How a name interpretation comes about at NamensArchiv, what it rests on, where its limits lie — and how you can report an error.
The basic principle of onomastics
German surnames are not an invention but a by-product of administration. As towns grew in the High Middle Ages, a first name was no longer enough to tell people apart: in a town with 300 men, 40 of them could easily be called Heinrich. So everyone got a byname — after whatever made him unmistakable. From the 12th century on, these bynames became hereditary surnames, earlier in the south, in the north and east sometimes not until the 16th to 18th centuries.
Name research sorts them into five basic types. Every interpretation at NamensArchiv begins with the question of which of these types applies:
Type
Derived from
Examples
Occupational name
trade, craft, office
Müller, Schmidt, Schneider, Wagner
Patronymic
the father's first name
Peters, Hartmann, Werner, Jansen
Name of origin
place or region the family came from
Bayer, Frank, Hesse, Böhm
Topographic name
location of the house in the settlement
Bach, Berger, Lindner, Brinkmann
Nickname
trait, appearance, character
Klein, Lang, Kraus, Schwarz
The classification is not always clear-cut — and that is precisely where serious name interpretation parts ways with the unserious kind. Schwarz can be a nickname for dark hair, but also a house name after a black house sign. Koch is usually an occupational name, but in some regions a Germanised form of a Slavic first name. Where several interpretations compete, we name the one predominantly held in the specialist literature and point out the alternatives.
What our information is based on
The foundation is the standard works of German name research and publicly accessible distribution data:
Rosa and Volker Kohlheim: Duden — Familiennamen. Herkunft und Bedeutung. The current reference work on the roughly 20,000 most common surnames in the German-speaking world.
Josef Karlmann Brechenmacher: Etymologisches Wörterbuch der Deutschen Familiennamen. Standard work with an extensive collection of evidence from historical documents.
Hans Bahlow: Deutsches Namenlexikon. Widely used, in parts academically disputed (Bahlow's "water word" thesis is considered outdated) — we use it only as a supplement and never for a sole interpretation.
Max Gottschald: Deutsche Namenkunde. Systematic account of name types and formation patterns.
Digitales Familiennamenwörterbuch Deutschlands (DFD), Academy of Sciences and Literature Mainz / TU Darmstadt — ongoing scholarly project on meaning and distribution.
Distribution maps based on public telephone-line and registration data (including Geogen by Christoph Stöpel) for regional clustering.
For non-German names, additionally the name research of the respective country and language — for instance the Słownik nazwisk for Polish names, the works of Hasan Eren and Turkish onomastics for Turkish names, and Arabic onomastics for Arabic names.
Beyond Germany
For names from other countries, we draw on official and academic distribution data wherever it exists: for France, the INSEE death register since 1970 with the municipality of birth; for Poland, the PESEL register with name bearers per voivodeship (2026); for the Netherlands, the Nederlandse Familienamenbank (CBG / Meertens Instituut) with figures per municipality for 2007 and 1947; for Belgium, Statbel with figures per municipality (2026). Everywhere else we work from etymological sources, place-name matching and regional history — honestly labelled as such.
How a page is made
The process is deliberately transparent because it includes automated steps:
Automated first analysis. An AI system evaluates the name along the standard literature named above: word root, name components, endings, language area, probable type. The system is expressly instructed not to invent dates, people or statistics and to label uncertainty.
Fact check. For the permanently published encyclopedia pages, the interpretation is checked against at least one standard work. Statements that can't be found there are cut rather than written in.
Editorial approval. Every encyclopedia page is read and approved by Jan Ozeir before publication. The spontaneous instant research on the home page does not go through this step — it is labelled as automatically generated information.
Correction in operation. Reported errors are checked; if the complaint is justified, the page is corrected or taken down.
Frankly: a name interpretation is a statement of probability, not proof of descent. It says how a name arose linguistically — not whom you descend from. As a rule, many families acquired the same name independently of one another. Anyone who sells you proof of a "family coat of arms" from the surname alone, for money, is selling you nothing.
Known limits
Rare names. The rarer a name, the thinner the evidence. We then provide the linguistic analysis (root, ending, language area, typical formation) and say openly that an individual record is missing.
Altered names. Migration, clerks and sound shifts have changed names massively — Kowalczyk became Kowaltschik or simply Schmidt. Such transformations often can't be reconstructed without family documents.
Regional double meanings. The same name may have arisen differently in Schleswig-Holstein than in Bavaria. Where this is known, we name both.
Distribution data. It reflects today's situation, not that of the 15th century. Expulsion, industrialisation and migration have shifted the map considerably.
Corrections and enquiries
If a statement on this site is wrong, we want to know. Report it via the contact page — most helpful with a source reference (work, local chronicle, church register, regional study). We usually get back to you within two working days.