Christenings
Hello, I refer to batches of christenings for UK Middlesex in general. Where a surname is listed for the child but not for the parents, the project instructions tell us to index the child's surname for BOTH parents. Many times only the father is shown, so do I index the surname for him ? As I have suggested previously, the instructions need to be updated on quite a few issues. Regards Jean.
Best Answers
-
I understand your dilemma, but even though we shouldn't index the gender field based on the name, there are lots of cases where we have to assume the gender in order to index the name in a particular field. That may not be desirable, but there is often no way around it.
For example, in the British parish records you will often find a male name spelled as Francis, while the female form will be spelled as Frances. So can you believe the clerk spelled the given names correctly? You have to index what you see, as best as you can determine it. You have a similar problem with marriages. I have seen the order of the bride and groom's names flipped in some records. If the bride's given name was Frances, but was where the groom's given name should have been, what do you do? Assume the clerk misspelled the name? Or seeing that the given name where the bride's given name should be, is a male sounding name, you index the male sounding name in the groom's given name field and Frances in the bride's given name field?
So if you have a baptimal entry, John, son of Frances Smith, do you go according to spelling convention, and assume this is a mother? Or what if you can't tell if the character between the "c" and the "s" is an "i" or an "e"? You have to make an assumption based on what you believe the gender of the name is.
So there are times you have to use your best judgement. But you do need to keep assumptions to a minimum.
My personal opinion, is that they should quit making exceptions to the normal rules in some projects and just index what you see. That would produce much more consistency across projects and probably improve the accuracy of the result. I keep saying that indexing is analogous to being a human Xerox machine. You should just copy what you see and let researchers interpret, deduce, and assume whatever details they want to based on comparison with other records. That is why people that have done genealogical research for years struggle with indexing -- they keep wanting to deduce, assume, and even correct what they see in the records and add information that isn't there or even change something they believe is an error as recorded in the record. You really have to wear a different "hat" when you index, when compared to when you do genealogical research.
In the cases we are discussing, if I were making the rules, I would just index the entry:
John Smith, son of Robert and Sarah as:
Child given name: John
Child surname: Smith
Child gender: male
Father's given name: Robert
Father's surname: <Blank>
Mother's given name: Sarah
Mother's surname: <Blank>
And if only one parent is listed, you would have to assume whether that listed parent is male or female. There is no way around that. But then you would still index the surname of the one parent as <Blank>.
Don't know if any of this helps, but those are my thoughts on the subject.
Jim
1 -
These changes to instructions still do not address the case when there is only one parent's given name and the surname is only associated with the child -- it still refers to BOTH and to plural "parents".
Someone needs to teach the people who make the instructions a course in logic -- If this, then do this, else do something else. And do it for each possible case. There are a comparatively small number of cases and it would be easy to do, although it would increase the length of the instructions, but at least it would be clear.
But I keep saying, in my opinion, it would be best and simpler if they just stick to "type what you see" and quit making exceptions to the General Indexing Guidelines.
2
Answers
-
It would be great to share the batch in question to more clearly show what you are writing about.
I am just a volunteer, but, I think the instruction is clear. It could be written again in the father and mother's surname field helps to make it even clearer. Maybe they didn't do this since it is an advanced project. In the olden days, even up until 1950 in the US, the mother's name often was not listed in a birth announcement. So, I never find it odd to only see the father listed. I suppose if the mothers were listed on the other records on the image, I would wonder if that mother had died before the christening. Anyway...
If a surname was recorded for the child, but not for the parents, index the surname of the child for both of the parents. This is an exception from the General Indexing Guidelines and should not be applied to other projects.
If only the father or only the mother is listed, then you would use the child's surname as their surname. For instance, if a record said: John Stanford, the son of James. Then James surname is also Stanford.
2 -
Hello, thankyou but are you sure because many of these batches have up to 100 entries and I do not want to index that many wrongly. A typical batch is as follows : 0080411034 00132, batch id - MSGB-8HC
0 -
I am fairly certain, but, perhaps we should tag a few people to see what they think. I'll tag @DHilary2 @Mirevo @Jim Hawker @barbaragailsmith1 @Dellory Matthews
0 -
Melissa is correct as we can see in both the Field Help and the Project Instructions.
This is the instruction in the Field Help, Mother's surname, bullet point 3,
"If only 1 surname was recorded for both parents, type the parents' surname in this field. If a surname was recorded for the child but not for the parents, index the child’s surname in this field. This is an exception to the General Indexing Guidelines."
This is the instruction from the Project Instructions, What to Remember about this Project, bullet point 6.
"When indexing baptism, birth, christening, burial, or death records, if a surname was not recorded for a child and only 1 surname was recorded for both parents, type the parents' surname in the Surname field for the child and both of the parents. If 2 different surnames were recorded for the parents, type the father's surname in the child's Surname field. If a surname was recorded for the child but not for the parents, index the child’s surname in the Surname field for both of the parents. If a surname was listed only once for both the mother and the father, index the surname for both parents. This is an exception to the General Indexing Guidelines."
2 -
Hello,
Yes, I agree. In this instance the father would be given the same surname as the child.
Hilary.
0 -
I have been working on Middlesex projects for almost two years now and I understand your frustration with the instructions.
A strict reading of the instruction to transfer the surname from child to parents, when it is only listed for the child, only applies to the case when BOTH parents are listed, not when only one parent is listed.
So my own convention for when a surname is only listed for the child is this:
1) If there is only a father listed, index the surname of the child in the father's surname field.
2) If there is only a mother listed, DO NOT index the surname of the child in the mother's surname field. Just leave the mother's surname blank. The reason for this is that you really don't know if the surname of the child is the mother's surname or the supposed father's surname. And indexing it this way, maintains the main general rule of indexing -- "to index what you see". And it doesn't violate the project/field instructions, because they don't apply to this case, since BOTH parents are not listed.
3) If there are both parents listed, do as the project and field instructions indicate, and transfer the surname of the child to each of the parents.
So I only review batches now, and there is no way to guarantee that the reviewer of your batches will follow the conventions above. They may transfer the child's surname to the mother, when only the mother is listed. Or they may even blank the father's surname when only a father is listed.
Good luck.
1 -
Thank you, Jim. I recalled that you had a theory about this when we had Indexing Chat. So sad that we can no longer search for those multitude of great answers!
So, I was half on target - Index the Child's surname as the father's surname in cases when his is the only name on the record. The only problem I have with this is that if we can't use given names to determine gender, then using that rule, we don't know if it is a mother or a father, i.e., Mary and James are just names.
John Riley, son of James
John Riley, son of Mary
Then they allow the transfer of a Surname from a child to BOTH parents, but, it doesn't work in reverse to transfer the surname of BOTH Parents to the child. Yikes!
For instance, the examples:
William, Son of Samuel & Elizabeth Walker, and yet they don't index Walker for the child's surname.
I do hope that there is a official final answer, with an explanation, provided on this subject - someday.
0 -
- If a surname was recorded for the child but not for the parents, index the surname of the child for both of the parents. This is an exception to the General Indexing Guidelines and should not be applied to other projects.
This is an update just today for this question. The examples in What to Index have also been changed.
4 -
This is such an interesting conversation. The whole question reminds me of computer coding when the rules you have set up (which work just fine for the majority of cases) are applied to some non-standard data. You can have some unexpected or illogical results.
And of course the problem of developing appropriate project instructions is made so much harder when the project spans so many years and therefore cultural practices.
On @Jim Hawker's point about being a being a human Xerox machine, I would only say that a researcher first has to find the document before they can examine and use the data and so indexing needs to enable that. If the index is too sparse then the document may never be found, particularly for individuals with common names. Of course a good search algorithm can overcome that to some extent, by making the sorts of assumptions that are driving the project-specific exceptions. So one way or another...
What I thnk about when I am faced with this type of ambiguity is how to follow the instructions and make the document findable while making the fewest assumptions (a kind of Ochum's razor approach).
Thanks for the stimulating Saturday morning reading - if anyone wants some happy diversionary viewing the Peregrine falcons on one of Melbourne's highrise buildings have newly hatched chicks https://www.youtube.com/watch?v=Un8f85yADAU
0 -
Well I happen to write software for a living, so that is why, if they want to insist on make exceptions to the Usual Indexing Guidelines, the exceptions should be explained in "if, then, else" types of conditional logic statements. But they don't seem to want to do that.
2 -
The project instructions and examples were changed today for the Herefordshire parish project, but not this Middlesex parish project.
https://www.familysearch.org/indexing/batch/19e70629-b999-4ff5-a154-eba39fba76e0
MS2Y-PHX
Now the field help for the Mother's Surname and the Father's both say:
If a surname was recorded for the child but not for the parents, index the surname of the child for both of the parents. This is an exception to the General Indexing Guidelines and should not be applied to other projects.
I think that pretty much covers the bases to index the child's surname for each parent. The instruction has been removed from the Child's Surname Field Help.
However, the Middlesex parish records remain the same with that instruction showing up in the Child's Surname field and not in either parental surname field.
I agree with you, Jim, there is not a lot of logic behind some of the instructions. Keeping It Simple Sweetheart just doesn't happen... and it should. This is especially true when reviewers are now indexers who have indexed 1000 records of any project. One batch of City Directories with 1200 names, and possibly a place and a date, and many blank fields and "poof" you are a reviewer.
1 -
@Jim Hawker that is pretty much spot on - having twizzled my brain doing "Intro to Matlab" I wouldn't want to have to write those rules that will cover all contingencies. But a more logical approach that discloses the underlying principles would be helpful (perhaps there is an article somewhere that I have not read)
@Melissa S Himes I also agree - I wonder how the new process stacks up against the old arbitration process. Garbage in/garbage out will eventually declare itself but surely someone evaluated the process before changing it.
0 -
@Ontymay I think it was a very well laid-out plan and well-investigated. I know the BYU computer scientists tested the systems in a field study, probably in 2012, and published their findings. As I recall, a single peer review wasn't quite as effective as arbitration for some fields. The paper suggested a double peer review would work better than the single one that was tested. Which probably led to the current process for triple peer-review, and an in-house arbiter, if necessary. I'm guessing the computer science and indexing departments evaluate the process frequently, if not daily.
1 -
Hello, as helpful as you all have been, the bottom line is that there is no DEFINATE specific answer to this. Until I actually see the exact answer to this question in the project instructions, I am still wary of indexing these types of batches. If the reviewer decides against my following your advice, who's to say who is right and who is wrong ? I am finding this very annoying now. Regards Jean.
1 -
How about this question for surnames?
UK, England, Northumberland—Nonconformist Church Records, 1613–1920 [Part B][MSPS-M5Z]
So, in the first record, George [Blank] is the child, Thomas Bell the father, Phillis [Blank] the mother. Or does the fact that the record clearly places her as the daughter of William and Mary Collinson give sufficient evidence to give her the maiden surname of Collinson? (In other words, does this juxtaposition carry the same weight as parentheses or "born as", etc.?)
And, secondarily, is there sufficient information that one could add a baptism/birth entry for Phillis [now blank because it's a birth record without a provided surname] as the child, William Collinson as the father, and Mary Collinson as the mother, with no dates, of course, because none are provided?
How about that one? And I have more when this one is answered....
0 -
"George, the son of Thomas Bell of Middleton, in the Parish of Middleton in the county of Durham, and of Phyllis his wife, who was the Daughter of William and Mary Collinson was born at Middleton on the second day of November... "
According to the parental field helps, Do not assume a surname from the surnames of others mentioned in the document. Unless the child has a surname listed then it can be used for both parents.
So, the mother would not have the surname of Collinson. If the sentence was his wife Phyllis Collinson, the daughter of William and Mary, then a surname could be indexed for Phyllis.
We would not create a baptism or birth entry for anyone in the entry except the person for whom the record was created - George. The mention of Phyllis' parents is not an event.
0 -
I agree with what you are pointing out here, Melissa. But there are plenty of indexers out there that would index the surname of Phyllis as "Collinson Bell". For example, in the Middlesex, Part A project, the field instructions for the mother's surname for a baptism says:
"Do not assume a surname from the surnames of others mentioned in the document.
Maiden names and former surnames should be typed in this field before the current surname. A maiden name or former surname was often indicated by the words "née," "formerly known as," or "born as." A maiden name was also sometimes written in parentheses."
The instructions for the same field in the Northumberland project is essentially the same - they just reversed the order of those two paragraphs.
Despite the first sentence, indexers will often put "Collinson Bell" in the mother's surname field for the example you show because of the second paragraph about maiden names. Or they also might put "Bell" as her surname. Most indexers just seem to have a need to put something in a field besides <Blank> if there is any way they can justify it. But this is yet another case where the instructions could be clearer. Why isn't the statement that Phyllis is the daughter of William and Mary Collinson as good a justification as the words "née," "formerly known as" or "born as" or "A maiden name was also sometimes written in parentheses".
Once again, if I were making the rules, I would index it as it is written:
Child's given name: George
Child's surname: <Blank>
Father's given name: Thomas
Father's surname: Bell
Mother's given name: Phyllis
Mother's surname: <Blank>
0 -
That is exactly how it should be indexed, Jim, and would be following the rules and the field helps. This is an advanced project. However, I agree that the instructions need to be clearer, especially since the new program doesn't really give indexers a lot of diverse experience before jumping into the review process.
It seems that FamilySearch lost the guidance article from nearly a decade ago on how to index maiden names. It was a beautiful instruction on how to index womens' names and what was necessary to identify a maiden name, like the woman's surname had to match the parental surname, or the words nee, born as, etc. were on the record.
If we were to look at the various was of indexing Phyllis's name:
Phyllis Bell (Collinson) is indexed as Given Name: Phyllis Surname: Collinson Bell. (Parentheses around Collinson indicate maiden name).
Phyllis Bell née Collinson is indexed as Given Name: Phyllis Surname: Collinson Bell (obviously, née being born as).
Phyllis Collinson Bell, the daughter of William and Mary Collinson is indexed as Given Name: Phyllis Surname: Collinson Bell. (Because the middle name matches the parent's surname).
Phyllis Collinson Bell, the daughter of William and Mary is indexed as Given Name: Phyllis Collinson and Surname: Bell (Because we don't know if Collinson is a given middle name or a surname because there is no surname for the parent).
Phyllis, the wife of Thomas Bell, or Thomas Bell, and his wife, Phyllis, are indexed as Given Name: Phyllis Surname: <blank> (Because we can't assume that Phyllis has taken her husband's name - even based on the timing of the records.) UNLESS the project instructions tell us differently.
So, on this batch, every other record needs to be deleted since we do not create a birth record for the mother of the child and probably the child's surname and the mother's surname have to be deleted on each record. (I only read the first three).
The instruction that leaves me Shaking My Head, it the County: "Do not index the county or city that may have been recorded with the parish." The child is born in the parish of Middleton in the county of Durham. That is correct, the parish of Middletown is in the County of Durham. So why would the Field Help say not to index the county where the child was born? I don't get it.
The instructions need to be reviewed for both of these projects. And it would be fantastic if there was a General Indexing Guideline just for Indexing Maiden Names, unless it is no longer a priority to get those in the correct fields.
0 -
Thank you very much for your wise counsel. I started this project doing exactly as Jim said. Then started rethinking it, called "Help", who said maybe I should do a birth record for the wife. And they didn't have any suggestions for the county question, except to submit a question here.
I will indeed delete the maiden names and the birth records for the wives. With pleasure.
The county question still has me scratching my head..... Seems like one should be able to index the county name.... but, you know, I try really hard to do just what I'm told. :)
1 -
Actually, the information "of the parish of Middleton in the county of Durham" was referring to where the parents are located, not the place of birth, so that particular question about whether or not to redord the county is answered for me.
I have already submitted that particular batch, corrected as noted in my last post, but the question about the county is found more glaringly in this batch, which I have not already submitted....because I'm hoping someone may respond to with a definitive answer that could alleviate head-shaking and head-scratching.... :)
UK, England, Northumberland—Nonconformist Church Records, 1613–1920 [Part B][MSPS-964]
0 -
record, not redord
0 -
Hello Trudy,
I would say the County in the last batch you mentioned should be indexed as "Durham". The instructions tell you not to include a County that is linked to the parish, so if it had been written as "Bishopwearmouth, Durham" or "Durham, Bishopwearmouth" where it says "in the Parish of" you wouldn't put "Durham" in the County field (no idea why you can't!) . For some reasons, also, you can't put the name of the parish, town or city either, which would be quite helpful to a researcher, I would think.
Interestingly, the name of the child has been left BLANK as it tells you you can give a child's surname to the parents if not mentioned, but not to assume a surname from anyone else on the record otherwise, so that is correct in my view. However, if you look at the example given in the Project Instructions for this project, under the heading "What to Index" the example given for a Baptism Register, just like this one, has the surname indexed for the child :
Baptism Year: 1837
2. Baptism Month: Dec
3. Baptism Day: 3
4. Child's Given Names: Elizabeth
5. Child's Surname: Huffer
6. Child's Sex: Female
7. Father's Given Names: Robert
8. Father's Surname: Huffer
9. Mother's Given Names: Alice
10. Mother's Surname: Huffer
11. Birth Day: 15
12. Birth Month: Nov
13. Birth Year: 1837
This is surely contrary to the instructions and not at all helpful!
Hilary.
1 -
They need to take down every one of these parish register projects and review and compare the project instructions, the field helps, and the examples. Parish registers are basically the same. There shouldn't need to be instructions for Middlesex, instructions for Herefordshire, instructions for Northumberland, etc. There needs to be general (and specific) indexing guidelines for indexing Parish Records, just like there are for indexing Obituaries. The same applies to Naturalizations which are federal documents, they deserve ONE set of guidelines.
1 -
You are preaching to the choir! But I doubt they will do it in the middle of the projects being indexed. But they should at least redo them for new projects when they are published.
1
