Adding context to census addresses

I’m currently going through an exercise to normalise the spelling of all the place names in my Unst Family Tree – every census seems to invent a different spelling for the same house! At the same time I’m adding a ‘district’ to each house as I’m aware of a number of house names that are not unique within the island of Unst; “Garden” is a house in Colvadale and a house in Snarravoe for example (as I’ve previously written).

The absolute need to do this was illustrated to me when I found this particular couple.

Andrew Smith (b. 1836)		m.	Jane Winwick (b.1835)
1881: Middleton			1881: Middleton
married
	1891: Middletown
	1901: Middletown

Above is what you would discover about the couple (from 1881 onwards) using just Ancestry transcriptions. You would be forgiven for thinking that they live in the same house before and after the marriage (spelling variations being what they are), AND that they lived in the same house as each other before they married.

If, however, you look at the census pages with the neighbouring houses in context, your understanding of this couples residence would be as follows.

Andrew Smith (b. 1836)		m.	Jane Winwick (b.1835)
1881: Middleton, Muness			1881: Middleton, Colvadale
married
	1891: Middletown, Muness
	1901: Middletown, Muness

So they lived several miles apart before they married, and after being married they lived in the house where Andrew lived, in Muness. The spellings of both houses use both spellings throughout the census returns over the decades.

If you then follow this up with the marriage record for this couple, from Scotland’s People, it pulls it altogether, giving their addresses at the time of marriage.

Andrew Smith (b. 1836)		m.	Jane Winwick (b.1835)
1881: Middleton, Muness			1881: Middleton, Colvadale
1883: Muness			1883: Colvadale
married 1 Mar 1883 Middleton, Colvadale
	1891: Middletown, Muness
	1901: Middletown, Muness

I can’t stress enough how important taking all records in their full context and understanding the local area where you are researching.

31 Days of Family History Fitness – Week 1

I only came across the blog post 31 Days of Family History Fitness at the very end of January so I decided to do it in February instead. I’ll update you with my progress on a weekly basis.

Day 1: Sync your desktop software and online family trees to make sure you have the latest version of your work in both places. If you don’t use desktop software, download a GEDCOM of your online tree instead. If you don’t have genealogy software or an online family tree, consider starting an online tree to help you organize your research.

I don’t use desktop software but instead build my private tree on the Ancestry website. I do download GEDCOM copies of it in an ad hoc manner, when I think I’ve made a sufficient number of changes to warrant doing so. This had increased a little recently when I’ve been playing around with TNG, but I could do with a more regular pattern as well as the ad hoc extras. So since I’m starting a new month to improve my family history fitness, I’ll mark the beginning of the month as the time when I take a new GEDCOM copy.

Day 2: Back up your genealogy data using an external hard drive or a cloud service. If you don’t currently use a backup system like Dropbox or Backblaze, take some time to learn about them.

With my new regular export to GEDCOM I’m also going to use that to save it off to an external hard-drive too. If I do them both together I have more chance of achieving this as a regular step.

Day 3: Create a source citation workflow so that you—and anyone who sees your research—will know where you got your information.

Citations are one of my strong suits, and possibly the reason why it’s taking me so long to finish my work. Nothing goes into the tree without full transcriptions and source citations.

Day 4: Set up a file-naming convention, and make sure all your documents follow it. This will simplify your filing and help you quickly find the information you need.

Some of my documents follow a good scheme and some don’t. Census pages and any statutory records (post-1855) are well organised including the year in the naming scheme. Old Parish Records are much less organised and every time I need to browse a particular year it takes me a while to find it. So I’ve used this prompt as a nudge to get that sorted.

Day 5: Find your local FamilySearch Center. Have you ever visited one? These branches of the Family History Library (FHL) have helpful volunteers, local resources, and computers with access to genealogy websites. From a FamilySearch Center, you can also rent microfilmed records from the FHL. Search for the nearest one, and call to check the hours.

This is my nearest Family History Centre. It is open Monday, Tuesday and Thursday 9.30am -3.30pm. I haven’t been there yet, but I definitely want to go.

Day 6: If you’ve taken a DNA test, look at your DNA matches (not just the ethnicity results) and review the match information for any fourth cousin or closer. Also check for messages from any matches who’ve contacted you (and respond to them).

Since I haven’t taken a DNA test, I’ve skipped this day. After all there are more days in January than there are in February so in order to complete it in February I need to get cracking.

Day 7: Choose a town your ancestors lived in and search the FamilySearch online catalog for it. Browse records the FHL has on microfilm and note any that might apply to your research. You might even find links to digitized versions on the FamilySearch website.

Searching the FamilySearch online catalog for Unst showed up a small number of records, but nothing that I hadn’t seen before. At least I know I’m not missing out on something major as a source!

FamilySearch Catalog, results for Unst

Day 8: Search a digitized newspaper collection for names on your family tree on the Library of Congress’ free Chronicling America website.

The best digitised newspaper site for British research is of course the British Newspaper Archives. It has the Shetland Times, from 1872 onwards, digitised which I use in an ad hoc manner at the moment, with an intention to go through more thoroughly once I finish processing all the information I have already collected. It can also be used for free from the library which is nice.

So that’s week 1 finished. I’m feeling fitter already, well better organised anyway!

Always check the real records

Transcribed records, provided by the various online genealogy websites are all very well, but part of your research should include checking the actual record image as well. For English records you may well be lucky enough to do this at the same time, and on the same website as the transcribed version, but for Scottish records you have to get them separately from Scotland’s People.

I had a recent person I was looking into that illustrated, again, to me, that it’s always important to look at the real records.

This lady was recorded as living in a house called Garden, in both the 1901 and 1911 census, and the transcription of the 1901 census said she was born in “North Unst”. That was in itself slightly unusual because most people recorded in Unst census returns have their place of birth recorded simply as “Unst”, without it being broken down any further. This is in contrast to the neighbouring island of Yell where birth places are broken down into “North Yell”, “Mid Yell” and “South Yell” because the island itself is not a single parish, and thus not a single registration area, unlike Unst which is.

When I found this lady’s birth record, it showed she was born in Garden, Unst. Not a surprise since this is where her parents, and later she, also lived.

Now there are two houses called Garden in Unst, one in Colvadale, and one in Snarravoe, neither of which I would consider to be in North Unst! See map for the two locations.

So, I brought up the actual 1901 census record, and it doesn’t say North Unst at all! It just says Unst. It would seem that the transcriber’s eye has been pulled offline to the record below her which records someone born in “North Yell”.

1901 Census

ROAD, STREET, &c., and No. or NAME of HOUSE.	NAME and Surname of each Person.	RELATION to Head of Family	CONDITION as to Marriage	AGE (last Birthday)		WHERE BORN
				Males	Females
Garden	Thomas Irvine	Head	Mar	34		Shetland, Whalsay
	Janet Irvine	Wife	Mar		41	Do North Yell
	Williamina Irvine	Daur			10	Shetland, Unst
	Thomasina Do	Daur			5	Do Do
	Andrina Williamsom	Sister in law	S		53	Do North Yell
	Cathrine Do	Do	S		45	Do Do

So, remember it’s always worth checking!

Conveying Accuracy

I’ve mentioned a few times that I plan to get my Genealogical research onto a website eventually. One of the things I’ve been considering for the design of my website is how to convey the accuracy of the information; after all, genealogical research is fraught with accuracy issues.

This is described well in 5 Ways To Tell If Your Genealogy Research Is Accurate.

I think there are several types of accuracy I’ll wish to convey about the records I’ll have used. Here are my thoughts on them.

Transcription Accuracy

I use a range of records, from those I have seen photographs of the original document and made my own transcriptions (5 star) through to facts I’ve been told by other people without any documentary evidence to back it up (1 star). I can imagine a rating scale something like this.

Star rating	Meaning
☆☆☆☆☆	I have seen the document, or a photograph of the document that the transcription was made from.
☆☆☆☆	Full transcription provided by someone else without a photograph, e.g. Ancestry website Scottish census records.
☆☆☆	Partial transcription provided by someone else without a photograph, e.g. Ancestry website Scottish Old Parish records.
☆☆	Fact provided from someone else with a note of the record where they found it, but without any transcription.
☆	Fact provided from someone else without any documentary evidence to back it up.

Recording accuracy

There are various ways to look at the accuracy of the recording; the distance in time since the event, e.g. age on a death record; the likelihood of the informant knowing the information, e.g. a neighbour instead of a relative; the kind of record, e.g. a statutory record versus something less formal or rigorous. I can imagine a rating scale something like this.

Star rating	Examples
☆☆☆☆☆	Birth/Marriage/Death date from Statutory record of the same.
☆☆☆☆	Age from Statutory Marriage Record. Parents marriage date on Statutory Birth record of child.
☆☆☆	Age from Statutory Death Record if recorded by a relative. Age from Census Record.
☆☆	Age from Statutory Death Record if recorded by a neighbour.
☆	Any other record/fact not listed above.

Overall accuracy of a Fact

There therefore needs to be a formula that conveys the likely accuracy of any one fact recorded for a person, for example if the birthdate comes from a Statutory Birth Record that I have seen the image of, that would be 5-stars on both of the above scales, and therefore the best possible score, however if it doesn’t match all the other records for the same fact, like all the ages on census returns, then that should reduce the score. I suspect I’m going to have some fun trying to come up with said formula!

Overall rating for a person

Having given each fact a score, the average of all the scores could be used to convey an overall rating for each person.

I suspect these ideas will solidify over time and, especially, when I try to put them into practice when I get to the point of creating my website. I will no doubt find other examples that need rated, and complications to any formula that I come up with. Should be fun though!

First names no longer in common use

When working with genealogical records for people who lived 100+ years ago, you get to see first names that are not now in common use. Many of the names rarely seen now are female versions of other male names. The table shows a selection, all of which are formed by adding “ina” to the male name.

Male form	Female form	Derivative
Andrew	Anderina
Jacob	Jacobina	Jessie, Jemima
Hugh	Hughina
Laurence	Laurina
Robert	Robina
Thomas	Thomasina	Tamar
William	Williamina	Willa or Mina

One of those names, Thomasina, has an oft used derivative, Tamar, which definitely confused some of the census transcribers. It’s a girls name, and yet frequently I have seen people with this name, listed in census returns as the daughter or wife of the head of the house, transcribed as James. Looking at the way the letters are written in examples of the old hand-writing it is almost forgivable, the “T” does look like a “J” in old script writing, then you have “am”, easy to assume it is “James”.

Example old hand-writing showing how Tamar could look like James

Mind you, the gender of the person ought to have made the transcriber double check, and also, in this case, six lines above there is a real “James” in the same handwriting. Perhaps we can forgive them however, as names are not always used the way we expect them to be, for example, would you expect ‘Bruce’ to be a female name?

To compare to the text saying Tamar, from the same page, here is text that says James

I’ve seen it so often now, that whenever I see “James” I will double check the gender, and if it’s female, smile to myself and think, “that’ll be another Tamar then”.

Reading Old Handwriting

Perhaps I shouldn’t be one to complain; in this day-and-age of using computers for everything, my hand-writing is likely even worse; but sometimes the writing on old records like census returns and birth/marriage/death records are really hard to read. You’ve got to feel for the people whose job it is to transcribe them for use on the plethora of websites now available for you to make your own family tree. Some of them, I’m convinced you can only decipher if you already know what it’s supposed to say.

Take place names, and worse still, house names; if you’re not from the local area where the places are being recorded, you’ve no chance of getting them right. However, they do a good enough job that someone local can work what it was meant to say, so I guess that’s half the battle.

Occupations are harder though; Sure the common ones are easy to decipher – with a tree of 5000+ people, 45% Knitters and 45% Fishermen, those I can spot. However, every so often I come across one that really makes no sense. In this one case it was transcribed as “Formerly Cond Trung Car”. Pardon? I thought, surely that’s got to be an error in transcription, that can’t be anything close to what was written. So I purchased the actual image the transcription was made from and it said …

Undecipherable Occupation on an 1881 Unst Census Return
Parish: Unst; ED: 4; Page: 3; Line: 19

“Formerly Cond Trung Car”. Pardon?

OK, so I’m none the wiser. I’ve asked on some ancestry web sites and had nothing that rings true either. The most common suggestion was “Formerly Conductor Tram Car” but since there are no trams in Shetland, that seems unlikely. I personally think that “Car” is short for Carpenter, certainly I have seen it used for that before. And, I’ve toyed with the idea that “Trung” is Training, since the guy in question is aged 23, but I can’t work out what “Cond” is.

So I’m putting this out to the world. Does anyone have any idea what this guys occupation might be?