Whole sources of Quaker records mis-indexed
I've been correcting Quaker birth and death dates on records that I found on FamilySearch for one of my lines. Whoever indexed the source records had not been educated on how to read dates on Quaker records older than 1752. Prior to 1752, Quaker dates are based on the Julian calendar that treats the first month of the year as March. What's more, since many of the months are named after "pagan" gods, Quakers refused to use the names of months. Their dates are all written numerically as the X day of the Y month in Z year, where XY&Z are all numbers. Here's an example: "the 31st day of the 11th month of 1712/3." Intuitively, an indexer would (and did) read that month as November, but of course November has only 30 days. The second hint of a problem is the confusion over the year. The latter clears up when you realize that, because March is the 1st month, the 11th month would be January. Since this date is in January, the year is 1713 (not 1712). Their year ran from March through February, so the last two months slipped into the next year on the commonly used calendar.
That's just a sample of why the records have been indexed improperly. A page of the document may be found at https://www.familysearch.org/ark:/61903/3:1:3Q9M-CSC7-HS2S-H?view=index&action=view
I have corrected all the indexed dates on that page. However, that document is 796 pages. Most dates probably do not predate 1752 (though all the dates on this page do). I looked around and there certainly are more pages needing correcting. It needs a dedicated person to go through it and make the corrections. I note that on your "community" page, many users are confused by this.
I suggest you make a point of educating anyone to whom you give Quaker records for indexing.
Attached is an explanation I created that may be useful to give to indexers (or it may be rewritten if you feel it can be improved upon). The first page is my summary, but the following highly technical pages of explanation are taken from a source I've cited at the bottom of my summary. I hope this is helpful.
Answers
-
First: I don't know who the "you" is that you believe yourself to be addressing, but you posted in the public Community of FamilySearch. I, like most people here, am just another user of FS.
With the current version of the Community's URL-mangler, you must use the "Display as Text" setting on all links that contain colons (and some other characters), else they will not work. (Rescued link: https://www.familysearch.org/ark:/61903/3:1:3Q9M-CSC7-HS2S-H?view=index&action=view.)
But it's all moot, as the film or image group in question is FHC-only, at least with a public FS account like mine. The new viewer/editor doesn't say so, but switching to the old one (by removing the question mark and everything after it from the URL) gives the more informative error message, along with access to a citation of sorts:
"Pennsylvania, Historical Society of Pennsylvania, Births and Baptisms, 1520-1999", database, FamilySearch (https://familysearch.org/ark:/61903/3:1:3Q9M-CSC7-HS2S-H?cc=4138679 : 20 July 2023), > image 1 of 1.
Unfortunately, the only thing the Collections search coughs up with "Historical Society of Pennsylvania" is their (unindexed) card catalog (https://www.familysearch.org/search/collection/2524622), so I still don't know the age or provenance of the index you're speaking of.
So, sticking to generalities: FamilySearch does not revisit published indexes, except to fix broadly-applicable metadata errors or the effects of collection-wide automated processes gone awry. The dates of individual entries in an index are not metadata, and they're not the result of any automated process, so I don't believe there is currently any mechanism by which FS staff could or would make any edits to these index entries.
There are currently no indexing projects with "Quaker" in their titles, and the only Pennsylvania project is Beaver county civil marriage records, so your suggestion and instructions do not appear to be applicable right now. Unfortunately, FS doesn't publicize its plans for future indexing projects, and I am not aware of any means of getting in touch with the people who prepare such projects, but perhaps you could check the FS blog for reports about indexing, and ask the author of one of those?
1 -
I have a follow up question about these misinterpreted / mistranslated / or mistranscribed Julian Dates.
Many of us have hundreds of Quaker ancestors whose profiles are littered with these false transcriptions of pre-1752 dates. And the sources section is littered with those indexed records.
I see these pages and unintended fraud probably due to ignorance or negligence. Whatever the cause, many of the indexed pre-1752 entries for Quakers are simply fraudulent. The result of human error.
My impulse is to simply delete those indexed sources from the dozens of Quaker profiles in my family. At best any indexed "source" is derivative and they carry almost zero evidentiary value.
Is there a serious argument against this proposal?
0 -
But what are you going to replace them with? After all, they indicate something… I'm not sure if it's possible to reverse engineer back to the correct date with certainty - and if you can't see the original, it will make it impossible to check any theories, of course.
And surely they indicate the correct geographical area…?
I also get twitchy with the references to "fraud" and "fraudulent". You simply can't have "unintended fraud" because "fraud" has a legal meaning about a deliberate attempt to gain advantage (usually monetary). The indexes may very well be erroneous, but they are surely not fraudulent.
As an aside, I would also take issue with this text in the original post…
"Here's an example: 'the 31st day of the 11th month of 1712/3.' … The second hint of a problem is the confusion over the year. … Their year ran from March through February, so the last two months slipped into the next year on the commonly used calendar."
Now, many people will already wince at what I just quoted - for those who didn't, there was no confusion by the writers of that text over what the year was. The whole purpose of the dual dating 1712/13 is to indicate that the year was 1713 if you regarded the year as starting on 1 January, and 1712 if you regarded the year as starting on 25 March (Lady Day). (If you enter that into FamilySearch FamilyTree, it will suggest standardising on 1713, i.e. today's calendar with the year starting on 1 January).
The quoted text refers to the "commonly used calendar" - but that's simply wrong - until 1754 the commonly used calendar in England & Wales and (as far as I know) in colonies such as the Americas, started the New Year on 25 March. So the years quoted by Quakers appear to be the same as the years that would be used by the Church of England - though it may be that the dual dating is less common elsewhere. Scotland started its year on 1 January after 1600 so after that, during Jan, Feb and most of March, the year was different on each side of the Border until 1754.
4 -
You're letting the best be the enemy of the good if you're proposing to throw away data just because some of the dates may be off by up to fourteen months.
As Adrian says, double dating does not represent any confusion. Quite the opposite, in fact: it's a clear specification of a precise date. Also, like most of what you describe, it is not in any way confined to Quaker records. Starting the year in March goes all the way back to the Romans, after all; that's why September to December are named as they are, based on the Roman numbers seven to ten — and that's why March-based month numberings for the fall months are ubiquitous in all European records, especially those written in Latin. (9ber is November, not September.)
The only part of the equation that can be considered uniquely Quaker is the extension of the month numberings beyond the fall months. Before the 20th century, nobody else referred to months like April or January by numbers. The confusion arises because the 20th-21st century numbering is different from the numbering used by the Quakers in pre-Gregorian documents. (In those old records, 11 is January, not November.) But as I said, the error introduced by such misinterpretations is small: less than two years, max.
0 -
Hi Adrian
Thank you very much for your reply -- getting someone else's view is always helpful.
Let me try to address the matters you raise in order.
1. What will replace the deleted index records?
In general one should always use the best evidence available. So an image of a birth record or a marriage record might be the best record available.
We're talking about Quaker records prior to 1752. In America, no other religious body kept better records of birth, marriage and death than the Quakers. Most of the records that have been indexed are available on microfilm. In some cases the best record available to me would be an abstract or another indexed version of the original record.
In the 1930 Quakers started indexing their records -- they were careful about it. They copied the dates of births, marriages and deaths exactly as they were originally entered by the various Quaker meetings in their records books. These were published in the 7 volume Encyclopedia of American Quaker Genealogy. H. Stanley Craig also published abstracts of the records -- Both collections carefully recorded dates exactly as they appeared in the original records.
But when the Historical Society of Pennsylvania and FamilySearch looked at those records they translated the numbered months, e.g., 1st month, 2nd month,...12th month as January, February, ... December. This is absolutely wrong. Prior to 1752 1st month was March, 2nd month was April and 12th month was February.
So I would replace deleted indexed records with an image from the microfilmed records or an image of the EAQG or from Craig who copied the dates, more often than never, correctly and accurately.
- About getting twitchy with the use of the terms 'fraud' and 'fraudulent'. Artifacts do not have intentions. But artifacts can deceive. The intention to commit fraud is serious and something to get twitchy about. I did not say that anyone is trying to deceive others with those index cards.
Your point here is well-taken — I need to be more careful in the use of those terms. However….
Consider an analogy. Suppose that someone creates an artifact that looks much like a $20 bill with no intention of deceiving anyone by it. They're simply challenging their printing skills or maybe its "money" that will be used in a movie--in a briefcase filled with $20 bills.
But someone, years later, perhaps, gets hold of one of those bills, not knowing it is not a real $20 bill, and "spends it". The intention to deceive, by definition, has not occurred anywhere in this chain of events. I tend to think of that $20 bill as fraudulent. It is not what it appears to be. The object does not intend to deceive. Objects have no intentions.
My guess is that these index records of pre-1752 Quaker records are legitimate index records-- but they only appear to be reliably accurate reflections of the information found in the Quaker records. They are not actually what they appear to be in that respect. Just like the "counterfeit" $20 bill. I tend to think of artifacts that are not actually what they are intended to appear to be, as fraudulent even if no one intended them to deceive.
3. Your asside on the "commonly used calendar" is well taken. I think the reference here might be that there was not always consistency in starting the new year after March 25. Even some of the Quaker records pre-1752, started the new year in with January. It depended on the clerk of the meeting and they, I think, didn't always know what was going on. Off the top of my head, you'll see this in the early records of William Penn's Council too.
I might be fooling myself on all of this -- please do not hesitate to disagree with me -- actually if you do still disagree with me please try again to get me to realize where I am going off the track.
1 -
@RobertTest - if you have replacement sources, that's fine (in fact, excellent) so long as the replacement goes onto the profile before you start hacking the original about.
There is a wrinkle over what you do to the original attachment of the source. If the original source attached to the profile comes from FamilySearch's Historical Records (or images? Not sure) then it must remain attached to the profile. I suspect from the descriptions above that the originals are indeed FS Historical Records. If the currently attached source is an external source, held elsewhere (eg on another website, in a physical book, etc) then, when the time comes, it can be detached entirely from that profile.
If you should happen to detach an FS Historical Record from the profile, then the system will start looking for a profile that matches that record and, if it finds one, it will mark it up as a Hint against that matching profile.
You could detach the Historical Record only to see it immediately reappear as a hint against the same profile. Worst case is that it finds another profile that is the wrong one and hints against that. If you do happen to detach the Record and it reappears as a Hint, do not reject the hint as that will cause the system to look for another profile with similar data. Just accept it back against the profile but remove any tagged data. (If you know all this, then fine but just in case....)
So far as I know, the best strategy is to leave the obsolete source attached to the profile but remove the tagging that says this source provides data about the birth, etc. That way, you stop confusion in any checking routines if they check the tagged values in the sources against those on the profile, because there won't be any tagged values to check.
As an aside, it's a valuable point to make that dual dating (eg 1748/49) can be found in contemporary documents because lots of people thought that the New Year ought to begin on 1 January but made allowances for those who didn't, making things clear by recording both years. I suspect any clerk with Scottish ancestry might have done exactly that given that Scotland had used 1 January as the New Year since 1600.
1 -
Adrian Thanks for the added information — I just tried replying to Julia but the system wouldn't accept my message. I might not have been signed in.
Actually I think I'm going to give up on FamilySearch
0 -
Julia -- let me thank you for comments on my earlier posting. I had said that many of the indexed records of 17th and 18th century Quakers faulty interpretations of the Julian Calendar dates. My impulse is to remove those "sources" from profiles of Quakers when better sources are available to replace them.
You wrote: "You're letting the best be the enemy of the good if you're proposing to throw away data just because some of the dates may be off by up to fourteen months."
I don't think I'm letting the best be the enemy of the good. First, in general I don't think that derivative "sources" are any good—not as sources per se. They simply, either accurately or in many cases inaccurately, repeat the information found in another source. Repetition does not increase evidentiary value. Indexed sources are have one one main use: a guide to find the original source. They have a secondary use: for those not able or willing to search for original sources it gives them they can use as a poor substitute for original an original source of information.
Nor does disposal of indexed records amount to throwing away data. Nor is my reason to remove those indexed records because the dates are off by up to 14 months. Again indexed records have one purpose: to help find original sources. Once they are found and posted the indexed records have served their purpose and they should be removed.
For 17th and 18th century Quakers ALL of the index records contain the wrong dates. There are too many profiles with the single citation referencing the original source hidden in the middle of a list of 10 or more indexed records repeating the same wrong date.
But there is a much better reason to remove those indexed records. Nothing is more destructive to the truth that the proliferation of misinformation. Let me quote Elizabeth S. Mill's Evidence Explained: Citing History Sources from Artifacts to Cyberspace: "our goal is to rely only upon the best possible source. When we recognize that a source is deficient or that a better source might exist, the better source should be sought and used.
0 -
Oh, I agree fully that the original document (or at least an image thereof) is far better to have as a source than the finding aid, but this doesn't make it necessary or desirable to get rid of all references to the finding aid, no matter how faulty. This is especially the case if the document is more difficult to access than the index, be that because of a paywall or a more physical barrier. (I don't even have any American ancestors, never mind any Quaker ones, so I don't know what the access requirements are for the faulty indexes versus the documents/images. [What little I know of Quaker documentary practices has been mostly picked up via peripheral involvement with historical interpretation/reenactment in southeastern Pennsylvania.])
On the FamilySearch Family Tree specifically, one reason to leave the index(es) attached is that that's what the hinting system works with. Yes, as I've written somewhere around here recently, this is analogous to using only the card catalog at the library, without ever looking at the actual books, but until computers can reliably read handwriting, I'm afraid we're stuck with it.
1 -
The real gap here at present is our inability to safely and accurately correct record metadata on the fly.
The single metadata edit I have ever done, over a year ago, changing a persona's middle initial on the 1810 US census, still leaves that persona 'padlocked', invisible to searching, and absent from the index data shown for the relevant census image. Until that resolves, I shan't be touching index editing, and the DQS will report unnecessary inconsistencies for quite a few of my sources.
1 -
Hi Julia
Let me respond to just two things 1) on not removing redundant and misleading index sources and 2) the value of Quaker records. I'm not removing indexed records -- here's an example https://www.familysearch.org/tree/person/details/LH6X-LXQ -- it has 10 indexed sources for the marriage occuring on 29 March 1751. The indexers took 3rd month as March when it was actually May 29. So I created a memory document referencing the Quaker marriage certificate --including two images of small portions of the record. Posting the entire certificate violates copyright restrictions. And I also explain the basic difference between the Julian and Gregorian Calendars. If you want to see examples of Quaker documents you could visit my genealogy website on this page: https://testgenealogy.net/Genealogy/Test/2%20Generation/FrancisTest/frame.html This page is about my 6th great grandparents Francis Test and Elizabeth Bacon. It includes images of the recorded copy of their marriage certificate and the Quaker Birth Register of their children. It also includes images of references in the Minutes of his Quaker meeting to his earlier marriages. The recorded copies of those marriage certificates have apparently been lost. You might also be interested in seeing an image of an actual marriage certificate -- old one's are very rare- Francis Test's mother (Grace Woolley Lippincott Test) was originally married to Jacob Lippincott. Her signature appears on her sister 1688 marriage certificate available here: https://testgenealogy.net/Genealogy/Woolley/RuthWooley/frame.html
0 -
Hi MandyShaw1
Thanks for the reply — I've been thinking about this more closely as a result. So I won't touch indexed sources. I'll post a memory document with the corrected dates and with an explanation of the calendar reform. You can see an example at the profile for LH6X-LXQ
https://www.familysearch.org/tree/person/details/LH6X-LXQ
Would like to see what you think — is this still a bit over the top?
0 -
I personally like the way you've approached this, very clear without disruption to the main sourcing approach. Just make sure you aren't including any images in these memory documents that aren't public domain, or FS might take them down, and remember (as Julia said) that your efforts may get shared far and wide without attribution (in case you care).
0
