If you set your fantasy story in 1100 AD and a character introduces themselves as "Robert Smith," you've made an anachronism. Hereditary family surnames — the kind where the entire Smith family shares the name Smith across generations — didn't exist in most of Europe until well into the 13th century. Before then, almost everyone was identified by some combination of personal name plus one of four alternative descriptors.
This guide walks through the four pre-surname identification systems that medieval Europeans actually used, when hereditary surnames emerged, and how to use the right system for the period you're writing.
The Problem of Identification Before Surnames
Imagine a village with 200 inhabitants in 1050 AD. How does the local priest, the tax collector, or the manor lord distinguish between three people named John? They don't have surnames. They don't have ID cards. They have one name each, often a fairly common name (variants of John, William, Henry, and Robert covered something like 40% of the male population in 11th-century England).
The solution was modifiers — extra information added to the personal name to disambiguate. There were four common modifier types, each of which eventually crystallized into the hereditary surnames we know today.
System 1: Patronymics
The most ancient system: identify a person by their father's name.
- John, son of William → John Williamson (in eventual fixed form: Williamson, Wilson, Williams)
- Edith, daughter of Henry → Edith Henrysdaughter (rarely fixed in English; survives in patronymic forms)
This is the system the Vikings used (see the post on Viking patronymics) and that survives intact in Iceland today.
In England, the -son patronymic was widespread among Anglo-Saxons and Normans alike. By 1200, names like Williamson, Robertson, Jameson, Johnson, Thompson were beginning to fix as hereditary — meaning the son part stopped reflecting actual sons of currently-living fathers and became a fixed family identifier.
In France, the patronymic prefix was Fitz- (from Norman French fils, "son of") for the aristocracy: Fitzwilliam, Fitzgerald, Fitzroy (the last meaning "son of the king," reserved for royal bastards).
System 2: Occupation / Trade Names
The second most common: identify a person by what they do.
- John the smith → John Smith
- Margaret the baker → Margaret Baker
- Robert the miller → Robert Miller
This explains why English surnames cluster so heavily around medieval trades: Smith, Baker, Miller, Carpenter, Tailor, Mason, Cooper, Wright, Fletcher (arrow-maker), Fowler (bird-hunter), Shepherd, Cook, Carter, Brewer, Sawyer, Thatcher, Dyer, Walker (wool-fuller), Weaver, Glover, Hooper, Sadler, Tanner.
The pattern repeats across European languages:
- German: Schmidt (smith), Müller (miller), Schneider (tailor), Schreiner (cabinetmaker)
- French: Lefèvre (smith), Boulanger (baker), Charpentier (carpenter)
- Italian: Ferraro (smith), Sartori (tailor)
- Polish: Kowalski (smith), Młynarz (miller)
When you're naming medieval characters, the trade-name approach is historically accurate even when you don't want a hereditary surname. John the smith describes a real role and a real person in 1100.
System 3: Location-Based Names
The third common modifier: identify a person by where they're from or where they live.
- John of York → John York or John Yorke
- William from the hill → William Hill
- Anne by the brook → Anne Brooks
This created a huge category of English place-based surnames: Hill, Brook, Wood, Field, Ford, Bridge, Town, Cottage, Hall, Wells, Pond, Greenwood, Ashfield, Oakridge.
For aristocracy, the location-based name was usually their fief or seat: Hugh of Warwick (a knight whose fief was Warwick), Eleanor of Aquitaine (a queen whose ancestral domain was Aquitaine).
The pattern in other European languages:
- French/Norman: de [Place] — Hugues de Montfort (Hugh of Montfort)
- German: von [Place] — Friedrich von Hohenzollern
- Italian: da/di [Place] — Leonardo da Vinci (Leonardo from Vinci)
- Iberian: de [Place] — Rodrigo de Castro
For nobility, the de/von/da prefix often stuck even after the family lost the original land — a vestige of medieval landholding identity.
System 4: Descriptive / Nickname Modifiers
The fourth modifier: identify a person by physical or personality trait.
- William the tall → William Tall (later: Long, Lang, Longshanks)
- John the bald → John Bald (later: Ball)
- Robert the red → Robert Red (later: Russell from Old French rousel, "red-headed")
- Margaret the gentle → Margaret Gentle (later: Gentle, Gentry)
These nicknames cover a huge range of attributes:
- Physical: Tall, Short, Strong, Brown (hair), Black (hair), Red (hair), Fair, Stout, Little (small), Long (tall), Lyttle (small)
- Personality: Wise, Bold, Quick, Gentle, Hardy, Goode, Trew (true), Wylde
- Status: Knight, Squire, Page, Yeoman, Bond (peasant)
For nobility, these descriptive epithets became attached to kings and lords: Edward the Confessor, William the Conqueror, Richard the Lionheart, Charles the Bald, Pepin the Short, Charles the Fat, Æthelred the Unready. See how knights got their epithets for the chivalric version of this convention.
When Did Hereditary Surnames Emerge?
The shift from "identifying modifier" to "fixed family surname" happened gradually:
- Norman aristocracy (1066–1200): Some Norman noble families started using their fief name as a fixed family identifier almost immediately after the Conquest. de Beaumont, de Clare, de Lacy — these became inherited family names within one or two generations of the Conquest.
- English commoners (1200–1400): Hereditary surnames became standard among commoners more slowly, driven by tax records, land deeds, and royal census requirements that needed consistent identification across years. By 1400, most English families had stable surnames.
- Scotland (1200–1500): Highland Scotland preserved patronymics longer than England did. Mac- prefix (son of) and Nic- prefix (daughter of) remained productive through the medieval period.
- Iceland: Never adopted hereditary surnames. Patronymic system continues to the modern day.
- Germany: Hereditary surnames spread later than in France or England, varying by region. Aristocratic von names were earliest fixed; commoner surnames stabilized 1400–1600.
- Eastern Europe: Hereditary surnames spread even later, sometimes not until the 18th–19th centuries when imperial governments imposed fixed names for tax purposes.
Implications for Your Fiction
If your story is set before 1200:
- Your characters should rarely have hereditary surnames
- Identify them using one of the four modifier systems: patronym, trade, location, or descriptor
- John the smith, William the bold, Robert son of Henry, Eleanor of Aquitaine — these are period-accurate
If your story is set 1200–1400:
- Mixed period: aristocracy has surnames, commoners are transitioning
- Aristocratic characters can use de [Place] or fixed Norman names (Beaumont, Lacy, Mortimer)
- Commoner characters often still use trade or descriptor modifiers
If your story is set after 1400:
- Hereditary surnames are standard for most of European society
- John Smith is now a real fixed family name; trades + locations + descriptors are all by now inheritable
If your story is set in a fantasy world inspired by medieval Europe:
- Pick a convention and apply it consistently
- Tolkien used patronymics for Vikings (Thorin son of Thrain), location markers for aristocrats (Aragorn of the Dúnedain), and almost no hereditary surnames for commoners
- Game of Thrones uses bastard surnames (Snow, Sand, Stone) and House names (Stark of Winterfell, Lannister of Casterly Rock) — closer to Norman aristocratic practice than to fully modern surnames
Examples to Borrow
For your medieval characters, here are period-accurate naming combinations:
England, c. 1100:
- Henry son of Robert (commoner)
- Hugh of Warwick (landed knight)
- Adam the carter (commoner trade)
- Edith the gentle (commoner descriptor)
- William FitzRobert (Norman aristocrat, illegitimate)
France, c. 1150:
- Pierre de Montfort (knight with fief)
- Geoffroi le Bon (commoner, "the Good")
- Jeanne de l'Hôpital (location-based)
Germany, c. 1200:
- Heinrich von Hohenstein (aristocrat)
- Konrad Schmidt (commoner, smith)
- Gertrud im Felde (location, "in the field")
The medieval name generator lets you pull period-appropriate first names; combine them with one of the four modifier systems for full period-authentic identification. For a chivalric formal name with epithet, the knight name generator handles the Sir + Name + (of Place / Epithet) pattern with the right Norman / Anglo-Norman flavor.
A medieval character without a hereditary surname doesn't feel anonymous — they feel anchored to their world in the way real medieval people were. By their trade, their place, their father, or their reputation. The system is older than surnames, and for pre-1300 fiction, it's the one to use.