Irish Baptisms: Get the child's surname from the father.
The instructions say:
- When indexing Ireland baptism and marriage records, if the principal's or spouse's surname was recorded and the parents' surnames were not, determine the parents' surnames from the principal's or spouse's surname. If the father’s surname is not recorded and the mother’s surname was, determine the father’s surname from the principal’s surname and index the mother’s surname as recorded. This is an exception from the General Indexing Guidelines and should not be applied to other projects.
I've understood this to include "If the child's surname is not recorded, then determine the child's surname from the father's surname."
However, some of the indexing that I'm reviewing shows "Blank" for the child's surname. I've been "correcting" this by pasting in the surname from the father.
Am I correct?
Answers
-
Good morning @BruceLaidlaw
Did you notice the instruction pop-up when you opened the batch?
And this thread - https://community.familysearch.org/en/discussion/169710/irish-baptisms-infants-surname-is-different-than-fathers-surname
0 -
Áine, I think Bruce is talking about the other direction: apparently, he's been reviewing batches where the child's (i.e., the principal's) surname isn't recorded. The question is, is the instruction to fill in the father from the child commutative? That is, can it or should it be turned around to filling in the child from the father?
1 -
Without the batch number, I wouldn't dare to guess. The style varied from priest to priest.
0 -
The records typically state:
"Baptised John, son of Michael Reynolds and Bridget Connelly."
I write the child's surname as "Reynolds" from the father. Some indexers are marking the child's surname as <blank>.
From the pop-up instructions, I suggest my indexing is correct.
0 -
The usual rule in FS indexing is not to assume anything: don't conclude a sex based on the name, don't fill in surnames based on family members, don't expand abbreviations, don't correct spelling. Index what you see.
Now consider these Irish records. In addition to the usual "X, son of YZ and MN", you also get "XZ, son of Y and MN". (And similar arrangements in marriage records.) In other projects, the first one would be easy: principal = [X] [blank], father = [Y] [Z], mother = [M] [N]. Depending on how Anglophone the project was, the second could either be [X] [Z], [Y] [blank], and [M] [N], or [X] [Z], [Y] [N], and [M] [N]. I think in order to explicitly avoid that second interpretation (which is erroneous for these records), the instructions for this project make an exception: that second formula is to be indexed as [X] [Z], [Y] [Z], and [M] [N]. But the instructions make no mention of that first formula, so it's not at all clear whether they intend for that blank to be filled in or not.
My long-time-indexer reflex is to not fill it in, that is, to consider the operation to be non-commutative: we've been told to fill a parent's name from a child's, but that doesn't necessarily mean that we should fill a child's name from a parent's. It appears that others share my instinct, given the batches you've gotten with the child's surname blank.
However, the fact that others appear to agree with me doesn't necessarily mean that I'm right. @Ashlee C., any chance of Word From On High on whether this project's exception is meant to be commutative or not?
1 -
In those days (1850) in Ireland the child's surname, in a regular marriage, would always come from the father.
0 -
This is the kind of issue that makes me so unhappy about the way FamilySearch's indexing projects operate.
Firstly, is the argument I commonly use about indexing needing to act as a finding aid, rather than later hinder a researcher in finding their relative.
From personal experience in consulting original parish registers, (and an example I have previously quoted) is where there is inconsistency in the format in the original recording a baptism event. So, for example, on consecutive pages of a register of a small parish one might find something like:
17 June 1793 John to William and Mary Blenkinsop
20 July 1794 Charles Blenkinsop to William and Mary
You might also have found a matching marriage in the same parish for William Blenkinsop and Mary Wilson, say in 1792. It would be highly unlikely, in these circumstances, that we are not looking at the same couple, who had children baptised in the same parish in consecutive years. However, I believe according to the usual / generic project instructions, the first event should be indexed (exactly as seen) with just JOHN being indexed as the child's name and BLENKINSOP only applied to the parents. Whereas, in the second example, the child would be indexed (exactly as seen) as CHARLES BLENKINSOP.
The problem here is that, using the family surname (last name) in a search on FamilySearch, would mean John would not be produced in the list of results, but Charles would. This would give the impression that William and Mary just had the one child baptised in the parish, during this period.
However, if I have to concede this is FamilySearch's way of indexing / producing records, at least I should be confident there is some consistency with these practices. However, it seems not so when it comes to these Ireland batches. I just cannot see any justification for the acknowledged comment:
"This is an exception from the General Indexing Guidelines and should not be applied to other projects."
Not only is this an acknowledged "breaking of the (general) rule", but the overall instructions are also so unclear that different indexers have seen it necessary to come here to gain clarification of the precise way in which the names are meant to be indexed!
For the sake of indexers and the researchers (who will eventually encounter these records through the FamilySearch search engine), please let there not be a continuation of such ambiguous and, in this case, what seems to be arbitrary instructions, which do not conform to generic project instructions.
3 -
@Paul W, you've illustrated one of the problems with trying to make indexes consistent: the records themselves are vastly inconsistent.
An indexing project's instructions need to be applicable to all of the documents in the collection. If some of the registers used the "baptised X of Y and MZ" formula, while others used "baptised XZ, child of Y and MN", what do you tell indexers to do with Y's surname? The indexer is looking at a small number of pages — sometimes just one; that may or may not be enough to determine what the formula is.
I've encountered a fun one in Hungarian civil registers of death, which in the later tabular formats have a single column for "parents". Normally, if there's just one name in that column, it's because the decedent was illegitimate, and the one name is the mother's. Except when it's the father's, because the informant didn't know the mother's name. Usually in the latter case they explicitly wrote in "unknown" below the father's name, but sometimes, they skipped that step. I wasn't indexing when I encountered this, and the record wasn't indexed, so I have no idea what the instructions would've had to say about it.
My (admittedly untrustworthy) recollection is that FS's Search function used to be a lot smarter about this: if you searched for John Blenkinsop, it'd show results for John, son of William Blenkinsop. I don't know why it no longer does so, especially given how its current behavior so directly contradicts FS's indexing structures/instructions.
2 -
I am coming to the conclusion that it's next to impossible to be certain what the names of the parents and child are in all possible circumstances.
Take, for instance, my grandpa's baptism in 1896. This is on pre-printed paper with column headings and, as a result, this version is pretty clear.
Child's Christian name: Jack son of
Parents' Name: George Slimon & Mary Ellen
Surname: BruceHis parents were (in full) George Slimon Bruce and Mary Ellen Bruce). But take the purely theoretical version of 100y earlier which would be written on plain paper. Suppose it had
Jack son of George Slimon & Mary Ellen Bruce
Is his father George Slimon? Or is it George Slimon Bruce? (Slimon is a Scots surname, by the way). It depends on whether you think the clerk is using the typical English formula for the names of the parents of an illegitimate child (so father is George Slimon) or whether you think the father has a middle name (i.e. 2nd given name) of Slimon and surname of Bruce.
This isn't just theoretical, I have seen similar issues, especially where the middle name is one that can also be used as a surname (e.g. Wesley, Seymour, Howard, etc. Even Bruce!). When I did have those issues, it was a case of doing research into whether there were other baptisms to those potential parents under various names - what formula was used then? - and was there a marriage betwen the parents?
Neither indexer nor software can be expected to do such research, so somehow the searches need to throw all possible surnames into the mix.
2 -
There will always be the occasional confusion, but we need a standard way of dealing with the usual practices. From the Irish baptisms that I have seen, the usual practices are either as:
Baptised John Moran, of Patrick and Bridget O'Hara. (Determine the father's surname from the child, as the instructions say. John Moran, son of Patrick Moran and Bridget O'Hara.)
or, more usually,
Baptised John, of Patrick Moran and Bridget O'Hara (The instructions don't specify, but they should say, "Determine the child's surname from the father." John Moran, son of Patrick Moran and Bridget O'Hara.)
Descendants searching for their ancestors need to have the surname there. Marking it <blank> helps no-one.
1
