Guide
Borrowing Real-World Names Without Getting It Wrong
From the cuename guides
Real names arrive freighted. Lift Aoife, Okonkwo, or Bjarni into a fantasy setting and every reader who knows those names brings their associations along. This guide covers how naming systems differ across cultures, where fiction usually fumbles them, how to research without flattening anyone, and when the honest move is to invent.
A Name Is a System Before It Is a Sound
The Western default of one given name plus one inherited surname is a local convention, not a law of nature. Iceland still runs on patronymics. Björn's son Leifur is Leifur Björnsson, his sister is Helga Björnsdóttir, and neither passes that second element down; their children build new names from Leifur and Helga. Icelanders address one another by given name, and the national phone directory alphabetizes by first name. A novel that hands an Icelandic family one shared surname across three generations has copied the sound and missed the machine.
Rome ran a different machine. A male citizen of the classical era typically carried the tria nomina: a praenomen drawn from a small recycled pool, a nomen marking his gens or clan, and a cognomen marking a branch or a nickname. Cue a batch from the Roman namer and three-part shapes like Publius Balventius Tetricus surface beside two-part women's names like Tullia Herena, because Roman women were commonly known by the feminine form of the family nomen, with sisters distinguished by epithets such as Major and Minor.
So the first research questions are structural. Ask how many parts a name has, in what order they come, which parts shift with gender or marriage, which are inherited, and who may speak which part aloud. Get those answers before you collect a single pretty syllable. The Viking namer bakes the Old Norse pattern in, which is why a cued list reads Valgard Vestarsson beside Hrafnhild Bjarkidottir, son endings for men and dottir endings for women, never swapped.
Name Order, Gendered Particles, and Chains of Descent
Family name first is the normal order across a wide band of the world. Japanese, Chinese, Korean, Vietnamese, and Hungarian names all lead with the family name in their home languages. A list cued from the Japanese namer reads Kawashima Mako and Mukai Noriko, family name first, and an acquaintance would be addressed as Kawashima-san rather than by the given name Mako, which is reserved for intimates. English media flips the order so routinely that writers forget the flip is an act of translation, not a fact about the names.
Irish surnames carry grammar inside them. Ó means descendant of and Mac means son of, but those are men's forms. A daughter takes Ní or Nic, a married woman traditionally takes Uí or Mhic, and the name that follows shifts form under Irish grammar, so a daughter of Ó Briain is Ní Bhriain. The Irish namer preserves all of this, returning Caitlin Nic Niocaill and Samthann Uí Bradáin instead of stamping one frozen surname onto every member of a family.
Arabic names historically chain rather than pair. An ism, the personal name, can be followed by a nasab tracing fathers through ibn or bint, a kunya honoring a child through Abu or Umm, and a nisba tying the bearer to a place, trade, or tribe. The scholar Western readers call Avicenna was Ibn Sina. Russian sits between systems with a given name, a patronymic built on the father's name, and a family name, and the respectful register of address pairs the first two, Ivan Petrovich, not Mr. Petrov.
The Mistakes That Keep Showing Up in Fiction
The most common error is treating a patronymic as an inherited surname. Fiction keeps handing dóttir endings to men, or passing Björnsdóttir from mother to daughter like an heirloom, when the ending means daughter and the stem must be the parent's own name. The mirror error treats a family name as a given one, so a heroine ends up with Sakamoto in the given name slot, a choice that reads to Japanese speakers roughly the way Postlethwaite would read as an English first name.
Era mixing is subtler. Tarozaemon is a name shape from the Edo period, and seating him at the same table as a sharply contemporary given name signals time travel to anyone who knows the culture, the way a medieval English village housing both Aethelstan and Kayden would to an English reader. Baby-name listicles invite exactly this failure, because they strip away period, region, and class, the three constraints that make a name sit naturally in its world.
Then there is weight. Some names are not neutrally available. Many Aboriginal Australian communities observe restrictions on using the names of people who have recently died, which is why Australian broadcasters air warnings before older footage. Names of deities, prophets, and revered ancestors land differently on a tavern keeper than names drawn from the same culture's everyday stock. Borrowing the sound without learning the status is how comic relief ends up wearing somebody's sacred name.
Research Habits That Hold Up
Go to records, not roundups. The Domesday survey of 1086 preserves thousands of personal names from Norman England in context. Icelandic sagas, Roman inscriptions, parish registers, and census rolls show names attached to actual people of known period, place, and station. A name met in a record arrives with its register attached. A name met on a listicle arrives with nothing.
Learn the romanization before trusting a spelling. The same man appears as Mao Tse-tung under Wade-Giles and Mao Zedong under Pinyin, and Japanese names drift across sources depending on how a transcriber handled long vowels. Mixing systems inside one cast quietly assembles characters from two different centuries of transcription. For living cultures the strongest single step is asking someone from the culture, whether a sensitivity reader, a cultural organization, or a speaker who can flag that a name is technically attested but reads the way Moonbeam reads in English.
Calibrate care to consequence. Nobody from classical Rome will read your book, but people who carry Hindi, Maori, or Yoruba names will, and they can tell a writer who learned how their names work from one who skimmed for texture. The research bar rises as the culture gets closer, more living, and more accustomed to being used as set dressing.
When Inventing Beats Borrowing
If your fantasy culture keeps a real culture's names while discarding its language, food, faith, and history, the names become costume. Readers who recognize them will expect the rest to follow, and when it does not, the borrowing reads as decoration at best. The durable alternative is borrowing structure instead of strings. Keep a patronymic system but build it on invented stems. Keep family-name-first order while inventing a phonology of your own.
This is the Tolkien move. He acknowledged modeling Quenya partly on Finnish and Sindarin on Welsh, taking sound patterns and grammatical flavor while inventing the vocabulary, which is why Sindarin names evoke Welsh without being Welsh. Sound symbolism then does real work. The bouba-kiki effect, the well-documented tendency for people across languages to match rounded sounds with round shapes and sharp sounds with spiky ones, means a culture of soft open syllables will feel different from a culture of clipped stops before a reader learns one fact about either.
A workable test: show only your names to a knowledgeable reader and ask what culture they belong to. If the answer is a real one, you owe that culture the full research treatment or a different set of names. Structure borrowed and strings invented clears the bar. Strings borrowed and structure ignored fails it twice.