* Question re: usage of Census, Residence and Occupation

[Gary_G]

When one has a Census record, it contains information that can be used to define entries for the Census, Residence and Occupation.
Somehow, using all three seems slightly redundant, but I suppose there may be a reason to do so.
Is someone aware of an article or video that discusses the pros and cons of using a particular combination of the three?
I've not been able to locate a concise discussion of this in a search of the forums and knowledge-base.

[ColeValleyGirl]

Typically, you would only create either a Census event or a Residence event for an individual from a Census record, plus an Occupation.

I prefer Census because (a) you can't be sure that somebody's location on census night is their residence; and (b) most plugins, queries, etc. to interrogate census data work on individual Census events. (Family census events are not recommended as they aren't handled by said plugins, queries etc. and they only apply to a couple, not all their children).

[Jean001]

For 'residential' information from the Census Return I use the Census Fact not the Residence Fact. (It is then easy to check for missing Census years).

Yes, I add an Occupation Fact and update all the other associated Facts (e.g. Birth, Parents, Death (from a widowed spouse), etc.)

[tatewise]

See FHUG Knowledge Base Recording from a Census Record that reinforces what Helen has said.

Many users prefer to use Ancestral Sources to capture Census Records because so much is automated.

[Gary_G]

Helen;
Thanks for the reply.
I'm still a bit confused by what you wrote

Since a Census documents the presence of an assemblage of people who may not be permanent residents, I understand why it's likely not wise to generate Residence entries from a Census record. One could safely generate Occupation entries from a census record, since the concept of occupation is not affected by their location at the time of the census.

In Canada; we have something called a "Record Of Employment". It is issued by an employer when a person leaves their employ and it explicitly states their occupation and residence on a given date (amongst other useful information). For such a record, one could generate both a Residence and an Occupation entry.

I guess that leaves me uncertain what creating a Census entry would accomplish. It could be misleading to use the data as a "fact". I recognize a census record provides valuable hints to the makeup of the family unit, their approximate birthdates and location etc. However; it sounds like it is more correct to treat a census record as just another source to potentially be attached to some other "fact" as support.

Could you clarify why you create a census record? I don't understand the value of being able to run queries on census entries, if the content can't be trusted.

[Gary_G]

Mike;
Thanks for the reference. I read it over.

So; what still bothers me is the actual evidentiary value of entering a Census "fact", as it sounds like the contained residence info may not necessarily be "factual" because of the rules regarding who was considered a resident.

[NickWalker]

Could you clarify why you create a census record? I don't understand the value of being able to run queries on census entries, if the content can't be trusted.

For the majority of individuals in my tree (and this will be the case for all researchers I think), there is very little documentary evidence about them (that is available and locatable at least). For some the only events I might be able to record are when they are born and died and without a census we have no idea what they did in between. A census is an event in someone's life and one of the few events that leave a record that we can access. A census gives us a snapshot of their life. It tells us that on one particular night they were resident in a particular location and it also gives us an age which may be the only evidence we have to help to give us an idea of when they might have been born. So a census event records those snapshots in someone's life. It also means you may be able to focus your research about that person to the area they were in around the date of the census.

[Gary_G]

Nick;

I would agree that the content of the record has some value, but it sounds like one has to combine it with other evidence to reach a conclusion. As we typically have so little other objective evidence of residence location, this situation is definitely an issue. The question in my mind is what does entering a fact for a census record with possibly erroneous info buy us. However; I could see looking at a sequence of census records and drawing a conclusion as to residence location. Anyways... It's an interesting question... Thanks for your comments.

[Jean001]

I agree with Nick's comments.

Factual information... I learned very early on in my family history researches that the information recorded is not necessarily accurate or true. Some of my forebears (and members of other families that I have researched) told lies! Especially concerning their age, marital status, and parentage of 'their' children. Add onto that people who moved away from their birth area to one in which the enumerator/registrar could not understand their accent so placenames (for example) can take some 'deciphering'. Plus a host of other reasons for causing information to be 'suspect': illiteracy, copied and recopied documents, etc.

For me, part of the pleasure of this hobby is gathering whatever information I can and building the families, and some sort of understanding of their lives.

[RS3100]

Any historic record can contain false or erroneous information. It is by seeking information from a variety of sources that we might obtain corroboration to add weight to suppositions and draw conclusions. Censuses are a valid part of that research.

Censuses are also a source of information about the members of a family and give clues that prompt us to look for additional information on those family members elsewhere. The occupations recorded in censuses might help identify the same individuals in other records. I know that many times I have discovered children or other family members who I probably would have otherwise not been aware of, and the census entries sent me off on other lines of research that helped expand my tree.

One of my Gx4 grandmothers was a woman for whom I had very little information with which to positively identify her, as there were three women with the same name in the small village where she lived, two of whom were married to men with the exact same name - one my Gx4 grandfather. Trying to establish which death related to her was only resolved when I discovered that the informant for one of the deaths - a woman with no obvious connection to her - was a near neighbour in the census a couple of years before her death, supporting the likelihood that the death did relate to my direct ancestor. I subsequently discovered that the informant's great granddaughter later married my GG Uncle.

So censuses can be a wealth of information that might be difficult or impossible to find elsewhere.

[tatewise]

Gary, all historical documents have varying degrees of factual accuracy, but if we don't record them because they might have errors then we won't record anything. I agree it is then a matter of assessing the records but if we don't even record them they cannot be assessed.

Nationally administered records such as Birth, Marriage & Death Certificates and Census Records are probably the most reliable and, as Nick says, are often all we have to work with.

Census Events record the address where the person resided on Census night but does not assert it was their residence.
Other information includes their age, place of birth, occupation and family relationships.
These are often the only details of family relationships available and provides clues to search for BMD Certificates.
If nothing else, recording Census Events allows you to Query the people for which you have not found a Census record for a particular Census year. It would be difficult to discover that any other way. The article I mentioned gives other reasons for using the Census Event. It is a standard GEDCOM event supported by every product I am aware of and is extremely useful.

[AdrianBruce]

When one has a Census record, it contains information that can be used to define entries for the Census, Residence and Occupation.
Somehow, using all three seems slightly redundant, but I suppose there may be a reason to do so. ...

I create "facts" for all three of those.

Yes, there are arguments about whether the Residence is anything more than where someone was on the day in question - but in my view that applies to any residence information. Once I've taken out the "Residences" that are clearly temporary such as "Visitor", "Boarder", "Lodger", and ignored the addresses that are clearly temporary such as hotels, prisons, hospitals, etc, then what I'm left with is, in my view, as good as it's going to get. (Two caveats - these are UK terms and some "temporary" residences may be regarded as being just as permanent as non-temporary residences).

Occupations should be reasonably straightforward and uncontroversial.

Both the relevant residence facts and occupation facts will have the relevant census source record cited against them. (And anything else necessary will be cited against the censuses, of course).

So why do I record census events? Simply to make it easier to look at someone's fact tab and see immediately whether or not I have captured and used their census data. If it's the UK, are there census events in (say) 1871, 1881, 1891, etc.? When it comes to producing narrative reports, I always exclude the census events from my narrative reports as it's only genealogists who would be particularly interested in them - the rest of the world is only interested in the residences, occupations, etc.

Very occasionally, there is something of general interest in the census - e.g. I have someone in the 1900 US census on 2 June 1900 in San Francisco and also on 22 June in San Mateo, not too far away. In that case, knowing that the census won't be printed, I've added some notes to the two residences speculating on the reason for the double appearance.

[KFN]

The term “fact” (as used in a lot of software) is really a misunderstanding of the process.

We are not recording ‘facts’ we are recording ‘events’.

We are not ‘concluding’, we are providing ‘evidence’.

Based on a the evidence we assert a hypothesis that may become a fact when other collaborative evidence is presented, but it could also stay a hypothesis or be refuted by additional evidence at a later date by others.

[Gary_G]

Nick;

I realize that the census record has some valuable info in it and I do create a source record for it. As I am redoing my database, I'm just making sure to ask any questions now rather when it's too late. I really don't want to re-do the database yet again.

I note that while the actual Narrative sentence for the census isn't really that informative, the words were carefully chosen and are factually correct. E.g. "He appeared in the census on 31 Mar 1926 in Saint-Gervais, Val-d'Oise, Île-de-France, France". So; it doesn't represent it as a residence. That's good. Also; it does give one a place to attach the source record, so it can be consulted for all the other info it contains.

I guess after all the data entry is done, I'll need to do a timeline analysis to see if any of the census records cause inconsistencies for which I need to create a note to explain them. To give an example; my grandfather was at his grandfathers house so often that the census made it seem like he lived there. However; he actually lived about 40 miles away.

[Gary_G]

KFN;

As you likely can tell from my quotes around the word "fact", I have a bit or an issue with the entries not being called Events and Attributes.
Part of the reason I am so careful about such things is that too many programs (and people) incorrectly use the term, "fact".
But... That's another topic that has been done to death in other threads.

[NickWalker]

The term “fact” (as used in a lot of software) is really a misunderstanding of the process.

We are not recording ‘facts’ we are recording ‘events’.

We are not ‘concluding’, we are providing ‘evidence’.

Based on a the evidence we assert a hypothesis that may become a fact when other collaborative evidence is presented, but it could also stay a hypothesis or be refuted by additional evidence at a later date by others.

The GEDCOM terms are 'attributes' and 'events' and in the early days of FH we always referred to either attributes (e.g. Occupation) or events (e.g. Census) but in a later version of FH (3?) the term 'Fact' was introduced as a convenient word to cover both those terms. But the word Fact in this case isn't meant to imply these are definitely correct.

[AdrianBruce]

The term “fact” (as used in a lot of software) is really a misunderstanding of the process.

We are not recording ‘facts’ we are recording ‘events’.

We are not ‘concluding’, we are providing ‘evidence’.

Based on a the evidence we assert a hypothesis that may become a fact when other collaborative evidence is presented, but it could also stay a hypothesis or be refuted by additional evidence at a later date by others.

As far as the word "fact" goes, I shall claim the Humpty Dumpty defence ("When I use a word, ... it means just what I choose it to mean — neither more nor less.") In this case, we have events, attributes, and other aspects such as names, relationships, etc, etc, all of which can have sources cited against them. Since "events, attributes, and other aspects such as names, relationships, etc, etc" is somewhat unwieldy as a phrase, I use the term "facts" as a replacement.

As far as I am concerned, when I enter a "fact" into my database, that is my conclusion unless I explicitly state that this is a working hypothesis. But it should be understood that all my conclusions are provisional. Like any science...

There are plenty of people in this UG who don't work in my way, whose "facts" are closer to "evidence". Only when they write up that person / family / whatever, will they attempt to reach conclusions. (And I'm sure that they also regard their conclusions as provisional).

So, different people, different processes.

I've long since forgotten the details but ages ago, a group did try to construct a genealogical model based on "proper" logic (whatever that word means - I only did a maths degree and successfully avoided logic as a subject in its own right). I could say the model crashed and burned but I think that's too dramatic a phrase - it just got utterly ignored as being over virtually everyone's heads.

[Gary_G]

Adrian;

I more-or-less agree with what you are saying. The census "fact" is a good place to attach the census source record, even if it doesn't say much in the actual report. At least one can find it later. Did I understand that one can turn off those sentences in the report? I might look at that.

As noted in another post, one has to be very careful with some census records. I should also note that France has a very different idea of who should be in the regular census and it is not like it is in Britain. In France, there were certain groups of individuals who appeared in their own census lists and the rules were all laid out on the first sheet of every census. The French are meticulous to a fault about record keeping, but it means one also has to be so careful.

[AdrianBruce]

... Did I understand that one can turn off those sentences in the report? I might look at that. ...

I believe that there are two means of suppressing sentences in narrative reports. The first is to use the code {blank} on its own as the sentence. That can be entered as the sentence for census events.

I don't use that option for censuses - instead, I use the Report Options and provide an Exclude List (for Individuals' Events / Attributes) - my Exclude List contains the Census events and also "zDiagram Note", which is an attribute designed solely to appear in boxes on Diagrams and contains free text.

On reflection this must be why I don't use {blank} for censuses - other non-narrative reports such as the Individual Summary Report also have an Exclude List, so I must have decided to use the Exclude List across the board in all report types.

... As noted in another post, one has to be very careful with some census records. I should also note that France has a very different idea of who should be in the regular census and it is not like it is in Britain. ...

Indeed - Rule 1 - understand your sources. I still say that my record was to understand the format of the reference number used on printed forms by the British Army, because that gave an ordering date for the forms, which proved we were looking at a later copy and not the original document.

[KFN]

The GEDCOM terms are 'attributes' and 'events' and in the early days of FH we always referred to either attributes (e.g. Occupation) or events (e.g. Census) but in a later version of FH (3?) the term 'Fact' was introduced as a convenient word to cover both those terms. But the word Fact in this case isn't meant to imply these are definitely correct.

I understand the history of the term as used by FH (and other programs). Having spent a lot of my time since the early 80's explaining the process to newbies, my point was that, many new researchers believe that they are actually only entering things that are true (a fact) because they see that term in the software and use the general definition rather than the FH definition.

I instead use other terms such as "evidence" to convey a better understanding of what they are doing. I'm not trying to be pedantic, but pointing out that it is okay to hypothesize that the census indicates a individual took residence at the location of the census data and that this is not truth (a "fact") in the general dictionary sense.

This is similar to the general misunderstanding that people have about the word "Theory" in science vs the common language use of the term. Cell Theory; Theory of Evolution; Germ Theory of Disease, among others, are supported by evidence, where as a the common language use of the term is more like a scientific hypothesis. The Scientific term uses evidence, the non-scientific use needs evidence.

[AdrianBruce]

... my point was that, many new researchers believe that they are actually only entering things that are true (a fact) ...

I am in complete sympathy with the idea that people need to understand that sources may not be "true" (shock, horror!) - that sources may contain accidental or deliberate errors, or that sources may mislead (not necessarily the same as containing errors if the source is a statement that "lies by omission"). And as a result, perhaps "fact" wasn't the best choice as a term. But I can't think of a better one.

Similar concerns may be directed at the word "proof" - as in "Genealogical Proof Statement". But then, the casual observer will be familiar with the court case proven in the first 5m of a TV drama which will then, by one of the laws of scriptwriting, have its verdict overturned by the end of the episode by the heroic defence attorney / forensics expert, etc. So maybe we don't need to worry so much about "Proof" as a word...

[Gary_G]

Adrian;

I believe you mentioned that you used all three tags; Census, Occupation and Residence. Could you elaborate on why you use both Census and Residence?

----

As for the general side-discussion about "proof", I'll put in my two cents...

Frankly; I'd be so very happy if people would speak about having documented "events" and recorded "attributes" in their genealogical program. At least those descriptions don't actually misrepresent what is being recorded; something that the term, "fact" does.

I feel that one is just recording "information", which hasn't even been established as "evidence" at this point. To be "evidence", one needs to view it in the context of "Does it answer the research question?". As each recorded piece of "information" could be relevant to several research questions, one can't just tag an individual "event" or "attribute" with "Direct", "Indirect" or "Negative" (which many programs forget to include) and transform "information" into "evidence". One really needs to look at the "information" in the context of one or more research questions.

I should note that I never allow myself to consider "proof" to be synonymous with "absolute certainty".
The concept of "proof" is based upon weighing all the "evidence" established to date and coming to a reasoned conclusion about an hypothesis.
That conclusion, as history has shown, is still sometimes flawed.

[ColeValleyGirl]

Adrian;

I believe you mentioned that you used all three tags; Census, Occupation and Residence. Could you elaborate on why you use both Census and Residence?

Not Adrian, but my answer is: Because there are sources which do give me direct evidence of Residence that aren't census sources. UK domestic telephone directories and electoral registers, for example, or the address given by the informant on a birth or death certificate. I don't use Residence for address information derived from Census sources however, even if it's overwhelmingly likely that the census address is somebody's residence, so that my handling of censuses is consistent.

One can't just tag an individual "event" or "attribute" with "Direct", "Indirect" or "Negative" (which many programs forget to include) and transform "information" into "evidence".

FH goes a little further than that; you can tag a citation (i.e. the link between source and event/attribute) as follows:

Screenshot 2023-04-29 150317.png (16.61 KiB) Viewed 958 times

There's a Wish List request to expand this to cover other classification from ESM's Evidence Analysis map:

https://fhug.org.uk/wishlist/wldisplay. ... lwlref=569

[fhtess65]

Putting on my historian hat here. NO document's content can be trusted. There's ALWAYS room for error, either deliberate or accidental. No matter what document we look at we have to ask who created it and why. One could make the same argument for a marriage certificate/parish register entry. I know for a FACT that my 2nd great-grandmother, her mother, and her grandmother ALL lied when they married, stating they were widows when they weren't. Does this mean I should devalue this record as evidence of the marriage? No. In fact, when I received the copy of my 2nd great-grandparents marriage certificate and saw she'd stated herself a widow, I questioned whether I had ordered the correct document and did more research, just to be sure I had the right one.

Her daughter, my great-grandmother, stated on her marriage register entry that her surname was Barlow. It was, in fact, Lawley - the surname under which she her mother registered her. She used her step-father's surname, Barlow, but he never formally adopted her as formal adoption didn't exist in that period.
Again, I'm not going to dismiss that entire marriage registration because Rose was continuing with the family's pretense that she was Jesse George's daughter, not step-daughter.

All that said, I've seen this Census debate before - is it fact/event or a source for residence. RM has gone one way, while FH has gone another.

<SNIP>

Could you clarify why you create a census record? I don't understand the value of being able to run queries on census entries, if the content can't be trusted.

[Gary_G]

Helen;

I appreciate your input. Thank you.
Actually; Adrian did say he uses all three in his post; "Post by AdrianBruce » 28 Apr 2023 13:53".