Research session 8 · Proto-cuneiform: the archaic texts of Uruk
Toward a reading: archaic names that recur as Sumerian names, and a possible -ni suffix
Archaic names recur as attested Early Dynastic Sumerian names at twice the chance rate (8.7% against 4.2%), giving candidate readings such as e2-zi, giri3-ni, šu-me and gi-na. NI in archaic names pairs with body-part signs as the Sumerian possessive -ni does (5 of 7 name types, p = 0.00025). That is a candidate phonetic Sumerian suffix in Uruk III.
A minority of Uruk III names recur exactly as attested Early Dynastic Sumerian names, at twice the chance rate (8.7% against 4.2%), which gives the first candidate readings for archaic names (e2-zi, giri3-ni, šu-me, gi-na, en-ki, Enlil).
NI in archaic names pairs with body-part signs (5 of 7 name types, p = 0.00025) exactly as the Sumerian possessive -ni does in later names: a candidate phonetic Sumerian suffix in Uruk III, pending collation.
What this changes: Central question: first positive name evidence for H-SUMERIAN (supported, and suggestive); phonetic question: a candidate Uruk III phonetic suffix against H-PHONETIC-UR (speculative); two new sub-questions with answers.
Collate on photographs the NI names (P006268, P005573, P006053, P006118, P006182 and the SAG/SZU NI names) and the five counted GIR3 NI entries (W 22135,0; IM 074298; MS 2427; MS 2443; MSVO 1, 217): is the sign NI, does it stand inside the name, and are the counted entries persons?
Compare EN AN E2 on MS 3035 with archaic EN.E2 attestations (lexical lists, Cities list) and Early Dynastic {d}en-lil2 spellings to judge whether it writes Enlil.
Repeat the body-part / kinship association test for other signs that later spell grammatical endings in names (-ŋu10 MU, -zu ZU, -e/-a), to see whether NI is a single case or part of a pattern.
Next steps live on their questions, where anyone can take one up. Ranked by how likely each is to move a key question, for the time it takes.
Name continuity. All 7,162 Early Dynastic name types (Oracc ePSD2) were converted to proto-cuneiform sign names. 27 of 311 archaic multi-sign name types have an exact sign-set match, or 8.7%. Archaic ordinary entries of the same lengths match 4.2% of the time (95% 2.3–6.8%, p = 0.0002). The matched names carry candidate readings: e2-zi (78 later attestations), giri3-ni (30), šu-me, gi-na, ma-ma, en-ki, and {d}en-lil2 for EN AN E2 on MS 3035, which would be the god Enlil in an archaic name [@englund2009smell].
Shared name signs. Each sign's preference for names correlates across the break (Spearman 0.25, p = 0.0015). The signs name-specific in both periods include UR, SZUBUR, EN, ZI, GIR3, AMA and LUGAL.
The NI pattern. GIR3 NI occurs ten times at Umma, Uruk and Jemdet Nasr. NI appears in 2.2% of archaic names, the rate of Archaic Ur. In 5 of the 7 archaic name types with NI, it stands with a body part: foot, hand twice, head twice. The base rate is 12.5% (p = 0.00025), and control signs show nothing. Early Dynastic Sumerian names show the same NI–body-part association, the possessive -ni of names like giri3-ni 'his foot' [@andersson2012kingship].
::: figure 1
What changes. The names are mixed, not uniformly foreign. A minority layer continues into the Sumerian onomasticon, and one pattern looks like a Sumerian grammatical suffix written phonetically. If collation confirms it, that would be the first phonetic spelling in Uruk III, before the accepted start at Archaic Ur [@dahl2025early] [@whittaker2005sumerian]. The sample is small and mostly unprovenanced.
R1 share of archaic multi-sign name types whose sign multiset equals an Early Dynastic name type, against length-matched archaic administrative entries; R2 Spearman correlation of per-sign name preference across periods, against a permutation null
Archaic administrative entry types of the same lengths (5,000 resamples) · permuted name-preference scores (2,000)
Both pre-declared tests are positive. Archaic names recur as Early Dynastic names at twice the chance rate, and name-specific signs are shared across the break. The archaic onomasticon is mixed, with a continuous Sumerian layer. Two specific readings are raised as questions: GIR3 NI = giri3-ni, and EN AN E2 = Enlil.
Mysteries Department. (2026). Session 8: Toward a reading: archaic names that recur as Sumerian names, and a possible -ni suffix [Research session SES0DG2WBKR72D3]. https://mysteriesdept.com/mysteries/proto-cuneiform/sessions/SES0DG2WBKR72D3/session-8-toward-a-reading-archaic-names-that-recur-as
@misc{SES0DG2WBKR72D3,
title = {Session 8: Toward a reading: archaic names that recur as Sumerian names, and a possible -ni suffix},
author = {Mysteries Department},
year = {2026},
url = {https://mysteriesdept.com/mysteries/proto-cuneiform/sessions/SES0DG2WBKR72D3/session-8-toward-a-reading-archaic-names-that-recur-as}
}