Researching a surname

How British surnames arose, the four main types, why spelling varied until the 1800s, how to use surname distribution maps to find a family's home county, and one-name studies.

Everyone wants to know what their surname means, and the answer is usually mundane: your ancestor's father was called Richard, or he lived by a hill, or he made barrels. The more useful question for research is where the name comes from and where it was concentrated, because a surname's geography can tell you which county to search when the records go quiet.

How British surnames arose

Hereditary surnames spread through England between the 1100s and the 1400s, later in Wales (where patronymics like "ap Hugh" persisted into the 1700s and became Pugh, Hughes, Price) and in parts of Scotland. They fall into four broad types:

A meaning tells you about one medieval person. It says nothing about whether two families with the same name are related; Smiths arose independently in every village.

Spelling didn't matter until it did

Before general literacy and civil registration, a surname was a sound, written down by whoever held the pen. The same family can be Clarke, Clark and Clerk across three generations, or Bowden, Boden and Bawden across one parish register. Fixed spellings only became usual in the mid-1800s. For research this means: search every plausible variant, use wildcards, say the name aloud in the local accent, and never dismiss a record because of one letter.

Surname distribution: the research tool

Most surnames were still concentrated near their point of origin in 1881, before railways and the twentieth century scattered everyone. Mapping where a surname occurred in the 1881 census (or in modern data) often shows a single county or a cluster of parishes. If your ancestor appears in Birmingham in 1851 with a name that is 80 per cent Shropshire in 1881, you know where to look for his baptism. Free tools include the surname maps on the Great Britain Family Names site (gbnames.publicprofiler.org), the National Library of Scotland's Scottish surname maps, and the distribution views built into Ancestry and Findmypast. The 1881 census on FreeCEN and FamilySearch lets you count a surname by county yourself.

Dictionaries and their limits

The authoritative reference is the Oxford Dictionary of Family Names in Britain and Ireland (2016), which gives origins and historical distributions for some 45,000 names and is available through many library services online. Older dictionaries (Reaney and Wilson, Bardsley) are useful but sometimes guess. Websites that sell "your family crest" are selling a coat of arms granted to one individual of that surname, not to the name; arms belong to a person and their descendants, not to everyone who shares the name.

One-name studies

Some researchers collect every occurrence of a surname everywhere, reconstructing all the families that carry it. The Guild of One-Name Studies registers these; if your surname is registered, the study's holder may already have your family worked out and will usually help. Rare surnames (a few hundred bearers) are particularly suited to this, and to Y-DNA testing to see which branches share a male-line ancestor; see DNA tests.

Using a surname well

• Rare surname: search the whole country; collect every instance and build families from them.
• Common surname: rely on first names, places and ages; the surname alone identifies nobody.
• Welsh patronymics: expect the surname to change each generation before about 1800, and search by first name and parish.
• Scottish and Irish names: watch for Mc/Mac and O' being dropped and resumed, and for anglicisation (Ó Briain to Brian to Bryan to Brien).

Take the geography back into the parish registers and the census, and the surname becomes a pointer rather than a puzzle.