. We’re looking for someone called Catherine or Kate Hay, and the family, again, this is crucial – this is really important – they knew the date of birth, 14 October 1874. That was something that had been handed down in the family. And that’s gold dust. A date of birth is gold dust.

You can probably question that. It might be a year or two out. But the date of birth, the actual month and the day, are unlikely to be wrong. Why would you lie about what day and month you were born on? You might not know about the year, but that’s gold dust, 14 October. So knowing that, just like we did in the case with James Edward Sharpe, knowing that was really important.

She had a daughter called Edith Florence Wells born 13 November 1899 in Brighton, and the father on the birth certificate was William George Wells. But in fact, as we find out through doing some research, the father was actually called William George Radley. I’ll come back to that and just explain a bit about that.

In the 1901 census we’ve got her – and I’ll show you that in a minute – aged 25, born in King’s Cross. In 1911 she says she’s born, aged 34, born in Clerkenwell. Let’s have a look at those records. So here’s the family. Here’s Kate Wells. She’s a visitor. She’s down as a visitor, aged 25, in this house in – where are we? We’re in Hammersmith. There we go. And she’s 25 and she says she’s born in London, King’s Cross. And there is the daughter, Edith Florence, who was registered in 1899 in Brighton.

What you might notice as well is that here is William Radley, William George Radley, as he was. And he’s married to a woman called Florence. And Florence has got two children there, John N – John Nathaniel, I think it is – and Florence Edith who is aged one and born in Brighton. They’re not the same person. They’re different people. Their births were registered within a few weeks of each other, and they’ve got the same father. Different mothers. The same father. And they’re all living together in the same house there. It’s very, very strange because this, the person who claimed, when she registered the birth of Edith and said that the father was William George Wells, he was actually William George Radley. And that’s him there with his wife and other children.

Now of course William actually isn’t married to either of those women. He’s married to another woman. It just gets more and more complicated. He says he’s born in Birmingham. He’s not born in Birmingham. He’s born in Bermondsey. That’s a different thing. That’s a different matter. But there we go. So we’ve got some information about Kate, born in King’s Cross. And if you look at the 1911 census, she’s now down as Kate Radley. And there’s Edith, the daughter.

I don’t know what happened to the other ‘wife’. I don’t know. I haven’t found out what happened to here. But here’s William now with Kate, and Edith, the daughter, is already born in Brighton, and then a long run of other children. And they then go to Canada, and that’s where my client contacted me, from Canada. And Kate here says she’s born in Clerkenwell. So that’s consistent with King’s Cross. We’re happy with that. And the age is one year out. It’s 25, and then ten years later it’s 34. But that’s not bad.

Now of course you’ll have guessed already there is no birth for a Katherine or Kate Hay or anything similar, any variance of Hay, anywhere around 1875 plus or minus a couple of years. So I tried looking for Kate in the census before 1901 to see if I could find her as a teenager, as a young girl in the 1891 and 1881 census. Couldn’t find anything in 1891, but eventually found this in 1881. You’ll see at the bottom of this page this actually says E Hay. And then on the next page we’ve got the rest of the family: AC Hay, E Hay, B Hay, C Hay, another C Hay, and then E Hay.

Now this C Hay interested me. It was a girl who is aged five in 1881, which would tie in with being 25 in 1901. She’s living in Clerkenwell, and it says she was born in Clerkenwell. So I thought: could that be her? Could I use this information perhaps to find them in the 1891 census and get some more information? So we’ve got a C Hay who is the daughter of an E Hay, and this E Hay very helpfully is born in Barking in Essex. That’s a nice – nothing against or in favour of Barking, but as a place of birth it’s quite distinctive. It’s not just London. It’s not just Shoreditch – somewhere like that.

So on the strength of that I was then able to find Edward Hay, born in Essex – doesn’t say Barking but it’s Essex – and here we have a 15 year-old daughter, Anne C Hay, born in Clerkenwell. So not C. Not Catherine. But Anne C. So I thought: could that be her? Could that be the person we’re looking for? It’s clearly the same family that we’d found in 1881. And this was the big breakthrough because on the strength of that I found this record of a baptism. Anne Catherine Bell – so we’re looking over here – daughter of Edward and Anne Catherine Hay, and look at the date of birth: 14 October 1875. A year out, but that looked very, very promising.

But it didn’t answer the question why couldn’t I find a record of her birth. And again you might say, well, perhaps the birth wasn’t registered. Again, no. That’s a low chance. That is not likely to be the case. The clue came from this strange third name, Bell, because although all the children of Edward and Anne Catherine were baptised as Hay, their births were all registered as Bell. Don’t know. Don’t know why.

I mean I know that Edward was born illegitimately, and his original name was Bell but his mother married someone called Hay. But why he would choose to register all the births under one name and baptise the children under the other, I don’t know. But here is Anne Catherine Bell’s birth certificate: 14 October 1875 in Clerkenwell, daughter of Edward Bell – not Edward Hay – and Anne Catherine Bell, formerly Birch. So I got there eventually, but it took a lot of lateral thinking and, again, treating the surname with disdain because it’s the least fixed thing in someone’s life, their surname.

It is quite tempting to assume the worst, to jump to the conclusion that the birth just wasn’t registered. And it is particularly tempting when it comes to searches for birth certificates in the first few decades of civil registration. Some of you may have heard that before 1875 registration of births wasn’t compulsory. Please disabuse yourself of that. It’s not true. Registration of births was always compulsory.

What changed in 1875 on the introduction of the 1874 Birth and Death Registration Act was that the ultimate responsibility for ensuring that births were registered was transferred from the registrar to the parents. Now that doesn’t mean that before 1875 they didn’t have to register births. It just means that it was up to the enumerator to ensure it happened. But people could see the benefit of registering births, and people wanted to register births, and people did register births.

I’m just going to show you a table. This is a table showing the number of births registered in England and Wales between 1870 and 1880 every year. So here we are, around about the 800,000 mark in 1870 and just a gradual increase as we go through the decade. Now if the 1874 Act, which came into effect in 1875, had had a massive and significant impact, if we can argue that before we have one situation, after we have another – I feel like Peter Snow here; ‘a swing over here’ – before we have one situation, after we have another situation, you would expect a significant increase here. Agreed? If we’re really saying before they didn’t have to, after they did, we would have a significant increase.

Here is what actually happened: slight decrease. Very slight. Not statistically significant, but a slight fall in the number of births registered. And then, carrying on through the decade, just what you’d expect with a growing population. That’s just exactly in line with what you’d expect. There is a slight fall towards the end of the 1870s – nothing statistically significant. The trend is steady and upwards throughout the decade.

This Act made no difference whatsoever to the rate of registration. Before 1875 if people were registering births, they continued to do so afterwards. If they weren’t registering births, they continued not to. OK? That’s the truth. That is the evidence that that is the case. So the next time someone tells you, oh, the reason you haven’t found a birth before 1875 is because they didn’t have to register them, tell them they’re wrong. That is not the explanation. It might be, but it could equally be the explanation after 1875. And it’s a very, very slim chance. It is not the most likely thing. By far more likely case is that the surname is not what you would expect it to be in the registration. OK? I’ll get down off my soap box now.

So we’ve looked at civil registration. We’ve looked a bit at census returns. Quite often the problem we have is when we get before, into the era before civil registration, before census returns. So within four years – the 1841 census, the 1837 Birth, Marriage and Death Registrations start. Before that what we hope is that we’ve got enough information from those sources to lead us back into parish registers, to lead us back into local county sources rather than the national ones. So you hope that your ancestor didn’t die two days before the 1851 census and therefore not tell you where they were born and all you’ve got in the 1841 census is yes, born in the county, or, even worse, no, not born in the county. But which county? Where? You don’t know.

You hope that you’ve got enough information. And the good news is that most people prior to the industrial revolution didn’t move that far. Most people. Obviously the problem is where you have ancestors who just uprooted and ended up in the other end of the country. Then you might have problems. But most people lived in what they called their country. That’s where we get the word countryside from. It’s the area around the market town. And people tended to move around from parish to parish within that area.

I had a problem with a family called Groves who lived in Dorset. And over 20 years ago I was doing this research when there was virtually nothing online. Really there was actually nothing available online. So everything meant going to record offices, looking through microfilms, looking at proper documents, real things. My former colleague Audrey Collins has this wonderful phrase: I’m going to a place to look at a thing. We don’t do a lot of that now in family history research.

Charles Groves, he married a woman called Jane Ellory on 9 March 1801 in the beautiful Dorset parish of Long Bredy. That’s all pretty straightforward. And he was a thatcher. I knew that. And he had children baptised at a variety of parishes in West Dorset. Now what I needed to do is to find out where he came from. The problem was not only could I not find a record of a baptism in Long Bredy or the couple of parishes either side of it, I couldn’t find what happened to him afterwards. He had children, and then he seemed to disappear. Just couldn’t find any trace of him.

I love maps, and I like to plot things on maps to get an idea of how people are moving and where they’re moving. And if we remember what I was saying about the country people lived in, this is good evidence of it. So he married in Long Bredy, which is down there. He then had children baptised – I love these Dorset parish names; they’re beautiful – Toller Fratrum and Wynford Eagle. That’s a joint parish.

So he had children baptised there. He then had a child baptised in Frome Vauchurch, which is in there. And then he had a couple of children in Toller Porcorum up there – so all very much in the same basic area. In December 1811 Jane was buried, also a Toller Porcorum, and then in 1813 Charles remarried, and he married a woman called Anne Laver in the parish of Chilfrome, which is just up there, so again very, very local, and not moving very far away – staying in his country. And then nothing. I couldn’t find anything more of him after that second marriage in 1813.

And so my challenge was (a) to find a record of his baptism but also to find out what happened to him because that might give me a clue. So I don’t know how old he is. I know he married in 1801, and statistically it’s likely that he was in his early to mid-20s when he married, probably closer to the mid-20s and for women closer to the early 20s. That’s just a rule of thumb. So I would assume he was born sometime in the mid 1770s, late 1770s, that sort of time.

So what I decided to do was to just do a sweep of all the parishes in the area and gather every reference I could find to the name Groves. This is something called family reconstruction, and the idea is that you just extract all the references to the family, and you build up little family groups. And you then can track mobility. You can see where the people are moving. You look for naming patterns – certain names that keep cropping up in the family. For some reason the name Levi is very big in the Groves family. There’s lots of Levi Groves.

So I knew the names of his children. Who did he have? He had Jane, he had William, he had Henry, he had Charles, and he had George. So I would expect his father’s name to be one of those. Not definitely but probably. That seemed likely. If you’ve got four sons and one of them is not named after your father, either your father has upset you or you’ve now upset your father, really.

So I spent an enormous amount of time going through census returns, going through parish registers, looking at wills, building up an immense amount of information about the Groves family. This was before things were available online. I have to admit that today this would have taken five minutes. 20 years ago it took five months, I should say, at least.

So I swept through all these documents, and eventually I found some possible baptisms. Sorry, I should just mention the family were also connected to this parish called Hook. They were also connected to Usk as well and miles out of area here down in Preston. They really – a traveller in the family there, definitely, moving all that way. So everything pretty much in that area, and that was the area I concentrated on.

And I found a record in 1765, which is a bit on the early side, of a baptism of a Charles Groves, son of Charles and Elizabeth Groves in West Compton over there. So that looked like a fairly strong candidate, but I didn’t like the fact that it was quite early. I also found another Charles Groves baptism in Portesham down there, and I found one right up there in Melbury Osmond.

The Portesham one was 1756, which really sounded too early. I didn’t like that one at all. The Melbury Osmond one in 1778 is a little bit out of the area. I wasn’t too happy with that, and the father’s name was Richard which wasn’t a name that was connected with the family. So this was my favourite, but I had to keep working because I haven’t proved that it was. And, more importantly, I haven’t proved that it wasn’t the right one.

It took me a long time, but eventually I made a major breakthrough, and it is a breakthrough I would have made much sooner if I’d had access to fully indexed census returns because in the 1851 census – and I’d been the whole way through the Dorset 1851 census – in the 1851 census in the small Somerset village of Hardington Mandeville, just over the border at the top of the map there, I found a Charles Groves, aged 73, whose place of birth was given as ‘Dorset, Milbury’ which must surely have meant Melbury. Most significantly of all he was a thatcher. And a bit more research turned up the clue that I’d been looking for because this Charles eventually died in 1860 at the grand old age of 83, and he left a will. And one of the executors and the chief beneficiaries was his son, William Groves, of Toller Porcorum, Dorset.

Just to set the seal on the whole case I discovered that in 1857 Charles married for the third time when he was 79 years old. He gave his father’s name as Richard Groves, thatcher, deceased – not surprisingly, I suppose – which confirms that the Melbury Osmond one was the right one, that it eventually managed to prove that it was. And I only managed to do that by this process of taking all the information I could find about people spreading out further and further and building up little groups, getting as much information as I could about the family.

Now I have to confess that I did have the West Compton one as my favourite. And I should really have listened to Charles Dickens about that because if any of you are familiar with Great Expectations – and I hope you are – there is a wonderful scene where our hero Pip confronts Mr Jaggers about the identity of his benefactor. And he says:

‘“I have been informed by a person named Abel Magwitch, that he is the benefactor so long unknown to me.”
“That is the man,” said Mr Jaggers…
“And only he?” said I.
“And only he,” said Mr Jaggers.
“I am not so unreasonable, sir, as to think you at all responsible for my mistakes and wrong conclusions; but I always supposed it was Miss Havisham.”
“As you say, Pip,” returned Mr Jaggers… “I am not at all responsible for that.”
“And yet it looked so like it, sir,” I pleaded with a downcast heart.
“Not a particle of evidence, Pip,” said Mr Jaggers… “Take nothing on its looks; take everything on evidence. There’s no better rule.”’

Now that could have been written for family historians, really. Take nothing on its looks. Take everything on evidence. It’s absolutely brilliant. It’s a lovely way of looking at it.

So family reconstruction can work. And that’s a sort of focused approach. That’s where you think, right, what I’m going to do is I’m going to look at particular documents for a particular thing. So I’m going to look in a parish record hoping to find a record of a burial. I’m going to look in a census return hoping to find this person. But the scattergun approach can also work very well, particularly thanks to our friend Google. Google is the most wonderful tool for family historians. Just throw things into Google and see what happens.

Now I use this – I’m not going to do this one; this one here. Thomas Annal – not actually one of my ancestors; I do a one-name study on the name Annal and this was a particular person who troubled me for a long time. He is the ancestor of a large group of Annals who live – large, comparatively speaking; there’s not many of us around – but lived in Yorkshire since the mid 1700s. My family come from Orkney. All the other Annals I know come from either Orkney or Fife apart from this little group who crop up in Yorkshire in the mid 1700s – and not even on the coast in Yorkshire but an inland parish.

First record we have of him is his marriage on 12 November 1741 to a woman called Isabelle Bridgewater. And I found this entry on the IGI [], which is now FamilySearch, the Church of Latter Day Saints database. I found it many, many years ago. He was buried – I looked in the parish register and I found that he was buried on 13 June 1787 in Well. The age wasn’t given in the parish register but, again, looking at a marriage in 1741 we can probably say he would have been born the mid 1710s – 1715, 1720 would have been likely. That’s when you’d expect him to be born.

Here is the record from – it’s actually the bishop’s transcripts of the parish register – Thomas Annal from North Britain, which was the wonderful way they had of describing Scotland, suppressing Scottish nationalism. It’s not Scotland; it’s North Britain. They tried to call Wales West Britain as well. You can imagine it didn’t really take on very well, but they did try it for a while.

So that is an important lesson in itself because the IGI just gives me the date of the marriage and the names. The actual register itself gives me this extra crucial piece of information that tells me he is from Scotland. So that’s really important. Never just rely on an index. Always look at the original. It might say so and so ‘of Brick Kiln Cottage’, which means you know which family it is. It’s not the Tithe Farm ones. It’s Brick Kiln Cottage. It might say a burial of someone ‘as an infant’, so you know that it’s not your great-great-grandfather because this child is an infant. It might say ‘the wife of’. It might say ‘the daughter of’. There might be extra information in the register that is not in the index that you’re looking at.

So we know then that Thomas Annal came from North Britain, from Scotland. And looking back at the FamilySearch website, or the IGI as it was then, I found a couple of candidates. I found one in St Andrews in Fife, but I was pretty sure it wasn’t him because I was pretty sure I’d found a marriage for that Thomas Annal in St Andrews later on.

But I found another one who was baptised in a parish called Anstruther Easter – I don’t know if any of you know the Fife coast around Crail; it’s a beautiful part of the world – and I thought he was the most likely –1715, just the right sort of age – the most likely one. But how could I prove it? How could I prove that this Thomas Annal who came from Anstruther Easter, born there in 1715 was the same one who turned up in this little village in North Yorkshire in 1741?

And I was stuck at that point for years and years and years until I tried a Google search. Now this branch of the Annal family normally spell the name with a double l at the end. So what I often say to people is just try putting in the name you’re looking at together with the place that they came from. It’s quite a good thing to do. Just put in the name and the place.

Now with my brick wall that I’m working on, which is a different family, the surname is Port and they came from London. Now putting in Port and London into Google is not going to get you any particularly useful results. But I tried it with this person here: Thomas Annal, Well. And I came across this reference here. Now this is some work that the Fife Family History Society have done, a very good bunch of people who have got together and they have indexed all these records, these deeds from 1715 to 1809.

And if we scroll down to the name Annal, we find this: Annal, Thomas, shoemaker in Well, county of Yorkshire, only son in life of deceased Thomas Annal, shoemaker in Anstruther Easter. And that was it. You probably heard me, my shriek of delight when I found this. It was magnificent. If I wanted to make up a document that proved that I had found the right person, that would be it. That would be what it would say.

And it was not thinking – I would never have thought, oh, I know, I’ll go through these deeds just in case there’s something. They’re a big bundle of things that haven’t been looked at for a hundred years. I would never have thought of looking in these. It would have been a speculative search at best. I wouldn’t have expected any reward from it, so I wouldn’t have done it. But there it is. There is the information. And it was simply because these kind people at the Fife Family History Society had indexed them and that they had been put online so that a Google search just brought them up.

It wasn’t a focused search. It wasn’t a clever search. It was just let’s see what happens if I throw this in. And remember, Google is a living thing. All the time it’s indexing more pages. As more pages appear on the web, more and more stuff is on there. So if you’ve done a search a year ago, try it again today because there might be something there that wasn’t there a year ago.

So, as I said, my particular brick wall that I’m stuck on is that little picture that we had right at the start is the village that I think they came from. This here is the village of Dorchester in Oxford – not Dorchester in Dorset. And I’ve traced a Port family back that I think are my family, back to this village. And I really want them to be my family because I’ve got very connected to them. I’ve got very involved with them and I’ve come to love them quite a lot. And it would be devastating if I ever found out that they weren’t my family.

But I’ve got the theory. It’s an illegitimacy, and I’ve got a theory that this particular person, my great-great-grandfather, was the son of this woman whose family came from this village. And I’ve never proved it. But I haven’t disproved it either. I’ve got a theory. It’s like a mathematician or a scientist. I’ve said this is my theory that this is the family that my Thomas Port comes from. Until I find proof that it can’t be, that will stay as my number one theory, the most likely solution to the problem.

I can’t claim it definitely is, but everything that I’ve found, all the bits of information I’ve found about the Port family coming from Thomas’ side and coming from his speculative mother’s side, none of them have disproved the theory. So it’s still alive. And that’s what you have to do: build up a theory; try to knock it down. If you can’t knock it down, as Sherlock Holmes says, once you’ve eliminated the improbable and the impossible, what’s left must be the truth.

I wouldn’t necessarily go exactly as far as that with family history, but it’s a good starting point to think if you haven’t disproved it, it could still be true. And you’ve got to keep working. It’s so much easier to find negative proof for something than it is to find positive proof. I can give you countless examples of that. Prove that someone isn’t someone rather than trying to prove that they are.

Transcribed by Mary Pearson as part of a volunteer project, January 2015