One of the most persistent challenges in Welsh family history research is not a lack of records, but a misunderstanding of how names worked before fixed surnames became the norm. The Welsh patronymic naming system, once perfectly logical to contemporaries, is frequently misinterpreted by modern indexing systems — sometimes with dramatic consequences for genealogy.
A recent discovery in the parish registers of Selattyn provides a clear and troubling example of how this happens, and why researchers need to treat indexed records with caution.
For centuries, Wales did not use hereditary surnames in the modern sense. Instead, individuals were identified by a patronymic formula, most commonly:
[Given name] ap [father’s given name]
The word ap simply means “son of”. For daughters, ferch (“daughter of”) was used. Under this system:
So:
Over time, many of these constructions solidified into surnames — ap Richard becoming Pritchard, ap Hugh becoming Pugh, and so on — but this process was gradual and uneven, especially in rural areas and border parishes.
The parish of Selattyn lies today in Shropshire, England, close to the Welsh border. Historically, however, it sat firmly within the Welsh ecclesiastical world.
This matters. Border parishes like Selattyn preserved Welsh patronymic usage long after fixed surnames became common elsewhere in England — making them especially vulnerable to misinterpretation by modern transcription projects.
An indexed baptism record on Findmypast appears as follows:
At first glance, this already raises questions. If the child’s name is Rich ap Rich, why is the father named David?
The answer becomes clear when the original parish register is examined.
The entry itself reads (in modernised spelling):
“Rich the son of David ap Rich by Elyn his wife was christened.”
This is a perfectly normal Welsh-style entry. The structure is clear:
The indexed version has effectively skipped a generation — assigning the child his grandfather’s name as though it were a fixed surname.
A similar error appears in another Selattyn baptism just a year earlier.
An indexed transcript presents the following:
Once again, the indexed name contradicts the named father. Under Welsh patronymic rules, Edward ap Robert would mean Edward, son of Robert — yet the father recorded in the same transcript is Thomas.
As with the previous example, the original register makes the structure clear. The wording follows the same pattern found throughout this register, identifying Edward as the son of Thomas ap Robert, not a child surnamed “Ap Robert”.
Correctly interpreted:
The index has again detached ap Robert from the father and incorrectly attached it to the child, creating a false patronymic and obscuring the true parent–child relationship.
This kind of error is not accidental. It happens because modern transcription systems are built around assumptions that do not apply to Wales:
In Welsh patronymic records, the grammar of the sentence matters more than the order of words. When that grammar is misunderstood, the result is not just a typo, but a broken lineage.
The consequences for researchers can be serious:
And this is not an isolated example. In Selattyn alone, there are hundreds of similarly transcribed entries, each quietly nudging family trees in the wrong direction.
Welsh genealogy does not fail because the records are poor — quite the opposite. Parish registers are often detailed and consistent. The failure lies in applying English naming expectations to a system that works entirely differently.
Large genealogy platforms perform a valuable service, but they cannot replace historical understanding. When indexes are accepted uncritically, especially in Wales and the Welsh Marches, they risk obscuring the very relationships they claim to reveal.
If you are researching Wales (or the Welsh Marches):
Most importantly: let the record speak in its own historical language.
Welsh patronymics are not a complication — they are a gift. They tell us exactly who someone’s father was, if we are willing to read them correctly. When we do, the past becomes clearer, not more confusing.
Linda
EYEonthePAST
Record images and transcript excerpts are reproduced here in small part for the purposes of research commentary, criticism, and historical explanation. All original records remain the property of their respective owners.
