Antiquity 2024 page 1 of 19

https://doi.org/10.15184/aqy.2024.165

Research Article

Seals and signs: tracing the origins of writing in ancient

South-west Asia
Kathryn Kelley1 , Mattia Cartolano1 & Silvia Ferrara1,*
1
Department of Classical Philology and Italian Studies, University of Bologna, Italy
* Author for correspondence ✉ [email protected]

Administrative innovations in South-west Asia during

the fourth millennium BC, including the cylinder
seals that were rolled on the earliest clay tablets, laid
the foundations for proto-cuneiform script, one of
the first writing systems. Seals were rich in iconog-
raphy, but little research has focused on the potential
influence of specific motifs on the development of the
sign-based proto-cuneiform script. Here, the authors
identify symbolic precursors to fundamental proto-
cuneiform signs among late pre-literate seal motifs
that describe the transportation of vessels and textiles,
highlighting the synergy of early systems of clay-based
communication.

Keywords: South-west Asia, Uruk, Susiana, proto-cuneiform, seal iconography, signs

Introduction
The origins of writing in South-west Asia are often sought in the accounting systems that
developed over the course of the fourth millennium BC, which physically documented trans-
actions using tokens, tags and bullae (clay balls), numerical tablets and seals (Nissen et al.
1993). Proto-cuneiform, first attested on clay tablets at the city of Uruk in southern Iraq
c. 3350–3000 BC (Table 1, Uruk IV and III script phases), is a complex accounting system
with hundreds of iconographic signs, many of which still defy interpretation. Elaboration of
tokens, long used throughout South-west Asia, may have both stimulated the development of
proto-cuneiform and served as models for several signs (Schmandt-Besserat 1992: 140–49)
but token-to-sign comparisons, numerical notations aside, are rarely demonstrable (Englund

Received: 4 August 2023; Revised: 24 May 2024; Accepted: 7 June 2024

© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd. This is an
Open Access article, distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives
licence (https://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and
reproduction in any medium, provided the original work is unaltered and is properly cited. The written permission of
Cambridge University Press must be obtained for commercial re-use or in order to create a derivative work.
1

https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press

 Kathryn Kelley et al.

Table 1. Periodisation of the fourth millennium BC in Mesopotamia. Absolute and relative dating
between sites is problematic (Dahl et al. 2013).
Date Northern Southern
(BC) Mesopotamia Mesopotamia Writing
2900 Jemdet Nasr Uruk III script phase (c. 3200–3000 BC)
3100
3400 LC-5 Late Uruk Uruk IV script phase (c. 3350–3200 BC) Proto-literate
3600 LC-4 Middle Uruk Uruk V earliest numerical and Pre-literate
3900 LC-3 numero-ideological tablets (c. 3500 BC)
4200 LC-2 Early Uruk Cylinder seals introduced (c. 3700 BC)
4400 LC-1 Terminal Ubaid
We refer to Late Chalcolithic 4 (LC-4)/Uruk V/Susa Acropole 1.18 dating or earlier as pre-literate and LC-5/Susa Acropole
1.17a–b, with Uruk IV script as proto-literate. Many seals are difficult to date as securely pre- or proto-literate.

1993; Zimansky 1993) and the origin of sign forms is likely to be diverse, emerging from the
multimedia environment of visual expression (Michalowski 1990: 59).
Specific symbolic precursors to signs, apart from tokens, have not been the subject
of concerted research, despite the potential importance that their identification could
have for our understanding of the evolution of symbol systems and writing. Visual parallels
to sign forms have been sought in the Uruk Vase of the Uruk IV period (see Table 1;
Cooper 2008; Hockmann 2009; Selz 2022) and broadly contemporaneous cylinder seal
images. While Uruk-period seal imagery might be seen as a semiotic “forerunner to
script” (Pittman 2023: 238), with clear parallels between some seal and sign forms,
relationships remain difficult to define (Pittman 1994: 190–91) and it is possible that the
‘pictography’ observed in seals is unrelated to the development of writing (Glassner 2003:
166–77).
In this article, we present a more nuanced perspective on the origins of the script through
the exploration of promising correlations between proto-cuneiform signs and earlier, pre-
literate seal motifs from sites associated with the Uruk phenomenon of the late fourth mil-
lennium BC. Our analysis offers evidence for the contribution of seals and sealing practices
to the incipient use and form of proto-cuneiform signs. The motifs we discuss showcase the
common interest of seal-bearing administrators and scribes in documenting the movement of
textiles and vessels within or between cities.

Information technologies and the Uruk phenomenon

Proto-cuneiform appears in the latter part of the Uruk phenomenon, an almost millennium-
long sharing of material culture from south-west Iran to south-east Türkiye (Figure 1). Urban
complexity developed substantially during this period, although trajectories varied even in
neighbouring regions (Rothman 2001; Postgate 2002). In southern Iraq, where cities were
bolstered by irrigation agriculture and marsh resources (Pournelle & Algaze 2014), Uruk
became a centre of immense importance with the Tigris and Euphrates rivers facilitating
the long-distance movement of people and goods. Exports from Uruk may have included
dried fish and textiles (Crawford 1973), and communities were established far to the

© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
2

https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press

 Seals and signs
Figure 1. Map of archaeological sites with pre- and proto-literate seals corpora included in this study. For further details, see OSM (figure by authors).
© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
3
https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press
 Kathryn Kelley et al.

north, east and west for the import of wood, stone and metal, resources scarce in the south
(Algaze 1989; Oates 1993; Butterlin 2003).
Across the wide area of the Uruk phenomenon, seals and clay accounting technologies
were a key cultural touchpoint; the shared symbolic conventions supported networks of
elites both locally and inter-regionally (Sauer & Sürenhagen 2016). Trace element
and isotopic analyses indicate the local scale of clay vessel production and practices for
sealing and also the long-distance movement of the vessels and their seals (Minc &
Emberling 2016).
Stamp seals were used on clay for administrative purposes from the late seventh millen-
nium BC in northern Mesopotamia and were widely adopted and adapted in the following
millennia. By the mid-fourth millennium, the cylinder seal was in use in Iran, Iraq and Syria
(Pittman 2023). In addition to sealing lumps of clay to secure containers and doors, seals
covered the surface of clay balls (bullae)—which could bear numerical impressions and
enclose tokens—and were eventually used on blank or numerical tablets. Cylinder seals
and impressed objects with Uruk-related iconography appear in Iran, Iraq, Syria and Türkiye.
Seals were used in myriad ways, identifying institutions or production units, the senders of
goods, specific administrative events, or products and their destinations (Le Brun & Vallat
1978; Brandes 1979; Dittmann 1986, 2013; Nissen et al. 1993; Pittman 1994, 2018,
2023). Although some motifs illustrate production and exchange, seals represent cultural
ideals as much as realities (Breniquet 2016a). Sealing facilitated control over products,
with the shared iconography of the Uruk period suggesting “an inter-related if not integrated
administrative system” (Pittman 1994: 191) involving goods and people who are themselves
sometimes depicted on seals.
Shared seal imagery may have contributed to the maintenance of inter-regional connec-
tions as a form of fourth-millennium ‘commodity branding’ (Wengrow 2007), and the
movement of goods stored or transported in sealed vessels may have helped construct the
social world (Collins 2001; Englehardt 2013; Balke & Tsouparopoulou 2016). People are
sometimes depicted on seals in front of elaborate buildings—probably temples—receiving
products, suggesting that some commoditisation practices involved the movement of
goods through sacred institutions. Although seals in this period did not (with rare exception)
bear proto-cuneiform inscriptions, a ‘priest/ruler’ figure appearing in ritual scenes on seals has
been linked to the title En which appears in proto-cuneiform texts as the sign EN. The proto-
cuneiform sign is also used to characterise products, for example possibly marking an
‘EN-quality textile’. In many cases, goods listed on clay tablets are understood as ‘offerings’
to deities represented by their cult symbols (Szarzyńska 1997: 58–77).

Inter-related systems of symbols

Although seals are widespread, proto-cuneiform is attested only in southern Iraq. The earliest
currently known proto-cuneiform tablets are associated with the archaeological level Eanna
IVa at Uruk, approximately Late Chalcolithic (LC-5) in broader regional terms (Table 1).
The stimulation for this innovation remains enigmatic, though it is possible that the growth
of the urban population at Uruk may have created organisational challenges that proto-
cuneiform addressed (Algaze 2008: 139–41; Nissen 2016).

© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
4

https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press

 Seals and signs

The development of proto-cuneiform signs relied on pre-literate administrative practices,

including numerical notations (Englund 2006: 28–30), while extending their recording
potential (Nissen 2011). The practice of sealing tablets and other clay devices continues
after the introduction of proto-cuneiform, though a smaller number of tablets are sealed
(Matthews 1993). Motifs on tablet seals are difficult to correlate with tablet contents, an
undertaking that is further hampered by poor understanding of the script. A single Uruk
IV tablet, in which the script and seal impression both reference boars (Brandes 1979:
29–30), provides marginal evidence for the use of signs to reiterate seal information.
Nonetheless, iconographic and use patterns may hint at shifting functions for seals after
writing was invented (Scott 2018; Pittman 2023), perhaps implying that writing adopted
some functions formerly performed by seals.
Many of the approximately 800 proto-cuneiform signs remain poorly understood, with
the iconic referent of more than half either contested or unidentified. Proto-cuneiform
signs mentioned in this article are listed in Figure 2. All proto-cuneiform signs may depict
objects, even if we cannot recognise each one (Englund 1998: 55), or there may have
been any number of “totally abstract signs” (Nissen 2016: 44). Schematic signs, such as
UDU (‘sheep, wool’), may have been inspired by token traditions (Schmandt-Besserat
1992), but there has been no systematic assessment of the extent to which earlier or contem-
poraneous visual arts or other symbolic systems are recognisable in the sign shapes.
Most signs are probably functional, indicating products, offices, people, administrative
processes or events like festivals (Englund 1998; Johnson 2015). This recalls the functions

Figure 2. Images and suggested meanings of proto-cuneiform signs mentioned in the article. Signs are referenced by their
conventional label derived from later Sumerian or when especially contested or unknown, by their ZATU number
(Green & Nissen 1987). For further details, see OSM. (Sign and tablet images courtesy of the Cuneiform Digital
Library Initiative (https://cdli.mpiwg-berlin.mpg.de/) (figure by authors).

© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
5

https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press

 Kathryn Kelley et al.

proposed for Uruk seal motifs, although Pittman (2023) contends that seals in this period did
not identify specific individuals. Where identifiable, proto-cuneiform administrative content
further highlights the inter-related functions of these two information technologies. Proto-
cuneiform tablets, like sealing practices, were primarily used to track goods. Some pre-
and proto-literate seal iconography, particularly from the Susiana region in Iran (Pittman
2013), depicts agricultural and textile production, grain storage and animal husbandry, all
of which are also important themes in proto-cuneiform texts.

Comparing seals and signs

Interpreting seal images and signs and understanding their connection is challenging. Simple
comparisons of shape lack a sound methodological basis. For example, the ‘loops’ extending
from human heads in a late pre-literate seal impression from Susa, the administrative capital
of Susiana, have been interpreted as strings of dried fruit related to a proto-cuneiform sign
(Le Brun & Vallat 1978: 26–7) or a belt loom (Breniquet 2008: 313). Equally a proto-cuneiform
‘house’ sign (E2/ZATU129) with parallels in seal imagery may depict either a temple façade
(e.g. Charvát 1998: 23) or a loom (Glassner 2003: 153), or it may be deliberately ambiguous,
drawing on associations between weaving and important households (Breniquet 2008:
306–12).
Nevertheless, defining context-supported correlations between seal motifs and proto-
cuneiform signs can help to determine the role of seals in stimulating the invention of
writing, as well as clarifying the subsequent parallel development of these recording systems.
The potential for the interaction of scribes and seal carvers (Ross 2014) means that some con-
tested proto-cuneiform signs may be understood through seal iconography, as much as texts
may elucidate seal imagery. Similarities of shape may reflect shared referents to objects that
provide detail on shared administrative concerns. For pre-literate motifs, comparisons
might reveal functional transformations from seal to sign, where aspects of the administrative
scope of a motif are carried over into writing. Signs, nonetheless, become symbols with new
definitions in script (see caution in Michalowski 1994: 54–5) and such ‘symbol-hopping’
remains difficult to confirm. Contextual evidence can, however, strengthen this analysis.

Late pre-literate seal motifs and proto-cuneiform correlates

Survey of a corpus of seals from LC-1 to LC-4 contexts (c. 4400–3400 BC) from across
South-west Asia (Figure 1) reveals a small number of convincing parallels between pre-literate
seal elements and proto-cuneiform signs, supported by contextual clues (Figure 3). For
example, a ‘pole and net’ sign (ZATU639) that appears alongside cattle signs in tablets is
similar to an object sometimes shown with cattle in pre- and proto-literate seals (Figure 4).
The ‘façade and flag’/pole motif, in which the pole may be a cultic element (Dittmann
2013), is also reflected in proto-cuneiform signs (Szarzyńska 1997: 58–107). Notably, this
motif was already in use shortly before the invention of writing. The façade finds parallels
in the rectilinear proto-cuneiform signs, including the ‘house’ sign (E2/ZATU129,
Figure 3B), and the flag/pole parallels a ‘generic’ pole sign (ZATU523/URI3) and the god
Nanna’s cult icon (ZATU595/ŠEŠ, Figure 3C) (Michalowski 1993: 120–1; Steinkeller

© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
6

https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press

 Seals and signs
Figure 3. Comparisons of proto-cuneiform signs with pre-literate seal elements. Signs are rotated for comparison with iconography. For further details, see OSM (figure by
authors).
© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
7
https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press
 Kathryn Kelley et al.

1998; cf. Green & Nissen 1987). In seals,

products are often depicted being carried
towards the façade and, in an example from
Jebel Aruda in northern Syria (Figure 3C), a
pole appears beside a human and objects
including a zoomorphic vessel resembling
cosmetic or essence containers from Susa
(Alvarez-Mon 2020: pl. 21–22). Such por-
trayals suggest that seals tracked the move-
ment of valued goods.
A compelling case for the transformation
of pre-literate administrative symbols into
proto-cuneiform signs can also be made for
Figure 4. ZATU639 is comparable in both shape and
context to an impressed bulla from Susa. For further
a related pair of motifs—the fringed cloth
details, see OSM (figure by authors). (Figure 5) and the vessel in a net (Figure 6).
These appear widely among the ‘inter-
regional’ motifs (Dittmann 1986; Pittman
2013) on clay sealings, bullae and numerical tablets (see online supplementary material
(OSM) Appendix 1). Their presence on numerical tablets, in particular—a document
type marking the transition between pre-literate accounting and proto-literate writing—
argues for a trajectory of symbolic representation from seals to inscribed signs. Comparable
shapes may be seen in the proto-cuneiform textile or fringed cloth sign ZATU662 (and per-
haps others, e.g. ZATU663, Figure 5), and what we identify here as a netted/suspended vessel
sign (ZATU190 and ZATU191–194, Figure 2).
The fringed cloth (ZATU662) and a similar shape (ZATU644) are used at Susa and Uruk
in two ‘numero-ideographic’ tablets and are thus early signs that appear to have been sche-
matic from the outset. It is possible that some late pre- or proto-literate tokens of the ‘parab-
oloid’ group (Schmandt-Besserat 1992: 146/31.21–22) are related representations
(Figure 5D). In fact, triangular tokens may be reflected in several different proto-cuneiform
textile-related signs, but specific correlations are difficult. The sign we identify with fringed
cloth, originating in seal motifs, could represent a hide/pelt or a woollen skirt related to later
depictions of Sumerian attire (Szarzyńska 2002), the so-called kaunakes (Breniquet 2016b).
Similarly shaped textile signs may have had different meanings (Figure 5): a v-shape with
single dissecting line (ZATU644) is interpreted as Sumerian sumun or ‘old’ (for heirloom
garments?) (Szarzyńska 1997: 154–67), while a different v-shape (ZATU186) is securely
identified as a forerunner to gada, ‘linen’ (possibly depicting a flax-airing tool used in the pro-
duction of linen, Englund 1998: N. 353). The collection and redistribution of garments,
both domestically and for trading abroad, was an important activity in urban households
of early Mesopotamia from the Late Uruk period onwards (Englund 1998: 150–53).
We suggest that ZATU190 (Figure 6) depicts a ‘netted-vessel’ comparable to the common
seal motif (Boehmer 1999: 29–33). The sign is recognisable as a vessel enclosed in a net
(represented by cross-hatching) and either set in a base or fitted with ropes for suspension
(sometimes without the hatching). Although the conventional proto-cuneiform sign label
is GAN, this is probably an anachronism derived from the later cuneiform sign form (Deimel

© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
8

https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press

 Seals and signs

Figure 5. Comparison of the ‘fringed cloth’ motif and proto-cuneiform signs. For further details, see OSM (figure by
authors).

1922: no. 644) which eschews clear indication of the netting and handles and prominently
depicts the stand, and has the syllabic value/gan/ (cf. Akkadian kannu, ‘pot-stand’ CAD 8,
154 with possible reflex in Sumerian, Cohen 2023: 431). Similar netted vessels occur in
third millennium BC seals and plaques from southern and central Iraq, where they may
have contained oils (Potts 1997: 147). Variation in proto-cuneiform vessel signs may have
both general and specific semantic implications; the inclusion of spouts, for example, may
indicate liquid products (Englund 1998: N. 371). We can thus propose that a category of
vessel signs depicting nets/suspension/stands might reflect their contents, transportation or
assignment (e.g. ‘vessel delivery’, perhaps).
In seals, the fringed cloth and netted-vessel often appear individually or together in scenes
of production, transport and presentation to a building/person, or in heraldic configurations
(Figure 7, full list in OSM Appendix 1). The cloth emerges from the vessel or grows out of
humans or animals. These are productive motifs with dozens of examples showing variation,
and are found at pre-literate Susa and Uruk, pre- or proto-literate Jebel Aruda, Habuba
Kabira and Chogha Mish, and many sites in the proto-literate period. A variation in the
fringed cloth is even found at Saqqara in Egypt on a cylinder seal from the Naqada IId period

© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
9

https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press

 Kathryn Kelley et al.

Figure 6. Examples of the netted-vessel and fringed cloth motifs from pre-literate documents (A–C) and a
proto-cuneiform account of wool/garments and linen (D & E). For further details, see OSM (figure by authors).

(c. 3450–3325) (Figure 7 and ZATU688 Figure 3D), approximately contemporary with the
emergence of proto-cuneiform signs.
The range of products indicated by these seal images and signs is unclear. The vessel with
rope suspension may be inscribed with the NI (dairy oil, perhaps clarified butter) sign (Eng-
lund 1998: 168) (Figures 2 & 8B) or with the DIN (wine?) sign, among others. The NI sign
is among the few signs that have a plausible complex token parallel that was actually found
inside a bulla (from Uruk, Schmandt-Besserat 1992: 122). Dairy products represented by
vessels are often recorded alongside textiles on proto-cuneiform tablets. Historically, lanolin
(derived from wool) and other oils used to treat fibres are also associated with textile produc-
tion. The accounting link between products in vessels and textiles can be seen in an import-
ant practice text, the ‘List of Vessels’, which first mentions containers of dairy oils followed by
netted-vessels, and moves on to textile designations, starting precisely with fringed cloths
(ZATU663–2). However, further clarification is still needed. An Uruk IV account of
wool/garments and v-shaped textile signs, including gada (linen) (Figure 6D), finishes
with a reference to an EN-official and UDUa GANb (‘sheep/wool’, ‘netted-vessel’), which
may suggest that textiles could be transported in the netted-vessels (Collon 1987: 16).

© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
10

https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press

 Seals and signs

Figure 7. Distribution of netted-vessel and fringed cloth motifs. For further details, see OSM (figure by authors).

© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
11

https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press

 Kathryn Kelley et al.

Alternately, the fringed cloth and netted-vessel might reference linen and linseed oil, not-
ing a parallel from Egypt where the two were brought together as temple offerings (Boehmer
1999: 29–32). The fringed cloth sign ZATU662 (Figure 5) looks similar to the linen sign,
and seal impressions from Habuba Kabira in northern Syria and Chatal Höyük (Amuq) in
southern Türkiye show the cloth and netted-vessel motif associated with humans and a
plant that may be flax (Figure 7). Flax, from which both linen and linseed oil are derived,
was cultivated widely across South-west Asia in the fourth millennium BC, and its import-
ance in southern Iraq is highlighted on the Uruk Vase (Miller et al. 2016). Linseed oil,
derived from flax, is consumable and also a base for paints and lacquer among other uses.
It has been argued that fourth millennium South-west Asia experienced a shift away from
linen production towards wool, with dramatic consequences on labour, social organisation
and regional trade dynamics (McCorriston 1997; Algaze 2008: 77–79). If the netted-vessel
sign sometimes references linseed oil, this pushes back the earliest known textual evidence for
the use of this resource (cf. Waedzolt 1985).
Despite the uncertainties, a relationship between these motifs and signs is demonstrable
through several close associations (Figure 8). The netted-vessel sign is found alongside the
façade and flagpole, echoing combinations in seals (Figure 8B & D). A delivery of dairy
oils is documented on an Uruk III tablet in association with the signs E2 GAN (‘house’, ‘net-
ted vessel’)—perhaps a ‘vessel delivery’ to an institution/temple. The account also references a
household by a flag/pole on a hut (ZATU648, Englund 1998: 102; compare Figure 9G) and
lists an EN-official associated with a textile (ZATU753), which is seen elsewhere with the
fringed textile (Figure 8C). These sign correlations are mirrored in a Susian seal impression
on a numerical tablet and a bulla which combines a façade, flag/pole, netted vessel and
fringed textile motif (Figure 8A & D).
Thus, a picture emerges in which seals and texts operate synergistically to document the
delivery of similar products to temple households. Crucially, these seal motifs pre-date the
introduction of proto-cuneiform and appear on transitional types of records (bullae and
numerical tablets). Their adoption in script draws a specific continuity between pre-literate
symbol systems and the invention of writing.

Symbol-sharing in the proto-literate period

More broadly, there is the potential for considerable interplay between seals and early script in
the late fourth millennium BC (Figure 9). For example, a vessel sign with a cloth covering
(ZATU506) has a clear proto-literate seal parallel (Figure 9G). This sign may reference a
type of priest (Selz 2022: 220; cf. Greco 2014: 368–74), perhaps in charge of certain con-
sumables (Englund 2001: 25), and may belong to a similar administrative context as the net-
ted vessel sign.
Plant shapes reveal connections between script and seals beginning with the proto-literate
material (Figure 9B & C), and the water in a boar-hunting scene parallels the proto-
cuneiform sign ZATU1 (Figure 9F). Stylistic similarities are also suggestive: compare, for
example, the bovine head on a seal impression from Nineveh and ZATU219 on an Uruk
IV tag (Figure 9A). Elements in seals that find comparison with signs are often built into
scenes in natural ways, such as the mace and bow(?) among kneeling prisoners that might

© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
12

https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press

 Seals and signs

Figure 8. Comparison of inter-regional motifs from pre-literate seal imagery and proto-cuneiform tablets. For further
details, see OSM (figure by authors).

parallel a rare proto-cuneiform sign (ZATU359, Figure 9E). Combining contextual evidence
from seals and signs may even aid further interpretation. The undeciphered sign ZATU675
(Figure 9D) appears to be a schematic representation of an object in a proto-literate seal from
Uruk, perhaps a cult statue (cf. also ZATU25, Green & Nissen 1987: 320).
Many seal-sign comparisons draw on seal motifs first found in the proto-literate period.
Reed poles (ZATU523 and ZATU595, Figures 2 & 3C) appear already in late pre-literate
seals but the pole with shawl or tassel, representing the goddess Inana (MUŠ3/ZATU374,
Figure 9B), is not found in seals before the proto-literate period. Seals and signs may also
represent architectural embellishments with cultic significance (van Dijk-Coombes 2024;
cf. Szarzyńska 1997) (Figure 9G). A ‘heap’ or mountain corresponding to ZATU328/
LAM appears in an impression on an uninscribed proto-literate tablet from Uruk. The
motif later proliferates, especially in the proto-Elamite seals and writing from Iran (Fig-
ure 9H). The introduction of these new motifs into seals alongside the invention of writing
shows how seals and writing reflect each other in stylistic preferences.
Despite this reflexivity, seals and script retain their independence. With few exceptions
(Matthews 1993; Nissen et al. 1993: fig. 18; Pittman 2023: fig. 9.14) seals do not contain
signs. Images of humans and quadrupeds illustrate one of many ways in which the systems
adopt different iconographic strategies: whole figures are frequently represented in seals,
whereas proto-cuneiform depicts only separate body parts. In script, tools can represent

© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
13

https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press

 Kathryn Kelley et al.

Figure 9. Examples of parallels between proto-literate seals and signs. For further details, see OSM (figure by authors).

humans (e.g. the threshing sledge for agricultural workers, Figure 9I) and women are repre-
sented by vulvas (compare an early alabaster plaque from Mari on the Euphrates in eastern
Syria, Steinkeller 2019), while seals depict whole, clothed individuals.

© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
14

https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press

 Seals and signs

Seals and writing in (inter-)regional context

A set of ‘city seals’, bearing symbols representing cities, appear impressed on Uruk III tablets
originating from two southern Mesopotamian cities. The tablets record small numbers of
commodities, including dried fruits. By the Jemdet Nasr period, therefore, seals appear to
have been used in combination with proto-cuneiform tablets to record the exchange of
goods, probably as cultic offerings, that maintained relationships between Uruk and other
cities of southern and central Iraq (Matthews 1993; Steinkeller 2002; see also Michalowski
1993 on these seal inscriptions as a separate symbol system from proto-cuneiform). Such
exchanges could be seen as a strategy for strengthening ties as inter-regional connections wea-
kened following the Uruk phenomenon (Matthews & Richardson 2018).
Similar inter-city cultic exchanges may contribute to the invention of writing in the Uruk
IV phase, but the extent is unclear. The small-scale numerical notations on bullae and
numerical tablets may point towards local economic concerns (Englund 2006: 30) and a
focus on agriculture and local industries in tablets may be emphasised (Nissen 2016). Yet
proto-cuneiform grew from the administrative techniques that structured communities living
at great distances from one another, from southern Iraq through to south-eastern Türkiye.
Seal use and imagery demonstrate a shared articulation of the production and movement
of commodities across this region, prominent among which are netted-vessels and cloth,
sometimes in the presence of temple façades or before important individuals (Figure 10B).
It is not clear whether these motifs represent commodity processing enacted locally or
imply long-distance or inter-city exchanges. The modest offerings marked with ‘city seals’
suggest that even small exchanges could be inter-regional.
A striking correlation between sets of signs and pre-literate seal motifs emerges from the
comparison of a jar sealing from Uruk and an Uruk IV tablet accounting 21 products or ‘ves-
sel deliveries’ (Figure 10C). The vessels might have contained honey, syrup or wine since HI
‘sweet’ (or GIŠ) DIN are inscribed inside (Figure 10C top image, bottom right panel) (on
DIN as perhaps wine see Englund 1998: n. 116; on regional trade in wine during the
Uruk period see Algaze 1996). Seven deliveries are possibly associated with different officials,
including one that uses a symbol later also used for the city Adab in the ‘city seals’. On the
reverse, an unusual, linear summary line (cf. Green 1981) appears to mimic a rolled-out
cylinder seal, offering the sequence ‘suspended vessel’–‘alkaline plant’(?)–EN (Soriga
2017). Alkaline plants are associated with textile processing, but this proto-cuneiform sign
(ZATU381/NAGA) can also reference an administrative office, as well as the ancient city
EREŠ2 (location unknown).
Substituting the EN sign for a large official-looking figure (compare Steinkeller 2019), the
summary line on the Uruk IV tablet therefore parallels important elements on the contem-
porary jar sealing (Figure 10B). An Uruk III tablet fragment similarly correlates the ‘temple
vessel delivery’ (E2 GAN) with the alkaline plant and EN (Figure 10A). Thus, although inter-
pretative questions remain, some correspondence between seals and writing may be seen dur-
ing the era of script development, when symbols and some of their semantic associations were
likely reflected in signs. Correlations between tablets and seals also suggest that Uruk IV writ-
ing may have sometimes connected cities through the movement of goods in a tradition that
has antecedents in seal iconography and a possible successor in Uruk III ‘city seals’.

© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
15

https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press

 Kathryn Kelley et al.

Figure 10. Comparisons between netted-vessels, a distinctive plant, and a priest/ruler(?) in proto-cuneiform (A & C)
and seal imagery (B). For further details, see OSM (figure by authors).

Conclusion: seals and the invention of writing

Seals are a key consideration on the path towards writing in South-west Asia. At times, shared seal
and sign shapes may reflect common referents in the real world. Though the schematic forms of
proto-cuneiform may obscure a myriad of connections with seal motifs from modern eyes, recog-
nising commonalities can elucidate the respective cultural traditions. The fringed textiles and
netted-vessels show how pre-literate motifs could be transformed into signs in script through
a process in which they retained some of their semantic associations. The correlation of fringed
textile and netted-vessel signs with ‘façade’ and ‘pole’ signs on tablets strengthens their connec-
tion to late pre-literate administrative traditions that connected Uruk with cities across South-
west Asia, adding to our understanding of the stimuli for the invention of writing in southern
Iraq. Early in the development of writing, seals and signs continued to share shapes and reflect
each other, while remaining highly distinct information systems. Documenting the extent of
shape sharing helps build a more integrated, cohesive and, ultimately, more compelling explan-
ation for the development of writing. Such investigation casts light on the late fourth millennium
as a time of intense technological innovation, during which different modes of communication
were not only exploited in new ways but were also deployed in synergy.

© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
16

https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press

 Seals and signs

Funding statement
This article is funded by the European Research Council INSCRIBE Invention of Scripts
and their Beginnings, awarded to Silvia Ferrara. The project has received funding from the
European Research Council (ERC) under the European Union’s Horizon 2020 research
and innovation program, Grant Agreement No. 771127.

Online supplementary material (OSM)

To view supplementary material for this article, please visit https://doi.org/10.15184/aqy.
2024.165 and select the supplementary materials tab.

References
ALGAZE, G. 1989. The Uruk expansion: CHARVÁT, P. 1998. On people, signs and states.
cross-cultural exchange in early Mesopotamian Spotlights on Sumerian society, c. 3500–2500 B.C.
civilization. Current Anthropology 30: 571–608. Prague: Oriental Institute, Academy of Sciences
https://doi.org/10.1086/203786 of the Czech Republic.
– 1996. Fourth millennium BC trade in greater COHEN, M. 2023. An annotated Sumerian dictionary.
Mesopotamia: did it include wine? in State College (PA): Eisenbrauns.
P.E. McGovern, S.J. Fleming & S.H. Katz (ed.) The COLLINS, P. 2001. The Uruk Phenomenon: the role of
origins and ancient history of wine: 110–118. London: social ideology in the expansion of the Uruk culture
Routledge. during the fourth millennium BC. Oxford:
– 2008. Ancient Mesopotamia at the dawn of Archaeopress.
civilization: the evolution of an urban landscape. COLLON, D. 1987. First impressions. Cylinder seals in
Chicago (IL): University of Chicago Press. the Ancient Near East. Chicago (IL): University of
ÁLVAREZ-MON, J. 2020. The art of Elam ca. Chicago.
4200–525 BC. London: Routledge. COOPER, J.S. 2008. Incongruent corpora: writing and
BALKE, T.E. & C. TSOUPAROPOULOU. 2016. art in ancient Iraq, in P. Taylor (ed.) Iconography
Materiality of writing in early Mesopotamia. without texts (Warburg Institute Colloquia 13):
Berlin: De Gruyter. 69–94. London: Warburg Institute.
BOEHMER, R.M. 1999. Uruk: früheste Siegelabrollungen. CRAWFORD, H. 1973. Mesopotamia’s invisible
Mainz am Rhein: Philipp Von Zabern. exports in the third millennium B.C. World
BRANDES, M.A. 1979. Siegelabrollungen aus den Archaeology 5: 232–41.
archaischen Bauschichten in Uruk-Warka. https://doi.org/10.1080/00438243.1973.9979570
Wiesbaden: Franz Steiner. DAHL, J.L., C.A. PETRIE & D.T. POTTS. 2013.
BRENIQUET, C. 2008. Essai sur le tissage en Chronological parameters of the earliest writing
Mésopotamie: des premières communautés system in Iran, in C.A. Petrie (ed.) Ancient Iran
sédentaires au milieu du IIIe millénaire avant J.-C. and its neighbours: 353–78. Oxford: Oxbow.
Paris: De Boccard. DEIMEL, A. 1922. Die Inschriften von Fara. 1, Liste der
– 2016a. Weaving, potting, churning: women at work archaischen Keilschriftzeichen. Leipzig: J. C. Hinrichs.
during the Uruk period, in B. Lion & C. Michel DITTMANN, R. 1986. Seals, sealings and tablets:
(ed.) The role of women in work and society in the thoughts on the changing pattern of
Ancient Near East: 9–29. Berlin: De Gruyter. administrative control from the Late-Uruk to the
– 2016b. Que savons-nous exactement du kaunakès Proto-Elamite period at Susa, in U. Finkbeiner &
mésopotamien? Revue d’Assyriologie et W. Röllig (ed.) Ğ amdat Naṣr: period or regional
d’Archéologie Orientale 110: 1–22. style?: 332–66. Wiesbaden: Reichert.
https://doi.org/10.3917/assy.110.0001 – 2013. Glyptic and patterns of urbanization: a
BUTTERLIN, P. 2003. Les temps proto-urbains de humble approach, in T.R. Kämmerer & S. Rogge
Mésopotamie. Contacts et acculturation à l’époque (ed.) Patterns of urban societies: 35–138. Münster:
d’Uruk au Moyen-Orient. Paris: CNRS. Ugarit.

© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
17

https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press

 Kathryn Kelley et al.

ENGHLEHARDT, J. (ed.) 2013. Agency in ancient https://doi.org/10.1080/00438243.2019.

writing. Boulder: University Press of Colorado. 1592018
ENGLUND, R.K. 1993. The origins of script. Book MCCORRISTON, J. 1997. Textile extensification,
review of ‘Before writing’ by Denise alienation, and social stratification in ancient
Schmandt-Besserat. Science 260: 1670–71. Mesopotamia. Current Anthropology 38: 517–35.
https://doi.org/10.1126/science.260.5114.1670 https://doi.org/10.1086/204643
– 1998. Texts from the late Uruk period, in J. Bauer, MICHALOWSKI, P. 1990. Early Mesopotamian
R.K. Englund & M. Krebernik (ed.) communicative systems: art, literature, and
Mesopotamien: Späturuk-Zeit und Frühdynastische writing, in A.C. Gunter (ed.) Investigating artistic
Zeit: 15–233. Gottingen: Vandenhoeck & environments in the Ancient Near East: 53–69.
Ruprecht. Washington: Smithsonian Institution.
– 2001. Grain accounting practices in archaic – 1993. On the early toponymy of Sumer: a
Mesopotamia, in J. Høyrup & P. Damerow (ed.) contribution to the study of early Mesopotamian
Changing views on Ancient Near Eastern writing, in A.F. Rainey & R. Kutscher (ed.)
mathematics: 1–35. Berlin: Reimer. Kinattâtu sa darâti (Raphael Kutcher Memorial
– 2006. An examination of the ‘textual’ witnesses to Volume): 119–35. Tel Aviv: Institute of
late Uruk world systems, in G. Yushu & C. Yiyi Archaeology.
(ed.) A collection of papers on ancient civilizations of – 1994. Writing and literacy in early states: a
western Asia, Asia Minor and North Africa: 1–38. Mesopotamianist perspective, in
Beijing: Oriental Studies Special Issue. D. Keller-Cohen (ed.) Literacy: interdisciplinary
GLASSNER, J.-J. 2003. The invention of cuneiform: conversations. New York: Hampton.
writing in Sumer. Baltimore (MD): Johns MILLER, N.F., P. JONES & H. PITTMAN. 2016. Sign
Hopkins University Press. and image: representations of plants on the Warka
GRECO, A. 2014. Considerations on some digressions Vase of early Mesopotamia. Origini 39: 53–73.
in the Early Dynastic lexical lists. Orientalia 83: MINC, L. & G. EMBERLING. 2016. Trade and
341–79. interaction during the era of the Uruk expansion:
GREEN, M.W. 1981. The construction and recent insights from archaeometric analyses.
implementation of the cuneiform writing system. Journal of Archaeological Science: Reports 7:
Visible Language 15: 345–72. 793–97.
GREEN, M.W. & H.J. NISSEN. 1987. Zeichenliste der https://doi.org/10.1016/j.jasrep.2016.03.032
archaischen Texte aus Uruk. Berlin: Mann. NISSEN, H.J. 2011. Schule vor der Schrift, in
HOCKMANN, D. 2009. Die Warka-Vase eine neue G.J. Selz & K. Wagensonner (ed.) The empirical
Interpretation. Altorientalische Forschungen 35: dimension of Ancient Near Eastern studies:
326–36. 589–602. Berlin: LIT.
JOHNSON, C. 2015. Late Uruk bicameral – 2016. Uruk: early administration practices and the
orthographies and their Early Dynastic development of proto-cuneiform writing.
Rezeptionsgeschichte, in R. Dittmann, G.J. Selz Archéo-Nil 26: 33–48.
& E. Rehm (ed.) It’s a long way to a historiography NISSEN, H.J., P. DAMEROW & R.K. ENGLUND. 1993.
of the Early Dynastic period(s): 169–210. Münster: Archaic bookkeeping: early writing and techniques
Ugarit. of the economic administration in the Ancient
LE BRUN, A. & F. VALLAT. 1978. L’origine de Near East. Chicago (IL): University of Chicago Press.
l’écriture à Suse. Cahiers de la Délégation OATES, J. 1993. Trade and power in the fifth and
Archéologique Française en Iran 8: 11–59. fourth millennia BC: new evidence from northern
MATTHEWS, R. 1993. Cities, seals and writing: archaic Mesopotamia. World Archaeology 24: 403–22.
seal impressions from Jemdet Nasr and Ur. Berlin: https://doi.org/10.1080/00438243.1993.
Mann. 9980216
MATTHEWS, R. & A. RICHARDSON. 2018. Cultic PITTMAN, H. 1994. Towards an understanding of the
resilience and inter-city engagement at the dawn role of glyptic imagery in the administrative
of urban history: protohistoric Mesopotamia and systems of proto-literate greater Mesopotamia, in
the ‘city seals’, 3200–2750 BC. World Archaeology P. Ferioli (ed.) Archives before writing: 177–203.
50: 723–47. Torino: Ministero Beni Culturali.

© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
18

https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press

 Seals and signs

– 2013. Imagery in administrative context: Susiana and sealing in the ancient world: 68–80.
and the west in the fourth millennium BC, in Cambridge: Cambridge University Press.
C.A. Petrie (ed.) Ancient Iran and its neighbours: https://doi.org/10.1017/9781108160186.005
293–336. Oxford: Oxbow. SELZ, G.J. 2022. Beyond speech: advocating a
– 2018. Administrative role of seal imagery in the non-logocentric view on the evolution of
Early Bronze Age: Mesopotamian and Iranian cuneiform writing, in D. Wengrow (ed.) Image,
traders on the plateau, in M. Ameri, S. Kielt thought, and the making of social worlds: 213–49.
Costello, G. Jamison & S. Jarmer Scott (ed.) Seals Heidelberg: Propylaeum.
and sealing in the ancient world: 13–35. https://doi.org/10.11588/propylaeum.842.
Cambridge: Cambridge University Press. c10812
https://doi.org/10.1017/9781108160186.004 SORIGA, E. 2017. A diachronic view on fulling
– 2023. Iconicity, composition, and semantics: a technology in the Mediterranean and the ancient
structural investigation of pictures in an early Near East: tools, raw materials and natural
writing environment, in I. Zsolnay (ed.) Seen not resources for the finishing of textiles, in S. Gaspa,
heard: composition, iconicity, and the classifier C. Michel & M.L.B. Nosch (ed.) Textile
systems of logosyllabic scripts: 237–71. Chicago terminologies from the Orient to the Mediterranean
(IL): Institute for the Study of Ancient Cultures. and Europe, 1000 BC to 1000 AD: 24–46.
POSTGATE, N. (ed.) 2002. Artefacts of complexity: Lincoln (NE): Zea.
tracking the Uruk in the Near East. Warminster: https://doi.org/10.13014/K2CJ8BNP
Aris & Phillips for the British School of STEINKELLER, P. 1998. Inanna’s archaic symbol, in
Archaeology in Iraq. K. Szarzyńska & J. Braun (ed.) Written on clay and
POURNELLE, J. & G. ALGAZE. 2014. Travels in Edin: stone (Szarzyńska Memorial Volume): 87–100.
deltaic resilience and early urbanism in greater Warsaw: Agade.
Mesopotamia, in A. McMahon & H. Crawford – 2002. Archaic city seals and the question of early
(ed.) Preludes to urbanism: the Late Chalcolithic of Babylonian unity, in T. Abusch (ed.) Riches
Mesopotamia. Cambridge: McDonald Institute hidden in secret places: 249–57. Winona Lake
for Archaeological Research. (IN): Eisenbrauns.
POTTS, D.T. 1997. Mesopotamian civilization: the – 2019. Texts, art and archaeology: an archaic plaque
material foundations. Ithaca (NY): Cornell from Mari and the Sumerian birth-goddess
University Press. Ninhursag, in G. Chambon, M. Guichard &
ROSS, J.C. 2014. Art’s role in the origins of writing: A.-I. Langlois (ed.) De l’argile au numérique:
the seal-carver, the scribe, and the earliest lexical mélanges assyriologiques en l’honneur de Dominique
texts, in B.A. Brown & M.H. Feldman (ed.) Charpin: 977–1011. Leuven: Peeters.
Critical approaches to Ancient Near Eastern art: SZARZYŃ SKA, K. 1997. Sumerica. Prace
295–317. Boston: De Gruyter. sumeroznawcze. Warsaw: Dialog.
ROTHMAN, M. 2001. Uruk Mesopotamia and its – 2002. Sheep husbandry and production of wool,
neighbours. Santa Fe (NM): School of American garments and cloths in archaic Sumer. Warsaw:
Research. Agade.
SAUER, C. & D. SÜRENHAGEN. 2016. Zählmarken, VAN DIJK-COOMBES, R.M. 2024. The standards of
Zeichenträger und Siegelpraxis: Einige Mesopotamia in the third and fourth millennia
Bemerkungen zu vor- und frühschriftlichen BCE. Tübingen: Mohr Siebeck.
Verwaltungshilfen in frühsumerischer Zeit, in WAEDZOLT, H. 1985. Ölpflanzen und pflanzenöle im
T. Balke & C. Tsouparopoulou (ed.) Materiality 3. Jahrtausend. Bulletin on Sumerian Agriculture
of writing in early Mesopotamia: 11–46. Berlin: 2: 77–96.
De Gruyter. WENGROW, D. 2007. Prehistories of commodity
https://doi.org/10.1515/9783110459630-002 branding. Current Anthropology 49: 7–34.
SCHMANDT-BESSERAT, D. 1992. Before writing. https://doi.org/10.1086/523676
Austin: University of Texas Press. ZIMANSKY, P. 1993. Book review of ‘Before writing’
SCOTT, S.J. 2018. Slave labour: Uruk cylinder-seal by Denise Schmandt-Besserat. Journal of Field
imagery and early writing, in M. Ameri, S. Kielt Archaeology 20: 513–17.
Costello, G. Jamison & S. Jarmer Scott (ed.) Seals https://doi.org/10.1179/jfa.1993.20.4.513

© The Author(s), 2024. Published by Cambridge University Press on behalf of Antiquity Publications Ltd
19

https://doi.org/10.15184/aqy.2024.165 Published online by Cambridge University Press