About sumerian names
Drawn from the cuneiform record, results range from compact forms like "Antum" and "Zambiya" to hyphenated scribal constructions like "En-me-nuna" that join several word elements. They suit Bronze Age historical fiction, ancient temple campaign arcs, or naming a riverine empire in a fantasy atlas. Reserve compound forms like Shatu-murrim for rulers and keep the two-syllable results for farmers and laborers.
How sumerian names are built
Markov synthesis builds each name letter-by-letter from two attested banks: 520 male forms and 152 female forms. The chain reproduces the reduplicated, theophoric shapes of the source corpus, so doubled syllables and hyphenated god-compounds survive. Bank outputs include Nanna, Silili, and Warad-Musa.
Nanna
Markov-synthesised from the female bank, showing syllable reduplication
Silili
Markov-synthesised from the female bank, showing syllable reduplication
Warad-Musa
Markov-synthesised from the male bank, a theophoric hyphen-compound
Naming notes
The phonetic markers are syllable reduplication (Nanna, Silili) and theophoric hyphen-compounds (Warad-Musa) inherited from cuneiform-recorded Sumerian names, which is why both doubled and dash-joined forms appear. The male bank is much larger than the female (520 vs 152), so feminine variety is limited. Treat results as plausible Sumerian-flavoured coinages for ancient or mythic fiction, not attested kings or priestesses.
Questions about sumerian names
- How are Sumerian names generated?
- A Markov chain walks 520 male and 152 female attested forms, producing each next letter, so each name is one synthesised word.
- Why do some names repeat syllables or carry hyphens?
- Sumerian names recorded in cuneiform often reduplicate syllables or compound a deity element, so the chain reproduces forms like Silili and Warad-Musa.
- Are these real Sumerian names?
- No. They match the recorded sound of Sumerian onomastics but are generated coinages for fiction, not attested historical individuals.