Attaching Non Head-of-Household members in 1840 & earlier US Census records?
The 1840 and earlier US Census records, only list and index the name of the Head-of-Household. All of the other people in that household are only indicated and counted in various sex / age / race categories. No other names are listed or indexed.
Attaching the source record to the Head-of-Household is understandable. But is it acceptable to attach the same source record to other family members (i.e. Spouse, Children, Parents, etc.)?
IMHO, the named Head-of-Household should be the only person attached. No one else.
Please advise.
Thanks.
Best Answer
-
Per an earlier post in this thread by Aine.ni.Donnghaile "When a census or other record where a person is not named is attached to that person, the hinting algorithm gets confused."
So, I'm going to adopt the practice of only attaching source records to the people who are actually named in the index, or in the full text, or by seeing their name (or nickname) written on the image. And editing the source Title or adding notations as described above will help to clarify those cases.
For the 1840 and earlier US Census records, I'm not going to attach the Head-of-Household's record as a source for any other people. While it's very likely the categorized tick marks/numbers represent other members in the Head-of-Household's family, there's no guarantee. Some of those marks/numbers could represent parents/in-laws, widowed/divorced siblings possibly with children, orphaned nieces/nephews, etc.
0
Answers
-
I have done this for my 5 greats grandfather's family, via the Source Box, but only after doing a calculation that demonstrated that the family properly matches the census (numbers, genders, ages), and pasting this as a note on the source as evidence.
I might not have done it at all if it hadn't been literally the only primary source linking him to my four greats grandmother, his daughter, however (there is one published genealogy that backs this up, though their source is unclear), and if there hadn't been previous incorrect FT information in place based on two (I belatedly realised) nonsense entries in the IGI.
I feel for you, it was this episode that made me realise quite how surprisingly difficult early 19th century US research can be.
0 -
@MandyShaw1 - I'm not sure if I understand what you describe here. If the 1840 (or whatever) census record contains an explicit persona for your 5G GF, then how do you attach that record to each of his family members, without having multiple attachments of the same 1840 (whatever) census persona record?
I do think that the idea is theoretically sensible because a census household schedule as a whole can provide information at a household level with implications for individual family members and the implications ought to be recorded with attached sources. I just worry about the implications of multiple attachments in FS (if that's what's happening)
0 -
I did end up with multiple attachments of the same persona. It doesn't seem to have broken anything (as far as I can see).
I just add the source to my Source Box and then attach it to the other FT profile(s) from there.
Here's an example persona (1790 census) safely showing its multiple attachments: https://www.familysearch.org/ark:/61903/1:1:XHKT-CQR?lang=en
0 -
@MandyShaw1 - my fear would be that multiple attachments of the same persona is teaching the system that each of the attached profiles represents the same real-life individual. Probably not something with immediate effects but …
2 -
@Adrian Bruce1 Hmm, you may be right. DQS might whinge, for example. Perhaps I should go back and change the extra ones to external links.
0 -
My personal opinion, and my process in over 20 years of research: the US censuses before 1850 name only the Head of Household. All other members are represented only by tick marks for age range, gender, and race. The census should only be attached to the Head of Household.
A note or discussion could be added to indicate who we think those tick marks represent.
10 -
That sounds like a good plan @Áine.ní.Donnghaile, thanks all for clarifying my thinking here.
3 -
When a census or other record where a person is not named is attached to that person, the hinting algorithm gets confused.
For example, on Ancestry, with individual trees, when one of my cousins attaches the 1840 for my 3rd great-grandfather to my 3rd great-grandmother in his tree, I'll receive a hint to attach that same record to my 3rd GGM in my tree. Rinse and repeat - there are hundreds of trees for my 3rd great-grandparents. My 3rd GGM died in 1833 and is not represented in the 1840 census, but I regularly see hints to attach that census to her profile.
2 -
If you do decide to continue adding these records to multiple members of the family, you are sure are represented in the statistics (but not individually named), I would edit the title of the source to (in this example) - "Thomas Eddy Household", instead of leaving it as just "Thomas Eddy".
When different individuals are named in a source I am inclined to reference all of them, to make clear what / who the source refers to. For example (and use of the Source Box is not even necessary in instances where FamilySearch suggests adding extra profiles to the source (usually parents or a spouse). In the example of a baptism (found from "Full Text Search") I would put the title as (say): "John Brown Baptism, son of Thomas & Mary Brown…" I then choose John Brown as the primary individual to whom the source should be added, then (when the drop-down appears on the next page), I take the opportunity to add the source to parents Thomas and Mary. The source then appears with the same, meaningful title in all three of their Sources sections.
I believe this detailing of the contents of a source should partly address @Adrian Bruce1 's point that:
"my fear would be that multiple attachments of the same persona is teaching the system that each of the attached profiles represents the same real-life individual"
Unfortunately, FamilySearch engineers left us with a general need to edit most of the source titles we now attach from the main database. For some unknown reason they changed the coding so that a source that was previously presented as (say) "Thomas Brown in entry for John Brown" now gets transferred to the Sources section as simply "Thomas Brown" and we are left to guess to whom the source (John Brown, in this case )primarily relates - unless we choose to edit it to the old (more meaningful) format ourselves.
In summary, it's usually best to create / edit a title that reflects the contents of the record (in Mandy's case, the household of Thomas Eddy) rather than one that names just a single individual, but where there are others (we have worked out are referenced, if not named, in the household), so to whom we wish to add the source, too.
3 -
What I have been doing is the following: I attach the source to the named Head of Household. Then I make a chart of the gender/ages/numbers with parenthetical notations of whom I believe those numbers represent. I put that chart in the Reasons in the Residences section for my Head of Household.
Then I copy the Residence date, place, and my chart into the Residences area of each person in the family that I know. In Description of Residence I put "See source attached to [PID of Head of Household].
I'm hoping this provides the information, while making it clear where it comes from. (Instead of having to make those charts every time, I have a document on my computer with the age breakdowns for each census year. So I just copy and paste.)
Perhaps something more about this in the Notes would be helpful too.
1 -
I agree with @Áine.ní.Donnghaile that: "The census should only be attached to the Head of Household." because: "All other members are represented only by tick marks for age range, gender, and race."
And I agree with @Adrian Bruce1 about the effect that: "multiple attachments of the same persona is teaching the system".
IMHO. If a person's name is not listed in the record or index, it should not be attached in FamilySearch.
0 -
Further to my earlier remarks, I would make it clear that my personal choice in cases like this would involve not attaching such a record as a source to individuals who are not actually named in the document. That said, I can see the argument for adding an individual who is clearly referenced - such as the (known) wife of the head of the household. As suggested (regarding this specific case), if a user insists on doing this (and I believe there is no "rule" against the practice) the source should certainly not be titled as "Thomas Eddy", but definitely as "Thomas Eddy('s) Household". This makes it clear that the source being attached makes reference to his known wife (and children), even though not naming them. The positive side of attaching as a source to other identified individuals (admittedly not named) in this record is that you are placing these family members exactly where they actually were resident at the time of the census.
Re-emphasising this would not be my personal practice, one has to recognise that most users would find it acceptable to attach a source to an incorrectly named individual - if completely sure their name had been recorded incorrectly in the original entry. As an example, I would not hesitate in adding a source to the known mother "Mary" if she was named as "Ann" in the baptism record of her child. The important point should be that we clarify any confusion by a note, or by amending the title of the source. In this example, the baptism record would possibly read:
"Thomas, son of William and Ann Wilson"
When adding this to the mother, the FamilySearch source title would now commonly be titled "Ann Wilson" (yes, without any reference to whatsoever to Thomas, the subject of the baptism!)
In this instance, I would edit the title to:
"Ann (should be MARY) Wilson in the baptism record of Thomas Wilson"
The two things I would not think of doing are:
(1) Not attaching as a source to Mary (when other evidence shows she - not an "Ann"- was Thomas' mother), just because Mary is not actually named (as such) in the source / record.
(2) Leaving the source title unedited, which would imply I had attached a source for an Ann Wilson to Mary Wilson (married name) in error.
I know many might consider this example goes way "off topic", but FamilySearch engineers made a truly awful decision in truncating meaningful source titles, so they now only include one name: that of the individual in the source to whom it has been attached. Users have now been left with the daunting task of deciding to whom a source primarily relates and then (optionally) to retitle / edit it to reflect what the source truly represents.
0 -
(Taking advantage of a currently suspended F1 race!)
Would be grateful if you could elaborate on that comment:
"my fear would be that multiple attachments of the same persona is
teaching the system that each of the attached profiles represents the
same real-life individual."Perhaps I am misunderstanding your line of reasoning, but surely while it remains possible to apply multiple attachments of the same source (same URL if the item has not been indexed) the "system" is likely to be confused about what action is being taken in these circumstances.
On the matter of source titles (headings), whilst I believe they should reflect the contents (primary individual / other named individual(s) found in the original record), surely this is only for our benefit, in being readily able to see (from the Sources section) exactly what the source represents, rather than it providing the "system" a way to differentiate in it applying either to one individual or multiple individuals?
Hope you are following me. If you have become used to attaching records as sources through using the Full Text feature, you will know one is actually "encouraged" to attach the same item to multiple individuals (from a drop-down list), whether they are actually mentioned in the original record or not!
0 -
@Paul W - re
"my fear would be that multiple attachments of the same persona is teaching the system that each of the attached profiles represents the same real-life individual."
I was told this by several of the stalwarts of this community whose names you can probably guess. Further, it seems logical to me.
In my head the possible logic is that if I attach the persona for Fred Bloggs, a child in a 1901 household (say), to children Fred (correctly), Billy (incorrectly) and Tommy (incorrectly) (who may not be in that 1901), then there is a bit of evidence in the system that says the persona for Fred is mapped to the profiles for Fred, Billy and Tommy. If the base assumption is that a persona is only ever attached to the correct profiles, then the system might deduce that Fred, Billy and Tommy are all the same person, since all of them are mapped to the single persona of Fred in the 1901.
Now, there's an awful lot of possibles and ifs, in that explanation, but that was what I was told and, running through the above logic in my head, it seemed perfectly possible.
It may be that the engineers have put defensive coding in place to avoid such multiple "cross threading" but, in the absence of anyone providing a conformation by reading what is now probably incredibly complex code, I feel that it still represents a risk to be avoided.
I don't know the Full Text attachment process so I haven't got anything in my head about it.
3 -
Thank you for that response.
I believe the problem in attaching sources partly lies with where we find the records. If we find them in the main database, it means they have been indexed, of course, and multiple names in the same record then have a different URL to distinguish them. The system also assigns a different title (that gets to the Sources section of each profile), now reflecting the name of the (one) individual referenced (e.g. "William Wilson" rather than the old format of "William Wilson in the record of Thomas Wilson".
If we attach from an image-only set (either from a specific collection or from a Full Text search), we make the decision on what title we give it. In the past, we had to put the source in our Source Box in order to add it to multiple individuals named within the record. Now (and I'll have to check if this only applies to Full Text), (in the process) we provide a title, assign it to one individual, but are then taken to a further page where close relatives are listed. In a typical situation with my work, I then click against the names of the parents of the individual I have made the prime person, so the source can be assigned to their Sources section, too. However, (as I believe has always been the case) if we change the title of the source for one individual, it gets changed for all the individuals we have chosen for it to be added. So, one has to assign a title that will be meaningful when viewed from the Sources section of the multiple individuals referenced. That is why I would use a format such as: "William Wilson Baptism, showing parents Thomas and Mary" - the alternative "William Wilson Baptism" would look like an error when viewed in the Sources section of Thomas Wilson (his father).
In summary, I don't think the engineers / developers have still quite worked out how best to deal with adding a source to multiple individuals. With image-only records the title gets changed for everyone if edited, whereas with indexed records, the name goes across in (say) the name of the child or that of the father or of the mother - but at least (if I have a year or two free!) I can edit each source title to at least indicate whether the first name applies to (say) the child or the parent.
As you indirectly imply, getting the process of adding a source to multiple individuals, without this producing potential problems in getting optimum results via the search engine, would probably involve some very complex coding!
0 -
I believe that @Adrian Bruce1 concerns are real.
And, per an earlier post in this thread by Aine.ni.Donnghaile "When a census or other record where a person is not named is attached to that person, the hinting algorithm gets confused."
So, I'm leaning towards adopting the practice of only attaching source records to the people who are actually named in the index, or in the full text, or by seeing their name (or nickname) written on the image. And editing the source Title or adding notations as described above will help to clarify those cases.
But, for the 1840 and earlier US Census records, I'm not going to attach the Head-of-Household's record as a source for any other people. While it's highly likely the categorized tick marks/numbers represent other members in the Head-of-Household's family, there's no guarantee. Some of those marks/numbers could represent parents/in-laws, widowed/divorced siblings possibly with children, orphaned nieces/nephews, etc.
2 -
@TGoymerac - I'd agree with you. On the principle of "Never say Never" I'm reluctant to predict that I'll never attach the Head of Household to anyone else, but I will say that I really can't conceive of any circumstances where I would do so.
I suspect that if I ever wanted to use the Household schedule for logic relevant to others, then I'd probably create Notes or Memories describing the Household schedule and the logic I used. That way I can add all the caveats that you justifiably list and allow the reader to make their own decision. Such tactics would avoid all the issues of multiple attachments.
4 -
@TGoymerac Please do not post a link to our profiles instead of tagging us.
Thank you.0 -
@Áine.ní.Donnghaile I'm very sorry.
I thought copying & pasting a respondent's link was the way to do that because the link to the responder (with the avatar) and the link to the tag (with the "@" symbol) are exactly the same.
I believe I know the proper method to include a person's a tag now.
Please accept my apology.
0
