Mysteries Department
Sign inRequest access
Proto-cuneiform: the archaic texts of UrukEvidence
Evidence · Proto-cuneiform: the archaic texts of Uruk

Name vocabulary turns over sharply between Uruk III and Archaic Ur, far more than between Archaic Ur and Fara

In a shared sign-name convention, the sign-frequency profile of Uruk III names has a cosine similarity of 0.47 with Archaic Ur names, against 0.73 between Archaic Ur and ED IIIa names; 8 of the 20 commonest Uruk III name signs are among the 20 commonest Archaic Ur name signs, all of them common logograms (A, AN, E2, EN, GI, HI, SAG, SI).

Suggestive0

What it means

Openunchanged
Answer this evidence bears onThe language is unknown (Englund's agnosticism)
Supports4 for, 1 against
Still possibleunchanged
What it implies
  1. The break in naming falls between Uruk III and Archaic Ur, the gap the site's open questions should now target.For: When and how did phonetic spelling begin in Mesopotamian writing?

Next steps 14

1
Labels or names: element reuse against Early Dynastic
Could shiftGood odds~4 h
On: What language, if any, do the proto-cuneiform texts record?
40score
Details

Compare how the archaic names reuse their elements (type-token curves, share of elements in two or more distinct names, compositional pairs) with Early Dynastic names at matched sample sizes; linguistic names, of any language, reuse a small vocabulary productively, while ad hoc labels should not.

2
Test the numeral-order argument on Sumerian accounts
Could shiftGood odds~6 h
On: What language, if any, do the proto-cuneiform texts record?
34score
Details

Englund argues that putting the numeral before the noun is un-Sumerian; Dahl notes that account formats may have their own order. In Early Dynastic administrative texts written in Sumerian (Fara and later, from CDLI or Oracc), measure how often the numeral precedes the counted noun. If Sumerian accounts also put it first, the syntax argument drops out and the Sumerian answer loses one objection; if they put it after, the archaic order stays as evidence for the unknown answer.

3
Officials' names in Uruk III
Could decideFair odds~10 h
On: What language, if any, do the proto-cuneiform texts record?
28score
Details

Identify name-like entries attached to officials and institutions in Uruk III (heads of households, seal owners, officials in the Professions-list vocabulary) and score them against the Early Dynastic benchmark. Sumerian-pattern officials with non-Sumerian dependents point to a foreign workforce; non-Sumerian officials point to non-Sumerian scribes or a labelling notation.

4
Officials' names
Could decideFair odds~10 h
On: What language, if any, do the proto-cuneiform texts record?
28score
Details

Test name-like entries attached to officials and institutions in Uruk III, to separate the scribes' own naming culture from the listed workers'.

Ranked by how likely each is to move the key question, for the time it takesAll 14 on the question

How it was found 1 run

Result. Semitic: the archaic names match the Akkadian signature no better than the Sumerian one, and far below either language's own names. That counts against H-OTHER's prediction, with low power. Archaic Ur: the Uruk III / Archaic Ur contrast holds in one convention under both conversions. Carry-over: there is a sharp turnover in name vocabulary between Uruk III and Archaic Ur..
Odds ratio of data-driven Akkadian and Sumerian name signatures in the Uruk III names; odds ratio of Englund's elements in ED I–II names converted to the proto-cuneiform convention; cosine similarity of name sign profiles across periods{'uruk3': {'95': [1.09, 1.6], 'OR': 1.35}, 'cosine': {'ed12_names_vs_ed3a_names': 0.73, 'uruk3_names_vs_ed12_names': 0.47, 'uruk3_names_vs_ed12_words': 0.28}, 'AKK_signature': {'uruk3_95': [1.48, 2.06], 'uruk3_OR': 1.76, 'train_names': 852, 'heldout_OR_akk': 12.31, 'heldout_OR_sux': 1.29}, 'SUX_signature': {'uruk3_95': [1.49, 2.67], 'uruk3_OR': 2.1, 'heldout_OR_akk': 5.4, 'heldout_OR_sux': 70.17}, 'ED12_pc_crosswalk': {'95': [4.68, 7.06], 'OR': 5.84}, 'ED12_pc_conservative': {'95': [2.75, 4.89], 'OR': 3.82}}Semitic: the archaic names match the Akkadian signature no better than the Sumerian one, and far below either language's own names. That counts against H-OTHER's prediction, with low power. Archaic Ur: the Uruk III / Archaic Ur contrast holds in one convention under both conversions. Carry-over: there is a sharp turnover in name vocabulary between Uruk III and Archaic Ur.
Figure 1

Neither a Sumerian nor an Akkadian name signature is present in the Uruk III names at anything like native strength

Odds ratio of each data-driven signature (15 signs) in names against non-name words; held-out Old Akkadian-period names and Uruk III names, 95% intervals for Uruk III

051015odds ratio (names vs other words)Akkadian signature in Akkadian names (held out)Akkadian signature in Uruk III namesSumerian signature in Uruk III namesEnglund's list, Archaic Ur names, archaic conventionEnglund's list, Uruk III names
051015odds ratio (names vs other words)Akkadian si…Akkadian si…Sumerian si…Englund's l…Englund's l…
odds ratio95% low95% high
The Sumerian signature reaches 70 in held-out Sumerian names (off the scale). The Akkadian signature is mostly syllabic signs, so its absence from pre-phonetic names is weak evidence.
Drawn by the site with Semitic signature test and Archaic Ur in the archaic convention
Archaic personal names in proto-cuneiform (editor-marked, CDLI)Early Dynastic Sumerian personal names, lemmatized (Oracc ePSD2 admin corpora ED I–II, IIIa, IIIb)semitic_and_ur.pynames_test.py2 outputs

Sources 2

SRC
SRC

Cite it

Mysteries Department. (2026). Name vocabulary turns over sharply between Uruk III and Archaic Ur, far more than between Archaic Ur and Fara [Evidence EVD0DG2PSG85PGC]. https://mysteriesdept.com/r/EVD0DG2PSG85PGC
Private beta Downloads are for beta members

We're in a private beta while we get things right. Everything here is free to read; downloading datasets, sources and code is open to members for now.

Apply to the betaSign in
It takes a couple of minutes. We'll reply by email.