Interpreting family names

Introduction

Family names or surnames are notoriously difficult to interpret, and individual names may well not be what they seem. For instance, Lamb might have originated with a man who was known by a pet form of Lambert, or who was the son of Lamb, or who was lamb-like, or who lived at the sign of the lamb. It has also been adopted as an anglicised form of an Irish name (by a mis-translation of Ó Luain) and as a romanised and anglicised spelling of a Chinese name. Even without the complications arising from language contact, forms of many English and Scots family names have also changed over the centuries, so that their modern appearance can be very misleading as to their origins and meaning.

The range of skills and knowledge required for interpreting the family names of Britain and Ireland is quite formidable, and even the most experienced scholars in the field would emphasise that they frequently need to consult colleagues, as well as constantly turning to reference works. A particular difficulty has always been that even if a medieval name can be explained quite easily, guessing whether a similar-looking modern family name is of the same origin can be a leap in the dark. On the brighter side, however, the digital era has made the research infinitely easier than in the past, with online resources such as editions of medieval records, major dictionaries, parish register and census data, and Steve Archer’s British 19^th Century Surnames, which maps the distribution of family names in the 1881 UK Census. It is partly for this reason, and partly because it’s the product of experts working within a large, well-funded collaborative project, that the recent Oxford dictionaries of family names have been able to make major advances on anything that has gone before. These are The Oxford Dictionary of Family Names in Britain and Ireland, The Concise Oxford Dictionary of Family Names in Britain and The Oxford Dictionary of Family Names of Ireland.

Spelling evidence

Interpreting family names has a good deal in common with interpreting place-names (see Interpreting place-names). Both are essentially a specialised form of linguistic research, depending above all on the evidence of early spellings. Using this evidence entails collecting possible instances from documents, deciphering the letters (for example, telling u and n apart and working out whether u is a vowel or a consonant in a particular spelling) and forming theories. The best explanation of a family name will account not only for the origins of the name, but also how it has developed. This is all worked out in the light of the known history of the languages concerned — which of course entails correctly identifying what languages those are — and awareness of how family names in particular vary and change through time.

One very useful type of pointer is that early forms of surnames in documents are often accompanied by small grammatical words that provide clues about the kind of surname they are, and hence about their origin and meaning. William le Lamb (recorded in Cambridgeshire in 1279) was evidently known as ‘the lamb’, for whatever reason, and it is le ‘the’ that tells us that this is a nickname, whereas the Londoner William atte Lamme was ‘at the Lamb’ in 1320, presumably a tavern with a sign of a lamb.

Contextual and comparative evidence

Importantly too, both family names and place-names need to be looked at in other kinds of context. For family names this means especially the known geographical and social distribution of the name across the centuries, and ideally this is coupled with tracing the history of individual families. Researching a family’s history is usually a very big project in itself, and so far it has been undertaken for only a minority of families. It involves not just identifying who is descended from whom, but also seeing how families have moved from their original neighbourhoods, and often how their social and economic fortunes have fluctuated (our Surname stories on Attenborough and Hamilton illustrate very different kinds of family history). In the last half-century the genetic evidence of DNA, especially of Y chromosomes, has become a very useful addition to the toolbox. Finally, for both family names and place-names, comparative evidence such as the history of similar individual names or larger patterns in naming behaviour can be invaluable.

Challenges and questions

The fact that family names are attached to people rather than places gives rise to two particular challenges in tracing their history: the possibility that the families bearing the names may have moved around a great deal over time, and the high degree of variation in spelling and pronunciation. Put quite dramatically:

Family names are much less stable than place-names. Over time, a family name may be borne by hundreds of thousands or even millions of individuals. These individuals move around, have offspring, change their names, and die.

Oxford Dictionary of Family Names in Britain and Ireland (2016), Introduction 2.3

As a result even the most dedicated family historians can often only get as far back as the 16^th or 17^th century and many theories about the etymology of family names have to remain tentative.

Bearing in mind this potential for movement and change in names and their bearers, a surname researcher needs to bear in mind several questions, and one of the most fundamental is:

Single or multiple origins – monogenetic or polygenetic?

Does the name in question have a single origin, stemming from the name of a single ancestor, or does it have multiple origins? In more technical words, is it monogenetic or polygenetic? (These terms come with the caveat that not all members of families are necessarily related genetically – individuals could have adopted a particular name for various reasons, or there may have been ‘non-standard paternity events’.) In general, surnames that are extremely common today are often polygenetic. A famous example is the occupational name Smith, with about 400,000 bearers in the 1881 Census. It is found widely across Britain, reflecting the fact that every village or hamlet needed a smith in medieval times. Monogenetic names on the other hand arose in a single locality and may often have relatively few bearers. These simple oppositions, between ‘polygenetic, widely distributed, numerous’ and ‘monogenetic, highly local, rare’ are far from universal, however. For instance the surnames Clegg and Kershaw have around 10,000 bearers each in 1881, residing mostly in Lancashire and the West Riding of Yorkshire, yet they are suspected to be monogenetic.

Most family names are uncommon

A surprising and at first sight paradoxical finding of research on the 1881 data is that common names are uncommon and uncommon names common, even when pronunciation and spelling variants are counted as the same name. Only a dozen family names, like Smith, run into hundreds of thousands of individual bearers, while even relatively common names (with around 10,000+ bearers, such as Clegg and Kershaw) number little more than 500, well under 1% of the total of different surnames (excluding spelling errors) in the 1881 census. By contrast, over 85% of the surnames had fewer than 500 bearers, many with fewer than 20, usually representing just a handful of households. This has a profound impact on the ability of current dictionaries to give an authoritative and unambiguous explanation for every well-attested surname. Establishing a clear chain of documentary evidence back to the medieval period for thousands of uncommon or rare names can be challenging when records are lacking or incomplete and families move location. Researchers are sometimes looking for needles in haystacks, not knowing whether any existing haystack contains the right needles. All surname dictionaries in Britain and Ireland are therefore effectively works in progress, awaiting future research by surnames specialists and family historians.

Puzzling distributions and homonyms

Where the ‘same’ family name is not very common but has more than one epicentre according to the mapping, what is the explanation? Possibilities include:

The same name arose independently in more than one place and more than one family.
The name arose in a single place and family but some of the family moved.
Names that were originally different have become identical (‘homonyms’) because one or both have changed through time.

Problems like this require painstaking work on local documents, perhaps aided by DNA evidence, but many remain unsolved and both the etymology of the name(s) and the course of the families’ histories may be uncertain.

Despite all the pitfalls, a great deal is known about the histories of families and their surnames, and the pursuit of that knowledge leads us along fascinating trails. As our Surname Stories show, these histories interweave language history, aspects of daily life in the past, major historical events and movements, and the careers of outstanding individuals.