The British Surnames Down to Fewer Than Twenty People

Most surname writing is about scale. A hundred and six million Wangs, thirty-nine million Nguyens, a name covering a fifth of a country.

This is the other end. In 2011, around the completion of that year's UK census, MyHeritage published a study of British surnames surviving on very small numbers of bearers. Names with fewer than twenty holders were classed as on the brink of extinction. Names under two hundred were endangered.

Sallow was one of them. So were Fernsby and Tumbler.

What the words meant

Sallow is the old English word for a willow, so the surname started as a description of where a family lived. Tumbler was an acrobat or performer, which makes it an occupational name of the same type as Fletcher or Thatcher, attached to a trade that stopped being a recognised profession.

That is part of the pattern. A surname built on a job nobody does or a word nobody uses loses the thing that kept it comprehensible, and comprehensibility helps a name survive spelling drift. Nobody misspells Smith. Plenty of people misspell a name they have never encountered, and each misspelling that sticks removes a bearer from the original.

A caution about the numbers

Figures for individual rare surnames get repeated online with a confidence the underlying data does not support, and precise counts for specific names circulate widely without any source attached.

The counts depend heavily on the register being searched, whether variant spellings are folded together, and which year the snapshot came from. A name recorded as having twelve bearers on one electoral roll may have several dozen once you include a variant spelling that a clerk introduced two generations ago.

Treat the specific numbers as indicative. The pattern they describe is real and the individual figures are softer than they look.

The arithmetic that kills a name

Surname extinction follows a well-understood process, and it does not require anything dramatic.

In a system where surnames pass down the male line, a name survives only if enough men in each generation have sons who themselves have sons. A family with three daughters and no sons ends that branch of the name. Do that a few times in a row in a name carried by a dozen families and the whole thing disappears.

The mathematics was worked out in the nineteenth century by Francis Galton and Henry Watson, who were interested in exactly this question about aristocratic families. The result holds generally. Any name below a certain population is likely to vanish eventually, and rare names cross that threshold constantly.

Rising female surname retention and blended naming practices change the picture somewhat, though not enough to rescue a name down to single figures.

Why the pool keeps shrinking

Britain is doing at a slow pace what China did over several thousand years. Rare names drop off the bottom. Common names hold their share. The distribution concentrates.

There is nothing pushing in the other direction. Surname invention effectively stopped in Britain centuries ago, so no new names enter the pool to replace the ones falling out of it. New arrivals bring surnames from elsewhere, which adds variety overall without doing anything for the native names already at the bottom.

A name at twelve bearers is one generation of daughters from being gone, and once it is gone nothing brings it back.