I am researching my 2x great-grandmother.

(Ann) Norah O’Brien
BIRTH ABT 1860 • St Pancras, Middlesex, England
DEATH 16 JAN 1923 • Exeter, Devonshire, England, United Kingdom

Her life has been quite a mystery.

I’ve posted a few questions about her before.


Basically I cannot trace my tree back anymore because there is ambiguity surrounding her parent’s names, her place of birth, and date of birth. I also can’t find any matches on the previous censuses.

She says different things about her age and her birthplace on her marriage certificate, and on the 1901 and 1911 censuses.

I found her death certificate and I’m confident that it’s hers because the name and the name of her eldest son matches. It would put her birth year at about 1861.

On her marriage certificate it says that she was the daughter of John O’Brien but when I look at this suspected birth certificate it says James O’Brien.

I looked in the GRO Index. I left the place of registration as a wildcard. I tried all years between 1858 and 1868. I also tried both of her names Ann Norah O’Brien and Norah O’Brien.

I found one record for Ann Norah O’Brien in 1860 for St. Pancras.

If I assume that the birth certificate is correct her parents’ names were James O’Brien and Ellen O’Donnell.

I have looked in the 1861, 1871, and 1881 censuses. I only found one record that kind of matches.


enter image description here enter image description here

enter image description here

2 Answers 2


To expand on @JanMurphy's first method, there's a set of rules that genealogists use when they get stuck:

  • If you can't go backwards, go forwards
  • If you can't go forwards, go sideways
  • If you can't go sideways, go to the pub

You've been unable to work back from Norah's later life, so try another approach. Start with something you do know. You have a candidate for Norah's birth. That birth certificate describes a family: James, Ellen and Ann Norah O'Brien, living at 12 Church Way, Somers Town, St Pancras. The question is, is that your family or another? So the goal is to research that family, and see if their Ann Norah can be confirmed or eliminated as your Norah.

From that certificate, I would assume that James O'Brien married Ellen O'Donnell some time before 1860. A quick search brings up a good candidate on Ancestry in their "London, England, Church of England Marriages and Banns, 1754-1932" record set:

Married on June 05 1853 at St Pancras Church, St Pancras were:

James Joseph O'Brien   Full Age  Bachelor  French Polisher  Father John O'Brien
Ellen O'Donnell        Full Age  Spinster   --              Father Jeremiah O'Donnell

The names are a good match, as is James' occupation, and the date and location are a decent fit for this family. One witness is Matthew O'Brien, who might become interesting at some point. And we gain a middle name for James, and the names of both fathers. Both are of "full age", which would normally be 21 then, so we might assume that both were born earlier than 1832 or so. Or perhaps a year or two later, these things weren't always exact, or checked.

Can we find them in the 1861 census? Ann Norah would be less than a year old. This is a little trickier, but there is a decent candidate in Somers Town, St Pancras (RG9; Piece 110; Folio 43; Page 17 on Ancestry)

Here we have, I think, spread over two pages:

James Brien   Head     27  Driver of Hackney Carriage  Middlesex St Andrews  
Ellen Brien   Wife     29  -"- Wife                    Middlesex St Giles  
Mary Brien    Daughter  7  Scholar                     Middlesex St Pancras  
William Brien Son       3  Scholar                     Ditto  
Hannah Brien  Daughter  7m                             Ditto  

The writing is very unclear, and James' occupation could also be something like "Donor of Wedding Carnage" but I think the more likely options are as transcribed above. Oddly, the census taker has written in an uninhabited property between James and Ellen and the children. I think he's put that out of sequence by mistake.

There are several problems when we compare this family to the information we already have. The surname is wrong - Brien instead of O'Brien. James' occupation is different from that on the birth certificate just a few months earlier. There's no Ann Norah.

But look at the address (it's clearer on the previous page): 12 Churchway, Somers Town, St Pancras. It's the same address as James and Ellen on the birth certificate. I doubt that's just coincidence. And people do change their jobs quite abruptly at times. And Hannah is essentially a version of Ann (or could even be a mishearing of Ann Norah). She is the right age for Ann Norah. I think it is plausibly the same family. (That's not to say you shouldn't look for a better match, of course!)

This gives us more to work with, anyway. What does the new GRO births index have for O'Brien, with MMN O'Donnell? Sadly, Soundex seems to think that O'Anything names are pretty much interchangeable, so a fuzzy search on those surnames brings up a lot of chaff. But confining our results to the London area from 1853-1861, we get:

Margaret O'Brian  Q3/1854 Bermondsey 
Mary O'Brien      Q2/1854 Saint Pancras 
Jeremiah O'Brien  Q1/1854 Strand 
Matthew O'Brien   Q3/1855 Strand 
Ann Norah O'Brien Q4/1860 Pancras 
John O'Brien      Q2/1860 Camberwell

There's clearly more than one O'Brien/O'Donnell family in London at the time, which is not terribly surprising. But there's a Mary, born in the right year in the right place, and Ann Norah of course. No William. Perhaps the surname 'went wrong' again. If we repeat the search for Brien/O'Donnell, we get one more result:

William Brien     Q2/1858 Saint Pancras

So we have candidates for Mary and William as well as Ann Norah/Hannah, all registered at (Saint) Pancras in years that match the census ages. This looks like a five-membered family group, ripe for further research. You might want to expand the search to look for later births, too.

We've also seen that we can't rely on specific names and dates being exactly correct on any given record. And you can't rely on Soundex to pick up all possible variants. That ambiguity isn't going to go away. Mis-spellings and mis-tellings are part and parcel of family research, and you have to accommodate them. Often, it's not really possible to "prove" that you have the right person, but you can try to use the weight of evidence to put the matter beyond reasonable doubt.

The challenge you have now is to follow that family forward. Find out what happened to each of them, if you can. You might find, for example, that Ann Norah/Hannah ends up on a path that means she cannot possibly be your Norah (e.g. dies young, or marries, or emigrates, or…) in which case you can eliminate this family from your research. If you can't rule her out, then keep on plugging away until the weight of evidence one way or the other is irresistible. And then go to the pub.


This is a good example of how we have all created brick walls for ourselves. We blindly accept the information on records -- we match names and call it a day -- and then when we can't find things that match up, we get stuck. One of our community members has dubbed this premature connectivitis syndrome (PCS).

Here are some methodologies you can employ to resolve your difficulties.

Go Forward in Time, not backwards

Have you researched her later life thoroughly? Sometimes we need to move forwards in time in order to make progress. I have found clues in records that include the grandparents and grandchildren but don't list the parent, households with aunts/uncles and nieces/nephews, etc. You might not be able to find a record with the direct evidence you seek. You may have to build a case using indirect evidence and inferential genealogy. See Determining what records are available about a individual?

Add a new record group

In this question you only list census records and certificates. Study the places in which she lived and find out what other records are available for you to look at. The FamilySearch Wiki and the FamilySearch Catalog are both good places to look when you want to expand your knowledge. Look for the research tools and strategy section in the article England Genealogy.

Study the localities in which she lived

How much do you know about the places she lived? Have you assembled a research notebook or toolkit? GENUKI, A Vision of Britain, and other gazetteers can help with jurisdictions, which you'll need in order to effectively do a place search in the FamilySearch catalog. See Determining what records are available in a particular locale?

Use Research Guides

GenGuide and the FamilySearch wiki can help you understand the records you have on hand. Do you understand who created the records you're looking at, and the purpose for which they were created? Have you collected catalog information from the repository that holds the original? Do you understand how the records are arranged? Answering all of these questions can help you see nuances in the records and get clues you might have missed.

For an example: your question said:

"She says different things about her age and her birthplace on her marriage certificate, and on the 1901 and 1911 censuses."

We don't know who gave the information on the 1901 Census, and for most of these records, the information has been told to someone else who has written it down. We all use this "she says" shorthand, but it is more accurate to say that the records say different things about her age and her birthplace.

Take classes

If you like distance learning, you're in luck. The British Isles Consultants at FamilySearch have produced a wide variety of classes about English research. The Wiki Article Classes in the Learning Center can be searched with Control-F in your browser or scanned quickly by eye. In most cases, videos can be downloaded as well as the handouts.

Join the local family history society

You can find a local society by visiting https://www.familyhistoryfederation.com/. While you're there, check the Resources section of the website for The Really Useful Guide and other free resources and research tips.

Study her friends, associates, and neighbors

Have you identified all of the witnesses for the certificates you're looking at? Have you traced all the people in her census households? Identify all her known associates, including extended family. See Quicklesson 11: Identity Problems & the FAN Principle for a worked case study.

Keep a journal and research log

The Family History Library has forms you can download here: Strategic Research Logs—England. Choose the log for 1837-present.

The form combines the prioritized sources with small research log tables for each type of source. In these tables, with a word processor such as Microsoft Word, you can easily add rows and expand them as you enter the results of your searchs. You can sort the rows within each table by name, date, or whatever will help in the research process. They also provide for linking to scanned copies of sources and to a Research Analysis Table.

You can also download a form for your Research Analysis Table from the FamilySearch Wiki.

For more ideas, see my answer to a similar question: Searching for UK records when only minimal information to go on?

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.