I commenced this (still very incomplete) section as a putative taxonomy to help me understand more about the subject, through its possible structure.

But from the start I encountered difficulties.

Any taxonomy imposes one particular perspective: a decision has to be made as to which aspects to group together, and which to separate.

These aspects might be termed:

  • Source Language
  • Ethnic Group
  • Type - (Patronyms/Occupational/Toponyms etc)
  • Morphology (form of the word e.g. prefixes/suffixes, plurals)

For example, Patrick Hanks’s taxonomy for family names of the United States rightly regards the type of name as paramount; thus bringing together occupational names of all cultural groups.

For Great Britain, however, the ethnic mix is less, whilst the potential time-scales are longer. For these reasons, the first cut has been arbitrarily made as:

source language
source language

This first division is ambiguous and open to criticism (e.g. are medieval Jewish names to be treated as assimilated, and 20th century arrivals as non-assimilated?).

Nonetheless, most surnames in Britain have origins in widely-differing early source languages, Old English/Continental Germanic/Scandinavian, that a non-specialist is unaware of, whilst discerning the origins of recent arrivals.

In essence, this assimilated/non-assimilated category is merely just a pragmatic device to separate post-Victorian name arrivals, as most name studies will involve pre-Edwardian source documents, and as you might gather, it is one that I am not too comfortable with.

Surnames, by their often unknown origins, can be impossible to uniquely classify. They are slippery eels.

“A surname like Stevens may mean:
    ‘son of Stephen’,
    ‘servant of Stephen’,or
    ‘servant at Stephen's house’, or it may be
    a metronymic (i.e., a name after one's mother) derived from a form Stevenes ‘Stephen's wife’...”

Source: Percy H. Reaney A Dictionary of British Surnames London, Routledge and Paul, 1958.

In this particular case, you need to decide whether to assign it to a primary category, or to all possible ones.

Another area of ambiguity is status nicknames. A name like Pope is not likely to be derived from an actual position, but more likely to have been a role in a pageant.

I mentioned different perspectives earlier. You might wish to look at all types of names that derive from

  • the same source language,
  • names by gender e.g. female names,
  • the morphology of a name.

I have included an alternative classification that groups these.


I have not included any notation, because this taxonomy is experimental and embryonic. Also, the idea is not for each aspect of a name to be recorded by a relevant notation, and perhaps stored in a database But here is an example



  • Occupational name by craft
  • Female name by Gender
  • Morphology - er ending
  • Source Language - English

The source language, gender (though perhaps not the morphology) could be notated as standard sub-divisions (in the same way that in the Dewey Decimal system .942 is the standard subdivision for England, thus 320-942 = Political History of England, 330.942 =Economic History of England)

Hence the code could be for the language .19107, gender .1922


Then there will also be temporal and spatial sub-divisions (for the hearth tax, census, electoral roll), and the notation becomes totally unwieldy. Rather, I would envisage a taxonomy with radio buttons that is linked to an Access database

Actually, I have decided no longer to pursue a taxonomy per se. Rather I want to convey an idea of the complexity of the subject, and an indication of which areas are being studied, and a flavour of the writing. There is no name for this - so I am using the term 'aspect map' instead. Pretty naff, and I may change it.


  • Personal surnames
  • Surnames of Relationship
  • Occupational
  • Topographical
  • Toponymic
  • Nicknames
  • Variant perspective
  • Non-assimilated