Continuing the discussion from New localized options for surname guessing:
If I understand the feature request correctly, the surname guessing will use the systems locale setting (‘DE’, ‘IS’ etc) for selecting which rule to use. I’m not particular fond with this, since genealogy will often involve several countries and thereby also several different surname guessing rules. Instead I suggest that the surname guessing should be based on Place typically from an event e.g. birth or census.
Here’s a couple of ideas for surname guessing:
- In northern Europe patronymic naming was widespread a couple of centuries ago, so the name guessing should cover all the countries - not only the few mentioned in Feature Request 13953.
- The patronymic naming changed slowly over time. For example in Denmark changing from patronymic naming to inherited family name became common in large towns in 18th century, while patronymic naming continued until late 19th century in rural areas.
- Therefore a rule for surname guessing should take the father’s or mother’s surname origin value into consideration. If the origin value is ‘Patronymic’ then the surname guessing will use patronymic surname guessing rules for a child, otherwise the surname guessing will copy the father’s surname to the child surname.
- The patronymic surname guessing should also know about exceptions (maybe user defined), Again as an example from Denmark: The general rule is to add ‘sen’ or ‘sdatter’ to the father’s given name e.g. given name ‘Poul’ becomes ‘Poulsen’ or ‘Poulsdatter’, but if the given name ends with ‘s’, then only ‘en’ or ‘datter’ is added - ‘Jens’ becomes ‘Jensen’ og ‘Jensdatter’. Further there are some special cases - given name ‘Ole’ becomes ‘Olsen’ or ‘Olsdatter’ but not always, and given name ‘Jeppe’ can either become ‘Jeppesen’ or ‘Ipsen’