* Citing English Census

[Gary_G]

I had a go at seeing what would geocode.
I took the actual address from the census return for the example I gave previously:
220, Tufnell Park Road, London, England

An online tool generated the following info and a map showing house numbers:
Latitude 51.5572145
Longitude -0.1343944

This was only off by about a city block and on the same road.
It sounds like using the street address, with the administrative county and country, worked fairly well.
So; I think I'll stick with stating the location based on what geocodes the best.

[Gary_G]

For those who really want to try to follow Evidence Explained style...

I think I may have found a documented answer to the intended meaning of the [ Jurisdiction] and [Civil_Division] parameters in the context of an EE citation.

The answer was thoroughly buried in the somewhat unrelated section 6.16, "Shortened Citations, Essential Elements". It's not surprising I missed it in a volume spanning many types of citations in some 893 pages.

ref.:
Mills, Elizabeth Shown. Evidence Explained: Citing History Sources from Artifacts to Cyberspace: 3rd edition revised (p. 264). Genealogical Publishing Company. Kindle Edition.

re: Jurisdiction
"COUNTY & STATE (OR PROVINCE)
Most censuses center upon some political jurisdiction. Frequently that was the state or province, then the county or an independent city."

This sounds like one would use the administrative county and parish; as previously suspected.

re: Civil Divisions
"Within a county, province, or equivalent, the enumeration is frequently divided into smaller civil divisions (towns, villages, or post offices), then into wards or precincts."

This sounds like one would use the city, town, village etc. (not including the address) for the [Civil_Division].

I should note that, in the case of the UK, the Jurisdiction and civil division names can sometimes be the same. So; in such cases, it may be wise to include a note in square bracket to avoid misunderstanding. eg. London [admin. county] vs. London [city].

[Mark1834]

Be careful though - the administrative county of London is different to the City of London (not unlike New York City and New York State), so context is key here.

Most of modern London wasn’t in London for censuses up to and including 1881, but in the historic counties of Middlesex, Kent, and Surrey. Many original documents, particularly pre-printed ones, give the wrong county as boundaries have been redrawn.

[fhtess65]

It all goes back to her usual line "Citation is an art, not a science." I thought the example I linked to was pretty clear, but as you know, her style isn't what I use for just the reasons you've outlined. Obviously I missed something.

Re: ESM's answers on her forum. I think that she wants to avoid people just doing things by rote, rather than fully understanding the source they're citing. Unfortunately, most people reading her 800+ page book, skip the first two chapters, which are the most important, and just focus on the templates.

Still, I agree, some of her answers on the forum only end up confusing me more.

Thanks, Teresa. Actually; Elizabeth Shown Mills' Evidence Explained website often tends to get me more confused than I was before. After years of reading those posts, I feel that she seems to seldom answer a question in a way that I can see a solid rule-of-thumb that will work elsewhere.

<SNIP>

Frankly; I think I'm going to simply look for something that both defines the census location and is able to be geo-coded. I'll still show the information in the reverse order she prefers. However; if that's not what she intended, then I hope she will strive to be more clear in future.

[Gary_G]

Mark;

Your caution is valid.

That's why I put:

I should note that, in the case of the UK, the Jurisdiction and civil division names can sometimes be the same. So; in such cases, it may be wise to include a note in square bracket to avoid misunderstanding. eg. London [admin. county] vs. London [city].

Britain can be confusing...

[Gary_G]

Teresa;

I understand. It's just that I never did buy the "Art, not a Science" blather. It was used too often to avoid admitting that the style was difficult to understand and put into practice. The very fact that a new revision was just released to address some of this (I hope) shows that the people expected more clarity and less verbiage in a standard.

While I no longer religiously try to follow that standard, it did have some useful bits that I still use.

Anyway; 'nuff said. With the answer, I can return to getting some actual work done.

[fhtess65]

All depends on your perspective - I find Canada and the US way more confusing, as the majority of my experience is with English records

If you really want confusing, try citing Polish records!!

<SNIP>

Britain can be confusing...

[Gary_G]

Well...

I've finally found something that helps me make sense of locations in the 1841 census of the Parish of East Peckham, Kent.

FindMyPast erroneously lists "Bush" as a street. In fact; it was one of 9 hamlets that later formed the place called East Peckham. Those hamlets were; Roydon, Hale Street, Beltring, Little Mill, The Pound, Snoll Hatch, The Bush, Goose Green and Chidley Cross. What helped me figure this out was that every one of the locations on the pages referred the set noted hamlets. I should note that the enumerator often dropped "The" in the census, when referring to "The Bush". The important point is to not take the FindMyPast location as gospel; check it out.

So, for the citation; the "Civil Division" is "The Bush" and the "Jurisdiction" is "County of Kent, Parish of East Peckham". However; I should likely enter the location in a fact as, "The Bush, East Peckham, Kent" and leave the address field blank. It will not be automatically geo-locatable, but one can locate it on Old Maps Online.

*** Later add ***
Perhaps, I should ask Adrian to jump in here.

It looks like one or more hamlets can occur on a given page in the 1841 census. That makes me think it might actually have been treated more like an address; much as one would treat "Torwood [House]" in the Scottish Census.

[tatewise]

You should be able to make it auto-geocode if in the Place record you enter a Standardised field place name that is the modern-day equivalent.

[Gary_G]

Mike;
If I enter "East Peckham, Kent, England", it would geocode, but where would I capture that he actually lived in "The Bush" at the time?

One other question that you might be able to answer...; I'm not sure how geocoding sites deal with places in a parish of the same name. When I get a hit, am I getting the location of the parish or of the place in the parish. This could complicate matters, unless the sites expect just the place, county and country.

[LornaCraig]

....It looks like one or more hamlets can occur on a given page in the 1841 census. That makes me think it might actually have been treated more like an address; much as one would treat "Torwood [House]" in the Scottish Census.

Yes, tiny hamlets often didn't have 'addresses', as such, within them. Especially not as early as 1841. Even if there was a road name the houses were unlikely to have numbers. If someone wanted to find a particular person they would probably just ask "where does Fred live?" and someone would point to his house!

When I get a hit, am I getting the location of the parish or of the place in the parish.

You are probably getting the place, which may or may not be centrally located in the parish. As the majority of parishes were named after what was at the time the main settlement within them this will happen a lot. And in some cases subsequent development means that other settlements in the parish are now much bigger than the one which gave it its name.

[Gary_G]

Lorna;

I agree. I grew up in a small place of 113 people (likely including the dogs and cats ). Nobody used addresses. They would often refer to my father as just "Al from Buckhams Bay".

So, I've been captioning my images as follows:
CENS: England: 1841: County of Kent: Parish of East Peckham: p. 5: household of George Bishop of Bush [hamlet]

I'll likely cite it in much the same fashion, but will likely be forced to "modernize" the way I show things in a fact.
Perhaps just using "East Peckham, Kent, England" would work for a place and entering "The Bush" for the address.

[LornaCraig]

Perhaps just using "East Peckham, Kent, England" would work for a place and entering "The Bush" for the address

Yes that's what I would do.

[ColeValleyGirl]

Mike;
If I enter "East Peckham, Kent, England", it would geocode, but where would I capture that he actually lived in "The Bush" at the time?

You'd either capture it in the Address, or enter a Standardised Place for it that would geocode, for example:

Screenshot 2024-02-19 145502.png (22.57 KiB) Viewed 617 times

(Llanfair Nant Y Gof was later subsumed in Trecwn, and doesn't geocode these days).

[Gary_G]

Helen;

Thanks for the image. I can see that I can put in the actual name and a standardized name, but that the standardized name is what gets auto-geocoded. If I need to "force" and "lock down" the location of "The Bush", I believe that FH7 allows that, too. Am I correct?

[LornaCraig]

If The Bush is the only hamlet you need to refer to within the parish of East Peckham you could simply enter the precise geocoding for The Bush in the standardized field for the place East Peckham. Otherwise you would need to include the hamlet name(s) in the Place name, creating a separate Place record for each hamlet.

Once you have done that and set the geocoding to your satisfaction you can prevent it being overridden by using Tools > Preferences > Map window and set Mark Auto-geocodes as Tentative If: to Always.. Then Block Refresh for Non-tentative Geocodes. Any geocodes you have adjusted manually will be non-tentative and therefore won’t be disrupted by any future automatic geocoding. Is that what you mean by 'locked down'?

[Gary_G]

Thanks, Lorna.

Yes; that's what I meant by "locked down".
I'm still coming up to speed with how to refer to various elements of FH7.

My image captioning is moving along much better, now. Clearing up how to document locations has helped.
Loading my pre-existing images into the genealogy app and then putting a fair amount of the citation info into the captions works well for me. It seems that it will speed up the citation process just a bit.

[AdrianBruce]

... I'm not sure how geocoding sites deal with places in a parish of the same name. When I get a hit, am I getting the location of the parish or of the place in the parish. This could complicate matters, unless the sites expect just the place, county and country.

I would suggest that if you try to geocode a place in a parish of the same name, then you will almost certainly get the physical place, simply because that's the most "popular" feature on the map. (Whatever "popular" means...)

Fundamentally though, you should always check on the map where somewhere has geocoded to. Don't rely on Google's (say) geocoding looking at the placename hierarchy that you've written. Rather I believe that it's much closer to a text-based Google search. I'm not saying that it is text-based, but sometimes it looks more text-based than hierarchically inclined. (Other geocoding sites may behave differently...)

One "gotcha" that you need to be wary of is when you attempt to geocode a tiny place in a county but there is a major road of that name in another town in the same county - then, every once in a while, your geocoding will point to the major road, not the tiny place. That's when you need to start dragging things on the map until the pointer is over the desired tiny place. This happened to me when trying to geocode somewhere in Gloucestershire - the tiny place was actually a hamlet next to a castle but there was a main road of the same name as the hamlet miles away. Fortunately, I was following my own advice and saw the problem on the map.

[Gary_G]

Good advice, Adrian. Thank you.
I usually double-check names and places, because England is such a challenge wrt. place-names.

[AdrianBruce]

... It looks like one or more hamlets can occur on a given page in the 1841 census. That makes me think it might actually have been treated more like an address; much as one would treat "Torwood [House]" in the Scottish Census.

Yes - that would work. Fundamentally, it depends on how precise you want to be with your geocoding and how widespread the hamlets are.

One example is that there is a feature on some maps called "Smithy Corner" just outside the village of Barthomley in Cheshire. If you were to go there in real life it is now just one building - a former blacksmith's, no surprise. I would code that up by putting "Smith Corner" in the address and "Barthomley, Cheshire, England" in the place. Anyone living at Smithy Corner would then get marked up with a pin on the map over the village of Barthomley.

Conversely, my great-grandparents lived in the hamlet of Vauxhall on the very edge of the town of Nantwich. For personal reasons, I preferred to see where they lived so I entered a placename of "Vauxhall, Nantwich, Cheshire, England" and left the address blank unless, by some chance, I actually got a street address as well. I can't remember whether "Vauxhall, Nantwich, Cheshire, England" geocodes accurately or whether it geocodes to Nantwich. If it's the latter, then I'd drag the pin for Vauxhall to be over where great-gran lived - it's easy enough to find on the map of the right scale. (They lived opposite the Workhouse - nice - not... )

[Gary_G]

England is still a bit confusing to me, but imagine that western Canada could be the same for someone living in England. I'll get it straight... eventually. Thanks, Adrian.

[Gary_G]

Before rushing ahead, I thought I'd take a look at how the way the National Archives structures the census and how the FindMyPast search page for the 1911 census locates records. After playing with the search engine a bit, it appears that there are two distinct ways to "re-locate" a record. I'm hoping that the general approach is applicable to other years of the census.

The first involves using the county/city, the registration district & sub-district, the enumeration district and an address that matches what FindMyPast has indexed. The latter can be a problem, because they don't always get it correct I noticed that using the schedule in place of the address doesn't uniquely identify the household to the search engine. (I suspect for other years, page number and schedule may be needed.)

The second is to use the content of the archival reference data. eg. "RG14PN4093 RG78PN158 RD50 SD3 ED1 SN44". Note that for other years, one needs to use the equivalent archival reference from the image, plus the folio, page and schedule number. This option is what Adrian suggested, works very well, and is not subject to address mis-indexing issues.

If one puts both together, one gets a relatively short, fairly robust, EE-like 1911 citation that looks something like the following. If this works, it should be easy to template and even to adapt to other methods like the Strathclyde citation approach.

1911 census of England, Kent, RD Tonbridge (50), SD Brenchley (3), ED 1, shed. 44, household of Thomas Wells, at the Post office, "Brenchley, Paddock Wood" [Paddock Wood]; imaged at FindMyPast (https://search.findmypast.com); archival reference RG78PN158 RD50 SD3 ED1 SN44.

I am looking into citing other census years in a similar manner. I hope that this approach will work for them as well.

Adrian: please feel free to chime in. I think this works better than trying to use my previous approach of trying Administrative County and Civl Parish.

[AdrianBruce]

... The second is to use the content of the archival reference data. eg. "RG14PN4093 RG78PN158 RD50 SD3 ED1 SN44". ...

I would never use that particular construction because it's not a reference at TNA but an image group key concocted by FindMyPast. FMP had a very good reason to do what they did but I suspect that most of us will find it overkill.

RG 14 is the class for the 1911 Census Household Schedules.
RG 78 is the class for the 1911 Census Enumerator's Summary Books.

"RG14PN4093 RG78PN158 RD50 SD3 ED1 SN44" contains enough data to find both the Household Schedule and the corresponding Enumerator's Summary book including that particular household, with the linking stuff "RD50 SD3 ED1 SN44" being common in some fashion (no idea how) to both.

I would suspect that most of us are only ever interested in the Household Schedules (in RG 14) and that therefore "RG14PN4093 RD50 SD3 ED1 SN44"
would suffice. Indeed, for the 1911 census the RD, SD and ED aren't necessary as
"RG14PN4093 SN44"
should find the household. (It does, I just tried it).

However, because you do have to use the ED for the 1921, I wouldn't disagree if you wanted to use the full
"RG14PN4093 RD50 SD3 ED1 SN44"
I did just that and confirmed that it's still the same household of 4 people.

[Gary_G]

Adrian;

Thank you for the feedback.

I understand your point. I'm trying to work through the rationale for citing things in a particular fashion and my EE-style background is showing.

Perhaps the following is more in line with what you are suggesting. For the UK, I think this style would work well, but keep in mind that it was designed to not require a short footnote or a bibliographic form. It also has a peculiar syntax which uses semi-colons where one might think to use commas. All that said; it fully avoids the confusion about jurisdiction and civil division.

Ian G. MacDonald would cite things in a pure Strathclyde manner; as in the following example from his book.

Census returns. England. Aston, Warwickshire. 02 Apr 1911. AARON, Francis, RD 385; PN 18210; ED 21; SN 101. Collection: 1911 England & Wales Census Collection Image. http://www.findmypast.co.uk.

Census returns is the class of citation and promotes proper sorting.
England is the country and sorts to give a sub-category of Census returns.
Aston is the name of the registration district.
Warwickshire is the name of the encompassing county.
02 Apr 1911 is the official date of the 1911 census.
AARON, Francis, is the name of the head of the household
The RG 14 is implicit, as the 1911 census has been identified earlier. However; adding it wouldn't hurt.
RD 385; PN 18210; ED 21; SN 101, are drawn from the archival reference.
[Note: 385 is the registration district number for Ashton, Warwickshire. This makes it easy to get the name correct.]
1911 England & Wales Census Collection Image, states the item is an image from a specific collection.
[I'd likely state it as; Collection: "1911 Census For England & Wales", image]
The inclusion of the main cite URL follows, but I'd likely add an access date. Some sites update collections over time.

[Gary_G]

Just a note:

Census returns. England. Watford, Hertfordshire. 2 Apr 1911. BUTTERFIELD, Herbert E. RD 140; PN 7715; ED 17; SN 101. Collection: 1911 Census For England & Wales. http://www.findmypast.co.uk : 8 Mar 2020.

The above seems to work fine for relocating the record on FindMyPast and should have enough info to be used elsewhere. It's a LOT easier than EE-style, too. The citation is per the MacDonald book with just a minor tweak to include the access date. Must see if Strathclyde style actually prescribes a specific way to include the access date.