How can I distinguish between name changes and missing records?

Every now and again I come across a person, who seems to have changed their name. For example, a birth/baptism record may have one name, while subsequent records (marriage, census…) use another for the only child of that age in the family. It’s a fairly uncommon thing but not always remarkable (exchanges of first and middle names aren’t that unusual, for example). But I have found a family where one child may have had up to three different names. This looks like a big red flag, but then so does the alternative – a stack of missing records – and I can’t find a clear way around it either way.

David Hill seems to have resided in the Gloucestershire village of Little Rissington for his entire life. He married twice, and had three daughters (Mary Jane, Amy Gertrude and Emily Florence) and up to three sons (Thomas, John and Charles). For his daughters, the records are mostly present as expected:

• each has an entry (or good candidate) in the GRO births index

• each has a baptism recorded in the parish church records

• Emily Florence died young, leaving a single census entry, a GRO death index entry and a parish burial record

• Mary Jane and Amy Gertrude both married in the parish church and have corresponding GRO marriage index entries

• Amy Gertrude Hill/Ferriman is found in Little Rissington records until her death in 1937

• Mary Jane Hill/Davis probably entered the Stow on the Wold workhouse around 1881, and I think she died there in 1892.

For David’s son(s), things are more problematic. In chronological order, I have:

1) 01 Nov 1857 – Thomas Hill born in Little Rissington to David and Jane, according to birth certificate from the GRO. (Thomas’ BC was filed under the wrong registration district, as asked about in another question). No GRO birth records are apparent for John Hill or Charles Hill.

2) 25 Dec 1857 – John Hill baptised in Little Rissington to David and Jane (two records on Ancestry – register and copy). No baptism records are found in the Little Rissington parish records for Thomas or Charles Hill.

3) 07 Apr 1861 – Charles Hill (aged 3) in Little Rissington census (Class: RG 9; Piece: 1789; Folio: 84; Page: 6 on Ancestry) with father David, sister Mary Jane and David’s sister-in-law Mary Ann Harris. (David’s wife Jane had died in 1859.)

4) 02 Apr 1871 – Thomas Hill (aged 13) in Little Rissington census (Class: RG10; Piece: 2654; Folio: 32; Page: 7 on Ancestry) with father David, stepmother Mary Ann and sisters Amy Gertrude and Florence.

5) 24 Dec 1880 – Thomas Hill killed Samuel Palmer after an altercation at a pub in Bourton on the Water (just up the road from Little Rissington). Thomas had been living in Bradford, Yorkshire for “several years” before this.

6) 14 Feb 1881 – Thomas Hill was sentenced to life imprisonment for manslaughter and was sent to Pentonville prison in London to begin his sentence.

7) 03 Apr 1881 – David, Mary Ann and Amy Gertrude, plus niece Selina Bolton, in Little Rissington census (Class: RG11; Piece: 2562; Folio: 102; Page: 22 on Ancestry) – no son(s) present. Thomas seems to be in the Pentonville Prison census (Class: RG11; Piece: 248; Folio: 104; Page: 33 on Ancestry).

8) 05 April 1891 – David, Mary Ann, with married daughter Amy Ferriman and her family in Little Rissington census (Class: RG12; Piece: 2037; Folio: 62; Page: 17 on Ancestry) – no son(s) present. Thomas is in Chatham Prison (Class: RG12; Piece: 665; Folio: 103; Page: 7 on Ancestry). Prison records show him transfer to Portland Prison later that year, and then I lose track of him.

The records I have found for the family are tabulated below. Birth dates in parentheses are inferred from baptism or census records (and so are not reliable), otherwise are from GRO records or candidate entries (I have certificate copies for Mary Jane and Thomas, but not Amy or Emily). Grey-shaded records relate to the Hill son(s), with bolded entries being confirmed records (just for clarity). Pink-shaded cells indicate missing records that we might expect to find for three sons.

So three sons of David (and Jane) are named, and have similar birth years, but no two appear in the same census. Thomas was born in Nov 1857, but less than two months later John was baptised. John was never recorded again, and Charles also appears only in a single record. All three sons have similar known or inferred birth dates. No local marriage, death or burial records are found for any of them. The common names don’t help, of course, and make it difficult to trace anyone who left the village. But from this table, up to twelve records are “missing”.

David does not appear to have left a will, and I can’t find any other likely records such as Thomas/John/Charles as marriage witnesses in Little Rissington.

If we merge the three sons into one individual, we end up with a complete, duplicate-free record set:

That’s a nice, simple model, but a fragile one: a single extra record would shatter it. So the impetus is to seek that extra record, or to explain why none will be found.

There are a few possibilities I can think of:

• Births of John and Charles may not have been registered (although all others were). Or their births may be registered in the wrong district… Both are possible, but I have looked at the Stow and Northleach indices and found no good candidate births for either. Thomas’ record filing means that I can’t fully trust the GRO index for this place and time, but I think it’s unlikely that too many more records would be misfiled into yet another registration district. Unregistered births would explain two missing records, but not the rest.

• Two of the boys could have died in infancy. But which two? There are no good GRO death or local burial record candidates. And this doesn’t offer a satisfactory explanation for the apparent name changes.

• Thomas and Charles may not have been baptised in the parish church (although again all others were) Also possible, although I’d expect consistency with this. They weren’t baptised in the neighbouring parish church of Bourton on the Water either. Further afield is perhaps possible but hard to trace with such common names.

• There could be more than one David Hill and family, and perhaps they moved around. There were quite a few Hill families in Little Rissington through the mid-late 19th century, and most or all of them were related. But I can find only two baptisms for “David Hill” and one of those died in infancy. There don’t appear to be any other David Hills in the census records at that time either, so I don’t think there is any chance of mixing up two similar families.

• Records have been lost. The Little Rissington parish baptism, marriage and burial registers look pretty complete for this timeframe. They are available as images on Ancestry, with no obvious missing pages or illegible entries. The baptism record for John appears in the original ledger and a transcript ledger, which do not appear to differ from each other significantly in this date range. There do not appear to be any missing census pages.

• The boys may have been sent away to live with other families – David being widowed with young children must have been a difficult time. This may be the best area for more research: the presence of Mary Ann Harris in 1861 does suggest familial support after Jane’s death. I have not filled out the families of Jane or Mary Ann, so there could be records showing Hill boys with Jane’s relatives, for example. And Thomas had been living in Bradford, Yorkshire for some time to 1881, after all, although I don’t know why he went there or if relatives were involved. I have looked for Hill boys in Jane’s home town of Shipton under Wychwood, and found nothing. But the family could have moved/spread around. It’s also possible that the boys could have been with several related Hill families in Little Rissington – I can’t see any obvious candidates but haven’t cross-checked every single entry yet.

Some of these options would explain missing birth/baptism records, while others would explain missing census entries. None explain both, so more than one would have to be in effect. Hence the appeal of the “One Son, Many Names” model – it explains everything in one go. But it is very hard to prove the negative of “no more records to be found”.

So are there any further strategies to consider? (Or am I asking for too high a degree of proof?)