By Emilia Wellesley · Published May 6, 2026 · Updated May 6, 2026
What Is the Proto-Elamite Script?
Proto-Elamite is the world’s oldest still-undeciphered writing system with a substantial corpus, used in southwestern Iran roughly between 3100 and 2900 BCE on small clay tablets that record the daily accounting of an early urban economy. Its inventory of more than a thousand distinct signs has resisted full reading for more than a century. Scholars treat it as the earliest writing native to the Iranian plateau.
The script first surfaced in the late 1890s, when the French archaeologist Jacques de Morgan (1857-1924) recovered the first two tablets at Susa, the long-lived city on the Khuzestan plain that would later become an Elamite capital. The corpus has grown unevenly since: roughly 1,600 published texts, the vast majority still from Susa, with smaller groups from Tall-i Malyan (ancient Anshan), Tepe Yahya, Tepe Sofalin, Tepe Sialk, and a handful of other sites. The script remains one of the great open files in the broader history of historical and archaeological mysteries.
What follows traces the Proto-Elamite tablets across their physical supports and their numerical scaffolding, the long arc of decipherment from Walther Hinz and Piero Meriggi to Robert Englund, Peter Damerow, and Jacob Dahl, the methodological leap of Reflectance Transformation Imaging at Oxford, and the unresolved question of whether the language behind the script is an early form of Elamite or something we cannot yet name. Where the evidence runs thin, the gap is named openly rather than papered over.
The Corpus: Where the Script Survives
The closest thing the field has to a single canonical edition is the running digital corpus maintained by the Cuneiform Digital Library Initiative, a consortium founded by the American assyriologist Robert K. Englund (1952-2020) at UCLA in 1998 and continued in partnership with Oxford and Berlin. The CDLI hosts photographs, line drawings, transliterations, and metadata for the published Proto-Elamite tablets and serves as the working reference for nearly all current scholarship; its open files at the Cuneiform Digital Library Initiative remain the practical entry point for anyone wanting to read the original signs [1]. Without the CDLI, comparative work across find-spots would still depend on copy drawings made eighty to a hundred years ago, many of them now known to be unreliable.
The Tablets and Their Sites
The largest single concentration is at Susa, where French excavations from the 1890s onward recovered well over a thousand tablets and tablet fragments, most of them now in the Louvre. Outside Susa, the corpus is small but geographically wide. Anshan at Tall-i Malyan, in the Marv Dasht plain near modern Shiraz, has produced about thirty-three Proto-Elamite tablets; the long-running American excavation at Tepe Yahya in southeastern Iran yielded twenty-seven, published as a unit by Peter Damerow (1939-2011) and Robert Englund in 1989; Tepe Sofalin, near Tehran, has added another dozen or so; smaller groups come from Tepe Sialk, Ozbaki, Shahr-e Sukhteh, Tall-e Geser, and Tepe Hissar. The geographic range matters: the script crossed the Zagros mountains and reached the eastern margins of the plateau, suggesting a writing tradition that travelled with administration rather than with a single city.
What the Tablets Record
The Proto-Elamite tablets are administrative documents, not literary or religious ones. They list amounts of grain, animals, slaves, and finished goods; they identify officials and institutions through ownership signs; they track distributions and receipts. This functional uniformity is a methodological problem as much as a curiosity. Decipherment thrives on textual variety. A corpus of accounting tablets, no matter how large, gives the would-be decipherer a narrow band of vocabulary, almost no grammar, and few of the proper-name strings that allowed Linear B and Egyptian hieroglyphs to be cracked. Susa was a working city; its scribes were balancing books, not telling stories, and that constraint has shaped every attempt to read them since.
The Sign Repertoire and the Numerical Scaffold
Estimates of the sign inventory range from around a thousand to roughly twelve hundred distinct graphs, depending on how variant forms are grouped, although the secure stable repertoire is smaller. Numerical signs are a class apart and have been read with confidence; the rest of the inventory has not.
Four Counting Systems Inside One Script
Proto-Elamite scribes employed at least four parallel numerical systems, each tied to what was being counted. A sexagesimal system, structured around progressive units of one, ten, and sixty, was used for most discrete commodities (livestock, slaves, finished objects); a decimal system, peculiar to Proto-Elamite and absent from contemporary Mesopotamian proto-cuneiform, was used to count animals and human beings under specific administrative regimes; a bisexagesimal system handled barley products and certain grain rations; a capacity system, with its own series of units, measured volumes of liquids and dry goods. The reconstruction of these systems, carried forward over decades by Damerow, Englund, and the Tepe Yahya team, is the one part of Proto-Elamite that can be read with rigor; the numerical signs sit on a stable arithmetic that constrains their possible values [2].
Why the Non-Numerical Signs Resist
Outside the numerical scaffold, the picture is harder. A small set of signs behaves like logograms for commodities, identified by their consistent placement next to specific quantities. A larger group seems to identify persons, offices, or institutions through compound graphs that combine repeating components. The remainder, which is the bulk of the inventory, has resisted any reliable phonetic or semantic reading. Three obstacles compound the difficulty: the corpus, while substantial, contains very few duplicate or near-duplicate texts that would let a scholar exploit redundancy; the underlying language is unknown; and no bilingual inscription has yet been recovered. The same three obstacles, in different proportions, blocked Cretan Hieroglyphic and Linear A. Proto-Elamite shares all three.
From Hinz and Meriggi to Englund and Dahl
The serious historiography of Proto-Elamite begins in the middle decades of the twentieth century with two scholars who came at it from opposite sides of the Mediterranean. The German Iranist Walther Hinz (1906-1992), holder of the Göttingen oriental chair from 1937 and author of the two-volume Elamisches Wörterbuch (1987) co-edited with his pupil Heidemarie Koch, took on Proto-Elamite as part of a broader project to reconstruct the script-history of the Iranian plateau, and is best remembered for his 1961 partial decipherment of the later Linear Elamite script which he then tried, with limited success, to extend backward into Proto-Elamite [3]. The Italian linguist Piero Meriggi (1899-1982), an Indo-Europeanist whose career had moved through Lycian, Cretan Linear A and B, and the Indus script before settling on Elam, produced between 1971 and 1974 the foundational three-volume sign catalogue La scrittura proto-elamica, which assigned numbers to the recurring signs and established the working repertoire that subsequent scholars have refined.
The Tepe Yahya Volume and the Berlin Method
A second methodological wave arrived in 1989 with the publication of The Proto-Elamite Texts from Tepe Yahya by Peter Damerow and Robert Englund. The volume is short on dramatic claims and long on careful reconstruction. Damerow, working at the Max Planck Institute for the History of Science in Berlin, brought to the tablets the same patient approach he had developed for early Mesopotamian numeracy: read the numerical systems first, hold the non-numerical signs at arm’s length until the arithmetic constrains them, and never assume a phonetic value when an administrative one is sufficient. Englund, who would go on to found the CDLI in 1998 and to lead the Iran-writing programme at UCLA through the 1990s, extended the same discipline to the larger Susa corpus. His 2004 essay “The State of Decipherment of Proto-Elamite,” widely circulated through the CDLI’s open archive, is the standard sober summary of where the field stands and what it would take to advance further [4].
The Oxford Project and Reflectance Transformation Imaging
In 2003, Englund handed the Proto-Elamite work to his doctoral student Jacob L. Dahl, who carried it to Oxford’s Faculty of Asian and Middle Eastern Studies. Dahl’s central methodological innovation, developed with the Southampton computer scientist Kirk Martinez between 2012 and 2017, was to apply Reflectance Transformation Imaging to the Louvre’s Susa tablets at high resolution. The technique uses a hemispherical dome with seventy-six independently controlled lights and a fixed camera; seventy-six photographs are taken, one per light, and the resulting image stack lets the viewer rake virtual light across the tablet surface from any angle. Surface marks invisible to ordinary photography, including faint stylus traces, partially erased signs, and shallow palimpsest layers, become legible. Around 1,100 Susa tablets were imaged this way and posted to the CDLI for open re-examination, and the project recovered earlier scribal corrections and miscounts that had been smoothed over by twentieth-century hand copies.
What Dahl’s Imaging Has and Has Not Settled
In short: the imaging gave the field a much better baseline image of the corpus, allowed earlier transliterations to be corrected against the original surfaces, and confirmed that the Proto-Elamite scribal tradition was less standardized than its proto-cuneiform Mesopotamian cousin. Dahl has argued, on the basis of the cleaned-up corpus, that the script may use a partial syllabary in which roughly half of the relevant signs were invented for sound values that have no logographic ancestry. The hypothesis is suggestive and consistent with some of the recovered patterns; it has not produced a confirmed reading. As of 2025, Dahl himself estimates that eighty to ninety per cent of the non-numerical signs remain undeciphered.
The Mahboubian Question and the Linear Elamite Bridge
A connected question has shadowed Proto-Elamite for decades: was it ancestral to the later Linear Elamite script, used in southwestern Iran around 2300 to 1850 BCE, whose underlying language is securely Elamite? If so, Linear Elamite values could in principle be projected backward to constrain Proto-Elamite. The hypothesis is old. Hinz and Meriggi both pursued it; Walther Hinz proposed graphic correspondences as early as 1975, and Meriggi argued in his 1971-1974 catalogue that the two systems might be a single tradition at different stages.
The Mahboubian Vessels and Desset’s 2022 Decipherment
The decisive Linear Elamite breakthrough came from a different direction. The French archaeologist François Desset, working with material that included silver vessels held in the Mahboubian collection, a private holding of Iranian antiquities assembled by the London-based dealer Houshang Mahboubian, published in 2022 a near-complete decipherment of Linear Elamite alongside Kameyar Tabibzadeh, Matthieu Kervran, Gian Pietro Basello, and Gianni Marchesi [5]. The team identified the underlying language as Elamite and read royal-name and divine-name strings on the vessels with cross-checks against later Akkadian-Elamite bilingual fragments. The provenance of the Mahboubian material has been a separate and live debate (private antiquities collections, particularly those holding Iranian objects, raise unresolved acquisition-history questions); the linguistic argument, however, has gained substantial scholarly support.
Why Linear Elamite Has Not Unlocked Proto-Elamite
A reader might expect the 2022 decipherment to flow backward and crack Proto-Elamite. It has not, and the reason is informative. Jacob Dahl, in particular, has argued that the apparent graphic similarities between the two systems are better explained by deliberate cultural imitation than by direct descent: Linear Elamite scribes, who were familiar with Old Akkadian cuneiform from their Mesopotamian neighbours, may have borrowed the look of an older local script to mark cultural distinctness, without inheriting its phonetic values. On that reading, Linear Elamite is a fresh writing system that wears Proto-Elamite costume, and projecting Linear Elamite values backward will mislead more than it will illuminate. Other scholars, including the Desset team, treat the link as more substantive. Both positions are still actively defended.
Why the Script Resists, and What Would Settle It
Three structural obstacles lie between the surviving Proto-Elamite corpus and a working decipherment, and they are worth naming together. First, the corpus is administrative. Roughly 1,600 mostly short tablets, almost all of them book-keeping documents, do not provide the textual variety that decipherers historically need; the script may simply not have been used for the kinds of names, narratives, or formulaic prayers that yield phonetic anchors. Second, the underlying language is unknown. Despite the geography, there is no positive evidence that Proto-Elamite records the same Elamite language attested in later Linear Elamite and cuneiform sources; the assumption is plausible but unconfirmed. Third, no bilingual inscription has been recovered, and the lack of standardized scribal training (no preserved sign lists, no exercise tablets, no scribal schools of the kind that produced Mesopotamia’s lexical tradition) means even the internal variation of the script is harder to systematize.
Either obstacle could be partially relieved by a single class of evidence. A new excavation that recovered a non-administrative Proto-Elamite text (a votive inscription, a name list, a royal dedication) would broaden the textual base. A securely contextualized bilingual, however brief, would anchor phonetic values. A Linear Elamite document that carried recognizable Proto-Elamite signs in a clearly transitional context would settle the descent question. None of these has surfaced. The prudent position, the one Englund argued for in 2004 and Dahl has continued through the imaging programme, is that the corpus may simply be too small, too administratively narrow, and too monolingual to be fully cracked, and that any future advance will probably come from new finds rather than from a dramatic reading of the existing corpus.
Frequently Asked Questions
When was Proto-Elamite in use?
The script was in active use roughly between 3100 and 2900 BCE, a window of about two centuries during the late Uruk and early Early Dynastic periods of the broader Near East. By around 2900 BCE its use contracts sharply and the system effectively disappears from the archaeological record, leaving Mesopotamian-style cuneiform as the dominant administrative writing across the region.
How many tablets survive?
Roughly 1,600 published Proto-Elamite tablets and tablet fragments are currently catalogued, the vast majority from Susa, with smaller groups from Tall-i Malyan, Tepe Yahya, Tepe Sofalin, Tepe Sialk, and a handful of other Iranian sites. The Cuneiform Digital Library Initiative hosts the working digital corpus.
How many distinct signs does the script have?
Estimates range from around 1,000 to roughly 1,200 distinct signs depending on how variant forms are grouped. The numerical signs (decimal, sexagesimal, bisexagesimal, capacity) form a stable, partially deciphered subset; the much larger non-numerical inventory remains substantially unread.
Has Proto-Elamite been deciphered?
No. Numerical signs and a small set of commodity logograms can be read with confidence, but Jacob Dahl estimates that 80 to 90 per cent of the non-numerical signs remain undeciphered. Phonetic values for most signs are unknown, and no bilingual inscription has yet been recovered to anchor them.
What language does Proto-Elamite record?
The language affiliation is unresolved. The script is geographically Iranian and may record an early form of the Elamite language attested in later Linear Elamite and cuneiform sources, but there is no positive evidence for this. Elamite itself is widely treated as a language isolate, with a contested possible affiliation to Dravidian.
Who deciphered the numerical systems?
The reconstruction of the four Proto-Elamite counting systems (decimal, sexagesimal, bisexagesimal, and capacity) was carried forward over decades, with major contributions from Piero Meriggi in the 1970s and from Peter Damerow and Robert Englund through their work on the Tepe Yahya tablets and the broader Susa corpus from the late 1980s onward.
What is Reflectance Transformation Imaging?
Reflectance Transformation Imaging is a photographic technique developed for ancient inscriptions in which a hemispherical dome containing 76 individually controlled lights and a fixed camera produces a stack of 76 images of the same object. The stack lets a viewer rake virtual light across the surface from any angle, revealing details invisible to ordinary photography. Jacob Dahl and Kirk Martinez applied it to roughly 1,100 Louvre Proto-Elamite tablets between 2012 and 2017.
Is Linear Elamite the same script as Proto-Elamite?
Probably not the same script, and the relationship is contested. Some scholars, including the Desset team that published the 2022 Linear Elamite decipherment, treat Linear Elamite as a descendant of Proto-Elamite. Others, including Jacob Dahl, argue that the graphic similarities reflect deliberate imitation rather than direct descent, and that Linear Elamite is a fresh phonographic system that borrowed the look of the older script.
Where can the tablets be seen today?
The Louvre Museum in Paris holds the largest single concentration of Proto-Elamite tablets, including the bulk of the Susa material. Other tablets are housed in Iranian museums (notably the National Museum of Iran in Tehran), the Peabody Museum of Harvard for the Tepe Yahya excavation, and a small number of other institutional collections. High-resolution images of much of the corpus are available open-access through the Cuneiform Digital Library Initiative.
Why was Proto-Elamite abandoned?
The script’s short two-century lifespan and the absence of any preserved scribal-school apparatus (no sign lists, no exercise tablets) suggest that Proto-Elamite never developed the institutional infrastructure that allowed proto-cuneiform to evolve into mature cuneiform. Administrative errors accumulated, the system did not train its replacements, and the region eventually adopted Mesopotamian-style cuneiform for its accounting needs.
