1.1.1 Document Forms

To focus our attention on the script and its writers, I have adopted the concept of “document form” as an analytical framework from the study of archival documents, a discipline called “diplomatics.” A document form is defined as any text whose external physical features, such as the materials it is made of and written with, its formatting, and so on, conform to “rules of representation” (see Table 1; Duranti, Reference Duranti1989, p. 15). Rules of representation come into existence by virtue of being taught and reproduced in specific institutional settings, either formal or informal ones, from one generation to the next. A study of document forms is therefore also a study of the institutional settings in which writers produced texts. Unlike most archival documents, which were written on objects designed to be written upon, Cypro-Minoan often appears on objects produced for other purposes. In such instances, instruction and text production probably occurred outside of institutions like scribal schools, whose primary purpose was instruction in the production of texts.

A document form is not the same as writing medium. Writing medium refers to the type and material of the object on which writing appears. The clay tablet is one example of a writing medium. It encompasses a wide range of document forms, some quite dissimilar from each other. Compare the cuneiform culture clay tablet and Linear B clay tablet pictured in Figure 1. The cuneiform tablet has a “pillow” shape, compact rule lines, and tightly packed wedge-shaped sign components produced with a triangular or square-tipped stylus. The Linear B tablet is flat and rectangular with clear, wide rule lines and “linear” sign forms drawn with a bladed stylus. Both are examples of the clay tablet writing medium but different document forms. Their material features, from their shapes, line ruling, and writing implements to their formatting, are distinctive.

Figure 1 Comparison between Linear B and cuneiform tablets. Linear B Tablet PY Vn 10, cuneiform tablet from Anatolia, probably Kanesh, Met. 66.245.10.

Photograph of PY Vn 10 courtesy of the Pylos Digital Tablet Project, Palace of Nestor Excavations, the Department of Classics, University of Cincinnati.

Institutional contexts play a defining role in dictating the rules of representation of a document form. In the cuneiform cultural world, instruction in the production of clay tablets was transmitted primarily in scribal schools. Surviving school texts indicate that scribal schools maintained similar curricula over a 3,000-year lifespan, spreading eventually from Mesopotamia to Anatolia. Scribes received training in how to construct clay tablets, which writing implement to use, and how to use it, with the result that cuneiform clay tablets show remarkable uniformity in construction, script, and formatting. Only experts can readily distinguish regional and chronological differences among tablets made in different times and places. How signs are drawn remains extremely stable throughout time (Taylor, Reference Taylor, Devecchi, Müller and Mynářová2015). Cypro-Minoan document forms, in contrast, are more varied, but their variations seem to occur within a rather narrow range. In other words, its writers followed rules of representation when constructing document forms, even on objects not purpose-made for writing. Take, for instance, the example of the vessel text document form (see Figure 2). Writers selected a rather narrow range of vessels to write on, often preferring to incise texts onto handles post firing (Hirschfeld, Reference Hirschfeld2008). Formatting is restricted to horizontal or vertical placement along the handle axis, and most texts begin at the top of the handle before it bends. Variation is expressed in sign shape and the width, depth, and shape of incision lines. The adherence to a broad set of rules seems to indicate that writers of Cypro-Minoan received instruction in how to write and how to make document forms but not the strict training characteristic of scribal school education. One aim of this Element is to describe the nonschool institutional settings where writers of Cypro-Minoan practiced writing and to think about how writing can be transmitted outside of the school setting.

A. CM 20.199/ T1907_193;

B. CM 1962_99;

C. ##115 ENKO Avas 008,

D. Melos inscription.

Photographs and drawings by author, courtesy of the Cyprus Museum and the Department of Antiquities, Cyprus.

Figure 2 Examples of the vessel text document form.

1.1.2 How We Define Writing Shapes What We See

The study of script can change the way we think about writing. Writing is often conceived of as a tool whose value derives from its ability to convey language in written form. Writers of Cypro-Minoan had a more nuanced practice of writing. Take the vessel text document form, once again. It includes around 100 texts labeled “inscriptions” by scholars of Cypro-Minoan because they contain multiple script signs in a row and therefore likely record a word in a language. In addition to the 100 inscriptions, there are over 1,500 single-sign texts. The single-sign texts comprise a mix of script and nonscript signs. Some texts with script signs probably record abbreviations of words, like the “B” on a Boston Red Sox cap mentioned above, but others were probably more similar to McDonald’s golden arches, sharing the shape of a script sign but able to convey meaning regardless of whether its viewer could read its script. The nonscript signs, for their part, were perhaps more akin to the Nike swoosh, capable of evoking a clear association in their viewer without the use of writing. The writers of the vessel document form apparently did not erect strict barriers between script signs that record language and other signs. They used script signs to communicate something besides language and nonscript signs in a manner similar to script signs.

But what can writing be used for if not to convey language? Several East Asian countries have long-standing calligraphic traditions that appreciate writing for its aesthetic qualities, like any other branch of the arts. Brands and graffiti employ a mix of script and nonscript elements to convey extralinguistic information. While Cypro-Minoan texts have a certain beauty, it is not apparent that their writers cultivated an art of calligraphic writing. Branding and graffiti may be a better fit for explaining Cypro-Minoan’s mix of script and script signs, but how do these very contemporary concepts translate into the world of the Late Bronze Age?

1.2.1 The 17–15th Centuries

The first Cypro-Minoan text appears long before the Bronze Age collapse, in the Late Cypriot I (LCI) period beginning roughly in 1650 if not earlier (see Figure 3). The 17th century on Cyprus marks renewed contact with eastern and western neighbors after a prolonged period of insularity. Increased extraction of agricultural and copper resources translated into increased social stratification and the concentration of population-dense towns, especially on the island’s coasts. Enkomi, which plays a central role in the use of writing on the island, may have been the main power center on the island, controlling copper extraction in the interior of the island from the coast, or one among several regional powers. Trade with the Levant and Egypt was brisk, while mercantile connections with Crete and Cilicia may have been limited to the trade of precious oils or perfume (Webb, Reference Webb and Bourogiannis2022).

Figure 3 Approximate synchronized chronological chart of the Late Bronze Age eastern Mediterranean.

The decision to begin to write may have been in response to renewed contact with neighbors, to internal social changes, or for no particular reason at all. There are no hard-and-fast rules for determining the circumstances under which people decide to invent or adopt a script. “Script transfer” is the process wherein a script is deployed to write a different language in a new context. Script transfer can come at the hands of an inspired individual/group or result from a cumulative process called “script diffusion” where adjustments and additions to a script happen slowly over decades or centuries (Salomon, Reference Salomon and Haralambous2021, pp. 153, 167). Often, script transfer results from a combination of a singular, conscious act of creativity followed by a long period of diffusion and reforms. The processes involved in script transfer are rarely transparent in the historical record and are highly variable. Script transfer can be conservative, wherein the shapes, sounds, and document forms of one script are retained and applied to writing a new language, or it can be selective. A script can be borrowed without its document forms, and sign shapes can be borrowed without their sound values, as was done by the inventor of the Cherokee syllabary, Sequoya, who adopted shapes of some letters from the Roman alphabet to represent entirely different sound values in his syllabary (Scancarelli, Reference Scancarelli, Daniels and Bright1996). Script transfer can also be combinatory, adapting signs from multiple scripts or even nonscript sources to create a new script.

The invention of the Cypro-Minoan script is believed to have involved script transfer from Linear A, one of the main scripts used on Crete and, to a lesser extent, the Cyclades from c. 1800–1450 BCE. Linear A was also the source script for the Linear B script used on Crete and the Greek mainland to write an early form of Greek called Mycenaean Greek from c. 1400–1190 BCE. Though Linear B and Cypro-Minoan were contemporaries, there is no evidence that the two scripts influenced one another at all. Some form of script transfer from Linear A to Cypro-Minoan is almost certainly assured, but its nature and extent are debated. Linear A may be one of several sources for Cypro-Minoan script signs. The small number of Cypro-Minoan texts from the 17th and 16th centuries, no more than four inscriptions and around fifty vessel texts, make reconstructing the circumstances of script transfer difficult.

Possible evidence for the setting of script transfer is a single-sign text on a discoid loom weight text from a small site in north-central Crete, Kalo Chorafi, dated broadly from 1800–1600 BCE. The discoid loom weight bears a single sign similar in shape to Cypro-Minoan script sign CM 061 (for a list of signs, see Section 2, Figure 7). The sign is definitely not a Linear A script sign, but neither is it incontrovertibly Cypro-Minoan. It is a sign of a simple shape that could also be a decorative motif. Until there is other evidence for Cypro-Minoan writing on Crete during the 17th century, the significance of the loom weight text is hard to interpret. Its presence could but does not necessarily indicate that Cypro-Minoan was invented on Crete. Alternatively, the sign could be a decoration or nonscript mark that later made its way into the Cypro-Minoan repertoire, or its similarity to a script sign could be coincidental.

The earliest certain Cypro-Minoan inscription, on a discoid loom weight from Enkomi, ##095 ENKO Apes 001, is datable to the LCI (see Figure 4). A second LCI inscription is on a clay tablet, also from Enkomi, ##001 ENKO Atab 001, but its script is not clearly related to Cypro-Minoan, only to (for an opposing view, Valério, 2018). Tablet ##001 has received a lot of attention from Cypro-Minoan scholars (see Steele, Reference Steele2019, pp. 19–25 for an overview), but ##095 is just as, if not more, interesting. The object’s function had previously been a matter of debate, with some proposing it is an economic label (Ferrara, Reference Ferrara2012, pp. 53–56). Its identification as a discoid loom weight was confirmed only recently when the first study of LCI weaving technology was conducted by Giulia Muti in 2024.

A. Loom weight ##095 Apes 001, h. 6.3, w. 7.2 cm. Note that the inscription is upside down as pictured.

Photograph and drawing of ##095 by author, courtesy of the Cyprus Museum and the Department of Antiquities, Cyprus.

B. Photograph and drawing of the Kalo Chorafi loom weight.

Courtesy of Anastasia Tzigounaki, Director of the Systematic Excavation of Kalo Chorafi, Mylopotamos, Crete, Director of the Ephorate of Antiquities of Rethymno, and Dr. Artemis Karnava.

Figure 4 Comparison between inscribed discoid loom weights.

Discoid loom weights were a technology associated with a particular form of textile production that began in Crete and spread throughout the Aegean, including to the Cyclades. Women likely played an important role in its spread, and therefore possibly that of Cypro-Minoan (Muti, Reference Muti2024). Caution is needed. though. Evidence for script transfer in weaving contexts is not quite there. The two texts are on the same writing medium but different document forms. The Kalo Chorafi loom weight carries a large text centered on the surface of the seal, while the Enkomi weight carries its small text placed along its bottom edge. Moreover, the Kalo Chorafi loom weight is not a certain Cypro-Minoan text. The Enkomi inscription ##095 is the only Cypro-Minoan inscription on a loom weight of any kind to date.

A potential candidate for the setting of script transfer could be the Cyclades, where Linear A was written (and where discoid loom weights were used). A likely Cypro-Minoan inscription has been identified as coming from the Cycladic island of Melos (Figure 2D). Its script was previously unidentified, but a recent Cypro-Minoan identification is compelling because a second Cypro-Minoan inscription on a vessel handle carries the same two-sign text (Figure 2C) and because the inscription fits the rules of representation of the Cypro-Minoan vessel handle document form (Donnelly, Reference Donnelly, Salgarella and Petrakisforthcoming). If the inscription is in fact written in Cypro-Minoan, then it is the only early inscription outside of Cyprus and the only one on a document form that continues into use past the period of script transfer. The strong associations between the vessel handle document form and trade could suggest that script transfer occurred in mercantile contexts (Hirschfeld, Reference Hirschfeld2008), but there is scant evidence of a trade relationship between Cyprus and the Cyclades in the 17th and 16th centuries.

Without the discovery of new evidence, whether from Cyprus, Crete, or the Cyclades, the earliest period of Cypro-Minoan will remain poorly understood. For now, all that can be surmised is that the script underwent a long period of script diffusion, during which some elements of the Linear A script were adapted. Traders and/or individuals involved in textile production might have participated in the early spread of Cypro-Minoan, but the current evidence is simply too sparse for a clear picture to emerge.

1.2.2 The Amarna Period (c. 1350)

By the mid 14th century, during the so-called Amarna Period, Cyprus is a major player in overseas economic exchange. A cache of clay tablets, known as the Amarna Letters, discovered in the newly founded Egyptian capital of Akhetaten – modern Tell el-Amarna (whence Amarna Letters) – testifies to the economic prowess and integrated overseas economy of the eastern Mediterranean from roughly around 1360–1330 BCE. Written in the cuneiform script, these tablets mainly record the Akkadian language, which was being used as a diplomatic lingua franca (the classic translations are found in Moran, Reference Moran2000).

Cyprus is a correspondent in the Amarna Letters. It is given the place name “Alashiya.” Petrographic testing conducted on clay from the Amarna Tablets from Alashiya finds a good source for the clay in the southwest of Cyprus (Goren et al., Reference Goren, Finkelstein and Na’aman2004). Nevertheless, no cuneiform inscriptions have been found on Cyprus aside from some short, rote texts on cylinder seals. The existence of the Alashiya letters indicates that someone(s) on Cyprus had some formal cuneiform scribal training in how to write Akkadian, but current evidence suggests cuneiform was not widely used on Cyprus. Meanwhile, the use of Cypro-Minoan was slowly gaining steam and, despite fewer than twenty Cypro-Minoan inscriptions dating to the 14th century, would have been the dominant script used on the island.

The Amarna Letters document the economic and diplomatic maneuvering of the major players in the Late Bronze Age Mediterranean and Near East (see Figure 5). The major players were ruled by “Great Kings” (LUGAL.GAL in cuneiform) who refer to one another as metaphorical “brothers,” distinguished by the term of address they use for their vassals, “son.” An economic hierarchy of the Late Bronze Age in the 14th century has been reconstructed through a study of how the kings address one another and their subordinates. The Great Kings are the king of the Hittites in Anatolia, whose empire is at its height during the Amarna period, the king of the Land of Mitanni in northern Syria (associated with the Hurrian language), a once-proud empire on its last legs, the pharaoh of Egypt, and, more peripheral for our purposes, the kings of Assyria and Babylon. The kings of cities on the Levantine coasts, including the city of Ugarit in northern Syria, where Cypro-Minoan texts would later be written, all had vassal or “son” status. The palatial centers of Mycenaean Greece, where Linear B was written, were not mentioned.

Figure 5 Script distribution in the Late Bronze Age eastern Mediterranean.

Scholars frequently omit Cyprus from the list of Great Kings, but Cyprus’s status as diplomatic equal with Egypt is clear (Humphrey, Reference Humphrey, Furlan, Husøy and Bohun2022). The king of Cyprus and the pharaoh call one another “brother,” the form of address that was restricted to the Great Kings. Cyprus’s equal diplomatic footing can be further intimated from the attitude and veiled insults – the shade – its king throws at the pharaoh, the likes of which subordinate kings would hardly dare. Yet the high diplomatic status of Cyprus and its king in the Amarna letters is belied by the material remains on Cyprus, where archaeological evidence for a king and a politically unified state is missing.

Many but not all archaeologists believe that 14th-century Cyprus was not a unified political entity but a “heterarchy” comprised of cities of relatively equal wealth and power, which each controlled its own hinterland (Peltenberg and Iacovou, Reference Peltenburg and Iacovou2012). Enkomi was one such city. What allowed Cyprus to rub proverbial elbows with the Great Kings was its vast copper reserves, which were being exploited to their fullest extent, and its traders, who played the role of middlemen between Mycenaean Greece to its west and the Levant and Egypt, circulating metals, oils, and wine (among other things). However this trade was conducted, it was likely not at the behest of a single centralized state with a king at its top (for an opposing view, Knapp, Reference Knapp2008, pp. 324–341). There is no royal iconography or architectural evidence of a single, centralized administrative and distributive center. Who the “king” of Cyprus was, whether he was a representative of one city or elected by all the cities to represent their interests in diplomatic correspondence, is a matter of continued academic debate.

The “internationalism” of the Amarna period increasingly gave way, in the early 13th century, to wars between the Hittites and Egyptians and general instability (for a summary, see Cline, Reference Cline2021). One source of instability must have been the climate, which was aridifying at a swift pace. Another was intensified attacks by the so-called Sea Peoples, a group of marauding pirates straight from central casting. Despite their fantastical element, the Sea Peoples are noted in Egyptian, Hittite, and Ugaritic literary sources as a genuine destructive force. Scholars today paint the Sea Peoples variously as climate refugees or economic opportunists taking advantage of an unstable economic system. By the end of the 13th century and beginning of the 12th, the instability culminates in the so-called collapse of the major Amarna players, either through destruction or political weakening – except for Cyprus.

1.2.3 Before, during, and after the “Collapse” (c. 1250–1050)

The peak of Cypro-Minoan script use coincides with the so-called collapse period, better conceived of as a period of transition, in which large-scale, state-backed globalization was replaced by the small-scale but far-reaching independent trade networks like the ones writers of Cypro-Minoan participated in before and then during and after the collapse (Sherratt, Reference Sherratt2003). In the Aegean, the major Mycenaean administrative centers of Mycenae and Pylos were destroyed and never used again. Same for the Hittite capital Hattuša and its vassal kingdom of Ugarit. Egypt sees the loss of its colonial empire and the division of the country into two parts, northern and southern.

What these places have in common is their reliance on centralized administration. The more decentralized, mercantile societies like Cyprus, and perhaps also Byblos and Sidon on coastal Lebanon or the traders of Ura on the Anatolian coast of Cilicia, could tap into far-reaching mercantile networks to maintain or create wealth. The breakdown of centralized administered states was accompanied by the loss or atrophy of the region’s administrative scripts, namely cuneiform and Linear B. Some scripts were wiped from the pages of history until looting and archaeological investigation in the 19th century CE brought them back to light. One such script is Linear B, which went out of use completely and suddenly with the destruction of the Mycenaean palaces c. 1190 BCE. When Linear B was rediscovered in the late 19th century, no one knew that it was used to record the Greek language until its decipherment a half-century later in 1952 (for a wonderful video summary, www.youtube.com/watch?v=iePEw_cHp8s). Another is the alphabetic cuneiform script of Ugarit, a locally invented, short-lived script (c. 1250–1180) that went out of use so completely that its existence was wholly forgotten until its rediscovery and decipherment in the 20th century.

An equal but opposite effect was seen in the scripts of mercantile societies, which eventually blossomed after a lag of a century or two. In Anatolia, the local Luwian hieroglyphic script survived the collapse only to resurface as one of main scripts of the Neo-Hittite empires of the Iron Age in the 10th century BCE. But Luwian hieroglyphic inscriptions from the 12–10th centuries are few and far between (Hawkins, Reference Hawkins2024). On the Levantine coast, the precursor to the Phoenician alphabet had been used since at least the 16th century, but inscriptions come in dribs and drabs until the end of the 11th century (Rollston, Reference Rollston2010). Only then did its use steadily increase until it gained enough steam to be adapted to write Greek, Etruscan, and Latin, becoming a template for our own alphabet. Of these scripts, only Cypro-Minoan is well documented during the late 13th and well into the 12th century. It therefore can serve as a proxy for thinking about how the scripts belonging to the other mercantile societies may have been transmitted in the absence of centralized administration.

Throughout the shifting landscape of the 12th century, Cyprus retains its wealth, conducts overseas trade, and exploits its copper resources, but it also undergoes changes. Some cities were abandoned while others underwent restructuring, like Enkomi, which centralized its copper workshops and revamped its sanctuaries. Sometime just before the destruction of Ugarit c. 1180, the king of Ugarit addresses letters to the “king of Alashiya,” a certain Kušmešuša (the only Cypriot king whose name has been recorded), by calling him “my father,” indicating that Cyprus and its king had a higher status than the king of the important port city (Singer, Reference Singer, Watson and Wyatt1999, pp. 719–720). As ever, though, the island’s political organization remains opaque, and there is little to no archaeological evidence for a unified state or a king.

The Cypro-Minoan script sees its furthest geographic extension during the collapse period. It was written locally in small amounts at Ugarit, just prior to its c. 1180 destruction, then at Tiryns in Greece just before and after the collapse of the Mycenaean palatial centers perhaps a decade earlier c. 1190. Short Cypro-Minoan inscriptions (which could have theoretically been written anywhere) are found on shipwrecks off the Anatolian and Levantine coasts. In the later 12th century, a single Cypro-Minoan inscription written on a northern Levantine vessel makes its way to Ashkelon in the southern Levant.

The use of Cypro-Minoan at Ugarit corresponds to the period of peak usage of the local alphabetic cuneiform script, Ugaritic (c. 1250–1180). The city was a major node in overseas and overland trade, an unhappy vassal of the Hittite kingdom of Anatolia. Texts written in at least five different scripts and nine different languages are found at the site, a testament to its connectivity and vitality (Boyes, Reference Boyes2021). Enkomi and Ugarit, facing one another a short one-day’s sail across the sea, had close cultural, economic, and even political ties. Yet in other ways the two cities are culturally distinct in their political structures and their main local scripts (and probably languages).

The close relationship between Ugarit and Cyprus raises interesting questions about how and why we find Cypro-Minoan texts in Ugarit. As discussed in Section 4, traders may have been integral to the dissemination and local production of Cypro-Minoan texts. At Ugarit, traders conducted trade both in their own names and in the names of the king or queen of Ugarit. Exactly how and on whose behalf Cypriot traders worked is not something we currently understand, but it is conceivable that the commercial initiative afforded to Ugaritian traders was echoed across the sea at Enkomi. The ability of individuals to move goods and strike deals on their own behalf could explain how traders from Ugarit and Cyprus became intimately acquainted enough to share writing practices and how writers of Cypro-Minoan were able to retain use of their script during the collapse that saw Ugarit destroyed.

2.1.1 Verifying Decipherment

For a full and successful decipherment, readings of proposed sound values and language identification must be verifiable. Proposed sound values should be tested and applied to multiple documents, not just one. Language identification cannot rest solely on the identification of individual words in a given language but also on grammar. In cases of scripts that record known languages, decipherments are relatively easy to verify. For instance, Michael Ventris suspected that his decipherment was successful when he was able to read the word tripod on a Linear B tablet. The word tripod is comprised of strong Greek roots (tria = three, pous = foot) and appeared on the tablet next to a drawing of a three-footed vessel. But it was not until his proposed readings for the signs yielded results matching Greek grammar that the decipherment was generally accepted. The true test of decipherment is grammar, not the ability to read individual words.

Grammar encompasses the way that different words or parts of words (morphemes) relate to one another in a sentence. Often, the position of words in a sentence and the form of words change depending on their grammatical role. Take, for instance, the changes in the simple sentences, “I give you a hot dog” and “you gave me the hot dog”: I/me changes forms based on its role in the sentence, give/gave changes according to tense, and the word “you,” though remaining unchanged in form, moves word position as its role changes. Grammatical features are a language’s fingerprint, distinguishing it from other languages even as certain features are inevitably shared with them.

Grammar is a much more reliable indicator of language identification than individual words. Michael Ventris’s identification of tripod is a case in point. The etymology of the word is certainly Greek, but it is a word that has been inherited into English, Italian, and French and could appear in texts recording any of those languages. Names can likewise be misleading. My first name has a clear Greek etymology, but my native language is English, and I have no familial roots in Greece. Individual words are not good indicators of the language recorded by a script, but grammar is.

Verification of decipherment requires replication of results across multiple documents. It is precisely on this count that Cypro-Minoan presents the most impediments. Very few texts are long enough to exhibit complex grammar: only six inscriptions are longer than 100 words (##097–098, ##207–209, ##215), which is an impediment to decipherment itself, and the longer texts, for reasons discussed in Section 2.3, might record different languages from one another. If this is indeed the case, then not only is the prospect of decipherment quite low but so is the ability to verify a decipherment.

2.1.2 What We Know about Cypro-Minoan

A first step to decipherment is determining the type of script (is it an alphabet? a syllabary?) and the number and nature of sounds in the script. The type of script can usually be identified by counting the number of unique signs it exhibits. Alphabets usually have the fewest number of signs (fewer than 50), syllabaries between 50 and 100 signs, and logographic writing systems hundreds. Insights into phonology, the number and nature of sounds in a language, are much more difficult to realize. One almost always needs to know additional information about the script, such as the languages spoken in and around its home region or if it is related to scripts with known sound values. In that case, one can make educated guesses about the script’s phonology, that is, the number and types of sounds it records. The field of linguistic typology, which aims to discover general tendencies in how languages work, can provide general guidelines for testing hypotheses but is not always a reliable tool because of the many exceptions that individual languages present.

In the case of Cypro-Minoan, the identification of the script as a syllabary has been inferred from a comparison to the related Linear A and Cypro-Syllabic scripts. Linear A provides the source for anywhere from twelve to thirty of Cypro-Minoan’s around ninety signs (see Figure 7), some of whose sound values are known through comparison to Linear B. Cypro-Syllabic is the “daughter” script of Cypro-Minoan, used c. 1000–300 BCE. Cypro-Syllabic adapted a significant portion of Cypro-Minoan sign shapes in its script transfer from Cypro-Minoan, reducing the number of signs to fifty-five. It is not clear whether Cypro-Syllabic inherited the sound values of Cypro-Minoan except in a handful of cases. The Cypro-Syllabic script was used to record at least two languages. The first is the main language recorded in Cypro-Syllabic inscriptions, Greek, which was widely spoken on Iron Age Cyprus, and the second an unknown language or languages, recorded on fewer than thirty inscriptions, most from the 5th and 4th centuries BCE, referred to as Eteocypriot, supposedly a language indigenous to Cyprus (Egetmeyer, Reference Egetmeyer2010). Comparing Cypro-Minoan to Linear A and Cypro-Syllabic reveals some basic insights but has not led to the ability to read most Cypro-Minoan signs nor to identify its underlying language(s).

Comparison of Cypro-Minoan to Linear A and Cypro-Syllabic suggest that it is a syllabary that records vowel signs and open syllabic signs (i.e., ba, pa, ra, and so on). Secure sound values for nine to eleven signs are obtained by comparing the sound values of signs with the same shape across Cypro-Minoan, Linear A/B, and Cypro-Syllabic. The nine to eleven signs suggest that Cypro-Minoan has at least four vowel sounds, a, e, i, and o, and the following consonants, -s, t/d, r/l, and -p. Attempts to derive new information from comparisons across the three scripts are common but often not successful. Miguel Valério, as recently as Reference Valério2016, advanced a more accurate method of sign-shape comparison, basing his comparison not on the abstract drawings of sign shapes (sometimes called “normalized” drawings), as his predecessors had done, but on the actual examples of sign shapes in inscriptions. Despite his improvement to the comparative method, his interscript comparisons result in readings for only about thirty signs, many of which are still highly contested, sometimes requiring signs to be flipped upside down, turned around, or otherwise reimagined to produce shape matches. The insights into the script gained by comparisons to Linear A and Cypro-Syllabic are limited.

Only a single feature of grammar has been identified with any degree of certainty: the ending -Co-ti (C = consonant of any value), which is believed to indicate a genitive or dative (“belonging to” or “to/for” X). The -Co-ti feature has been identified through comparison to bilingual Greek–Eteocypriot Cypro-Syllabic inscriptions (Masson, Reference Masson1971, p. 26). Although the Eteocypriot language remains unidentified, when its signs are read with and compared to the Greek words in the bilingual inscriptions, it is apparent that the -Co-ti affix indicates a genitive or dative, often with an inscription dedicated “to/for” or “belonging to” a name ending in -Co-ti. Luckily, the Cypro-Syllabic ti sign has the same shape and sound value as the Linear A sign for ti. It is therefore likely that the Cypro-Minoan sign of the same shape, CM 023, has the same ti sound value. CM 023 appears frequently in word-final position in Cypro-Minoan inscriptions. Although this is often taken as proof that the Cypro-Minoan inscriptions that show possible -Co-ti endings and Eteocypriot record the same language, other explanations are possible. It may be the case that a Co-ti feature was shared by two different languages because of their proximity, a linguistic process called areal diffusion, or through frequent contact.

2.2.1 Computational Approaches

Perhaps the most important prerequisite for decipherment is the accurate identification of unique signs and their range of sign variants. Accurate sign identification is necessary for getting a true count of the total number of signs in a script, for identifying how a sign behaves in words (i.e., its preferred word position), and for identifying features of the underlying language’s grammar. The shape of individual script signs can vary considerably and for many reasons: personal whimsy, physical limitations, regional differences in scribal training, or simply because training in a certain script does not impose uniformity on its writers. Different fonts illustrate . One of the main tasks in decipherment is to determine if signs of similar shape are the same sign with same sound value, that is, variants of the same unique sign, or if they are different unique signs with their own sound values. Is the serifed and dotted “i” of Times New Roman the same sign as the simplified line i with an acute mark from Papyrus as the dotted line of Comic Sans? Comparing only the shapes of signs to identify whether they are unique signs or variants of one another can be misleading. A much more reliable method is to observe how a sign behaves, its preferred word position, the signs it frequently appears next to, and repeated sign sequences in which it appears.

Computational analysis is one of most important tools that decipherers have for constructing a sign’s behavioral profile. The pioneer of computational analysis in the decipherment of Aegean scripts, working without a computer, is the American scholar Alice E. Kober (1906–1950), whose work was instrumental in Michael Ventris’s eventual decipherment of Linear B. Instead of trying to “read” Linear B by guessing sound values, Kober assembled and analyzed lists of Linear B words to observe sign behavior (see Figure 6). Due to paper shortages in World War II, Kober assembled word lists on homemade index cards cut from scrap paper and stored in cigarette cartons (Fox 2013). On her index cards, she catalogued how many times each sign appeared, its preferred word position, which signs it appeared next to, lists of repeated sign sequences, and more, conducting a decades-long, long-form computational analysis of sign frequency and word position, unaided by computers.

Figure 6 Image from Alice E. Kober’s Word List notebook with “Kober’s triplets” inset in black.

Notebook image courtesy of the Program of Aegean Scripts and Prehistory, University of Texas, Austin.

Positional analysis of signs and permutations of repeated sign sequences are potent tools for identifying individual signs and their variants. Imagine you are an alien beamed down to earth and given a code to decipher. The code consists of two sets of numbers, 1352, 8901352, 13526, 1752, and . Being an alien, you have never seen numbers before. One of your first steps toward decipherment will be to determine if the codes are identical. Looking at the shapes of the signs alone will lead you down a wrong path. The first sign in each set, “1 and ,” for instance, have similarities and differences in their shape which may or may not be meaningful. Comparing how the two signs behave in each set of code is simpler and more conclusive: Both signs appear frequently in word-initial position, are often followed by sign 3, and recur in a repeated sign sequence 1352. It is therefore highly likely that 1 and are variants of the same unique sign.

Next, you might try to identify the encoded language by looking for grammatical features recorded in the code, 1352, 8901352, 13526, 1752, and comparing them to known languages. A simple computational analysis would show that “1352” is a repeated sequence that can take a prefix (890), a suffix (6), and an internal modification (1352 v. 1752) called an ablaut. The language recoded in the code can thus be narrowed down to languages with suffixes, prefixes, and ablauts, such as Germanic languages. In fact, that’s what the code records: the English words “give,” “forgive,” “given,” and “gave.” Kober used a similar process in her study of Linear B, identifying a grammatical element called declension in sets of three repeated sign sequences that showed internal and word-final permutations, known today as “Kober’s triplets” (see Figure 6). Although Kober was not able to use this information to decipher the script herself, Kober’s triplets were instrumental in Michael Ventris’s realization that Linear B records Mycenaean Greek.

Cypro-Minoan has proven resistant to traditional computational analysis approaches like Kober’s. The inscriptions are simply too short and too few to ascertain statistically significant information about each sign’s preferred word position. There are also very few repeated sign sequences and even fewer with permutations. New technologies may prove instrumental in providing a way forward. Advances in statistical modeling can assist, among other things, in determining the positional frequencies of infrequently attested signs with a degree of statistical significance unreachable using older methods. Christina Skelton, in collaboration with a team of computer scientists from the University of California San Diego, for instance, refined computational methods to produce behavior profiles for Cypro-Minoan signs across all inscriptions (Skelton et al., Reference Skelton, Selvig, Chen, Srivatsan and Berg-Kirkpatrick2022). Skelton and her team concluded that individual signs had different positional frequency across different inscriptions, a result that they interpreted to mean either that different inscriptions recorded different languages or different topics/genres of documents (more on that in Section 2.3).

Advances in artificial intelligence technologies (AI), such as deep learning and neural networks, have made great strides in the realms of translation and language identification using less and less input data. Deep-learning-based neural networks tasked with identifying the sound values of unknown signs recording known languages have been successful, including one tasked with identifying the underlying language of Linear B (for an overview, Ferrara and Tamburini, Reference Ferrara and Tamburini2022). A team of script scholars and computer scientists in Italy has trained neural network technologies to analyze and identify Cypro-Minoan sign shapes (Corazza et al., Reference Corazza, Tamburini, Valério and Ferrara2022). The fields of Machine Translation (MT) and Natural Language Processing (NLP) have developed automatic translation tools to target translations for languages with small datasets and without known language data inputs. Any of these technologies, or a combination thereof, might one day be successfully applied to attempts to decipher Cypro-Minoan. For any of these new computational methods to one day be successful, however, the most important task is for scholars of Cypro-Minoan to develop as accurate a signary as possible.

2.2.2 Signaries

Computational analysis of sign frequency and position can only be as accurate as its data. The greatest barrier to Cypro-Minoan decipherment efforts at present is the lack of an agreed-upon inventory of signs, or “signary.” There is no agreement on the number of Cypro-Minoan signs because there is no agreement on which signs are unique and which are variants. Historically, the greatest impediment to the creation of an accurate signary was the inability of scholars to distinguish between script and nonscript signs. Today, the greatest difficulty is figuring out how to distinguish unique signs and their variants without the aid of traditional computational analysis.

Distinguishing script from nonscript signs was difficult for early scholars of Cypro-Minoan because the earliest inscriptions were few and short. As we know now, single-sign texts contain an assortment of script and nonscript signs, but for much of the history of Cypro-Minoan there was no method for distinguishing script from nonscript signs. The earliest signaries simply aggregated lists of every sign depicted on every known text. The result was that they often included what are today known to be nonscript signs.

The first method for distinguishing script from nonscript signs was developed by the American archaeologist and wartime spy John Franklin Daniel, who had been the lead excavator at the site of Episkopi-Bamboula in the 1930s, where his team discovered over eighty Cypro-Minoan texts, most of them single-sign texts on vessels, tripling the total number of Cypro-Minoan texts (Daniel, Reference Daniel1941). Faced with so many new texts, Daniel sought to impose order on a chaotic dataset. He developed a hierarchical assessment for determining whether a sign was a script sign or not. He counted a sign as a script sign with confidence, first, if it appeared in a multisign inscription, which he believed guaranteed that the sign was used to spell a word and hence to record language; second, with less confidence, if a sign with the same shape was attested in another Aegean script (such as Linear A); and third, with even less confidence, if the sign frequently recurred on single-sign texts.

After World War II, when a trove of longer inscriptions was unearthed, Daniel’s method for identifying script signs was no longer needed. Scholars simply could count a sign as a script sign if it appears in a multisign inscription. Since then, scholars have tried and failed to develop signaries using computational approaches based on positional frequencies and repeated sign sequences. One of the more ambitious approaches to the signary, of late, was taken by Miguel Valério, who combined computational analysis with cross-script comparisons of Cypro-Minoan signs with Linear A and Cypro-Syllabic. Valério’s signary produced a wide-ranging number of possible signs, fifty-seven on the low end and seventy on the high end, showing both the benefits and limitations of more traditional approaches to Cypro-Minoan (Valério, Reference Valério2016, pp. 166–171, 442).

A novel approach to the signary was taken by Martina Polig, who created a signary based on an analysis of sign shape alone (Polig and Donnelly, Reference Polig and Donnelly2022, see Figure 7). Sign shape has not historically been regarded a useful unit of analysis in decipherment (see the discussion of “1 vs. ”), but Polig harnessed the advantages provided by new technologies to create an accurate dataset analyzable on a greater scale than before. Using structured-light scanning technology, she created millimeter-accurate 3D models that recorded the shape of every undamaged sign from every available Cypro-Minoan inscription, and then she catalogued, measured, and compared each sign, which enabled her to observe “rules” according to which sign variants were produced from unique signs.

Figure 7 Polig and Donnelly’s “integrated signary.”

Image courtesy of Polig and Donnelly, Reference Polig and Donnelly2022.

The result is an eighty-nine-sign signary, called an “integrated signary” because it presents the unique signs along with their variants. It is the signary used throughout this Element when referring to signs. In Cypro-Minoan studies, it is common to refer to signs by their number in the signary since their sound values are not known. The numbers Polig chose for her signs adhere as closely as possible to the numbering system that was first put into place by Émilia Masson in the 1970s and which was followed, with modifications, in subsequent signaries (for a history of Cypro-Minoan signaries, see Polig and Donnelly, Reference Polig and Donnelly2022). I have chosen to present the integrated signary here because it is the only signary built from accurate images of the whole spectrum of Cypro-Minoan inscriptions and the only one to represent sign variants alongside unique signs. It shows the wide variation of Cypro-Minoan sign shapes in all its glory.

3.1.1 Earliest Discoveries

Cypro-Minoan can be seen as something of a Rorschach test of colonial ideologies. The origins of the script became a focal point for determining the “ethnic” character of the island’s ancient and modern inhabitants. Depending on the political situation, European and Greek Cypriot scholars alike posited European (specifically Greek/Cretan), “Oriental,” and sometimes indigenous Cypriot origins for the script. A template for a vacillating Cypriot identity was provided by its Iron Age past when its cities were split between Greek and Phoenician rule. Already in the 19th century, the many Phoenician inscriptions found on the island were taken to demonstrate the Levantine, “Oriental” ethnic character of the island’s inhabitants. That conclusion was upended in the 1870s, however, when the British autodidact George Smith deciphered Cypro-Syllabic and demonstrated that most Cypro-Syllabic inscriptions record the Greek language.

Smith’s decipherment impacted different communities in different ways. For many Greek Cypriots, the decipherment reinforced their sense of Greek ethnic identity. It also lent ideological backing to the enosis movement’s fight to join the Greek nation state. Many European archaeologists, at least at first, welcomed Smith’s decipherment as proof that writing on Cyprus was a European invention, another feather in their superior European caps. But exactly at the time enosis gained momentum in the 1940s and 1950s and posed a serious threat to colonial rule, as first noted by archaeologist Michael Given, they pivoted to the idea that writing on Cyprus was an “indigenous” tool (Given, Reference Given1998). The earliest discoveries of Cypro-Minoan and their subsequent interpretations were impacted by the broader political climate on the island at most every turn.

The first Cypro-Minoan texts, unearthed in the late 1890s by a privately funded British expedition to Cyprus, were short and single-sign texts on vessels and short inscriptions on clay balls. They contained no more than fifteen unique signs among them on fewer than thirty texts (Evans, Reference Evans1909, p. 68). Despite their brevity and small number, all were interpreted as writing. Sir Arthur Evans, the British archaeologist who established the discipline of Aegean scripts and whose belief in the racial superiority of Europeans permeates his scholarship, posited as early as 1909 that Cypro-Minoan descended from Minoan Linear A. In his hands, Cypro-Minoan inscriptions were taken as proof of European ingenuity and Europe’s pivotal role in the spread of writing from west to east (Vidal, Reference Vidal and Fenollós2014). Evans posited that a small but influential group of Minoans colonized parts of Cyprus and the Levantine coast, where they invented not only Cypro-Minoan but also the alphabet (Evans, Reference Evans1909, pp. 74, 77). If the invention of writing could not be claimed by Europe, since that honor belonged to the cuneiform writing of Iraq and the hieroglyphs of Egypt, Evans wove a tale in which “European” Minoans invented the most ingenious and efficient of writing systems, the alphabet.

Today, no scholars believe that “Minoan” Linear A was the progenitor of the Phoenician alphabet. Studies show that the alphabet is no more efficient than syllabic and logographic writing systems, and scholars question the value of applying the very modern, Western concept of efficiency to ancient contexts. Nevertheless, the myth of the alphabet’s efficiency persists (Baroni, Reference Baroni2011). The idea that Minoans created the alphabet was always a transparent attempt by Evans to disassociate the alphabet from the “oriental” Levant and transport it to the “European” Aegean. The signs of the Phoenician alphabet are clearly modeled Egyptian and hieratic signs according to the acrophonic principle, where the first letter of the word depicted by the hieroglyph becomes the sign’s sound value (e.g., the first letter of the Phoenician alphabet “aleph” derives from a hieroglyphic sign depicting an ox’s head, or alef, in Semitic languages).

After Evans, the next scholar to look at Cypro-Minoan was Menalaos Markides, the first Curator of the Cyprus Museum from 1912 to 1931. The Greek Cypriot Markides was a pioneer of scientific archaeological methods and applied the same level-headed circumspection to Cypro-Minoan (Pilides, Reference Pilides and Webb2018). Even during the war years when the museum’s budget was mean, Markides went to great effort to continue excavations and document the island’s past. He was rewarded with the discoveries of eleven new Cypro-Minoan texts, comprised exclusively of short multisign and single-sign vessel texts. Markides published them, along with a signary, in short reports in consecutive years, 1915 and 1916. Markides’s signary counted some twenty-two unique signs, thirteen of which he attributed to Aegean origins and the rest not (Markides, Reference Markides1916).

Subsequent European scholars returned to Evans’s Aegeo-centric interpretation of the script. Swedish excavations in the 1920s added to the list of Cypro-Minoan texts, again mostly in the form of short and single-sign vessel texts. These were published along with a signary by the Swedish archaeologist Axel Persson, discussed in Section 2, who took an unabashedly Eurocentric approach to the script, so sure the inscriptions were written in Greek that he attempted a decipherment to that effect. The British colonial officer and scholar Stanley Casson was much more measured than Persson in his inventory of signs and interpretation of the script but did little to advance the study of Cypro-Minoan beyond the Evans paradigm. Of the four scholars responsible for the early studies of Cypro-Minoan, only Markides, the Greek Cypriot, questioned the wholesale European origins of the script.

Perhaps the most innovative pre-War approach to Cypro-Minoan was taken by the American archaeologist and World War II spy John Franklin Daniel, mentioned in Section 2. Daniel created a classification system that sorted texts by their material features, the method in which they were inscribed (incised after firing vs. impressed before firing), and the origins of their writing medium (imported vs. exported). Like Markides, he theorized that some classes of Cypro-Minoan inscriptions were not Aegean inventions but Cypriot ones.

3.2.1 “Achaean” Tablets?

Some of the fiercest debates over the ethnic identity of Cypro-Minoan writers focused on the seven clay tablets, three from Enkomi and four from Ugarit, discovered in quick succession in the 1950s (Ferrara argues there are four clay tablets from Enkomi, Ferrara, Reference Ferrara2012, pp. 192–195). The Enkomi tablets were, in the hands of Dikaios, evidence of the Greek colonization of; in the hands of Schaeffer, evidence of Semitic colonizers. That two scholars could arrive at diametrically opposed conclusions looking at the same artifacts, both conclusions based in colonialist fantasies, speaks more about the political climate in which the discovery of the tablets took place than the features of the tablets.

Porphyrios Dikaios was the Curator of the Cyprus Museum (1932–1960) and the first Director of the Department of Antiquities, Cyprus after independence (1960–1963). Dapper and cosmopolitan, the Nicosia-born Dikaios was selected by the colonial government, like Markides before him, to receive archaeological training abroad. Like Markides, Dikaios returned to the island and implemented a level of scientific rigor in his excavations that superseded most of his contemporaries working on Cyprus and elsewhere. Dikaios was assigned by the Department of Antiquities, Cyprus, to excavate at Enkomi concurrently with French Expedition excavations at Enkomi led by Schaeffer as a stopgap against what were considered the latter’s destructive excavation methods (Papasavvas, Reference Papasavvas2023, p. 41).

Dikaios excavated two of the site’s three clay tablets (##207 and ##208) in consecutive years, 1952 and 1953, finding them in contexts he dated roughly to the 12th century just after the “collapse.” The dating coincided well with his theory that the island was colonized by Achaean Greek refugees in the aftermath of the destruction of the Mycenaean palaces. For Dikaios, the “Achaean” authorship of the two tablets was never in question. Given the opportunity to publish the tablets, as is customary for the lead excavator, he selected Aegean scripts specialists John L. Myres, nicknamed “Blackbeard” for his World War I naval espionage, and Michael Ventris, the decipherer of the Linear B script, to collaborate with him (Dikaios, Reference Dikaios1953, Reference Dikaios1956). Although Dikaios was known for his impartial and apolitical public persona, it is hard not to see his theory of Achaean colonizers and his choice to call them by the Homeric term “Achaean” as an attempt, conscious or not, to ground modern Cypriot identity in a Greekness dating back to the imagined age of Homeric heroes.

Claude Schaeffer, the lead excavator of the French Expedition at Enkomi and at Ugarit, disagreed vociferously with Dikaios’s dating and interpretation of the Enkomi tablets. Schaeffer’s disagreements were rooted in his belief that an Achaean colonization of Cyprus had happened earlier in the 14th century and that the 12th-century “collapse” period marked the colonization of Cyprus by the Sea Peoples, whom he believed to be ethnically Semitic. An earlier 14th-century date for the Achaean invasion allowed Schaeffer to posit that Ugarit was a Mycenaean (i.e., Greek) colony, an echo of Evans’s theory that the Phoenician alphabet was invented by Europeans.

Schaeffer recruited Cypro-Minoan to support his theory of successive colonizations. He “read” the shapes of sign on the Enkomi and Ugarit tablets as if they were clues indicative of the ethnicity of their writers. In the Ugarit tablets, he found “linear” paleography characteristic of ethnic “Mycenaeans,” and in the Enkomi tablet a “cuneiform” aspect characteristic of ethnic Semites. His method of reading ethnicity into sign shapes which is a reduction to absurdity, nevertheless supplied the germ for a theory cultivated and popularized in the subsequent decades, that Cypro-Minoan represented not one single script but three subscripts representing three languages, one belonging to Cyprus (CM 1), another to the supposed immigrants to Enkomi who made the tablets (CM 2), and another to Ugarit (CM 3).

That the clay tablets became a focal point for fantasies about ethnicity and colonization in the 1950s and 1960s echoes the intensity of the anticolonial struggle of the 1950s and the tumultuous years of independence in the 1960s. Colonial fantasies were expressed in scholarship, however, in ways that were sometimes confounding or contradictory.

3.2.2 The Paleography of the Enkomi Tablets

The Enkomi tablets (“CM 2”) became the focus of ethnic theories in the 1970s (Figure 11). They were interpreted as the products of a distinct ethnic group because of the perceived unity in their material features, but even this unity is, to an extent, in the eye of the beholder. The paleography across the three tablets is certainly similar, with its compressed size and simplification of sign forms, but the same paleography was described by Émilia Masson, the main scholar of Cypro-Minoan in the 1970s, as “squared and squat” but by Myres as “Minoan,” that is, drawn and linear (see Figure 12; Masson, Reference Masson1974, p. 15; Dikaios, Reference Dikaios1956). Aside from the paleography and line ruling, the tablets are different from one another in the color of their clay, tablet shape, and formatting. The perceived uniformity between the Enkomi tablets nevertheless became the basis for the claims that they were written by a distinctive ethnic group.

A. ##209 ENKO Atab 004;

B. ##208 ENKO Atab 003;

C. ##207 ENKO Atab 002. Note that ##209 is a cast that does not reflect the true color of the tablet.

D. Close-up of ##208;

E. close-up of ##207. Note the similarities and differences in sign shape, ductus, and ruling.

Photographs by author, courtesy of the Cyprus Museum and the Department of Antiquities, Cyprus.

Figure 11 Enkomi tablets on display at the Cyprus Museum with close-ups.

Figure 12 Close-up of ##207 showing distinctive Enkomi tablet paleography.

Photograph by author, autopsy courtesy of the Cyprus Museum and the Department of Antiquities, Cyprus.

3.2.3 Masson’s Trifold Division of the Script

Émilia Masson’s trifold division of Cypro-Minoan into the CM 1, CM 2, and CM 3 subscripts is heavily influential in Aegean scripts studies and Cypriot archaeology (Table 2). Masson distinguished the subscripts by what she believed to be their distinctive paleography and sign repertoires, which she associated with the different linguistic and therefore ethnic identities of their writers: CM 1 was the script of local Cypriots, CM 2 of Hurrian (i.e., Anatolian) immigrants to Enkomi, and CM 3 of Ugaritians (Masson, Reference Masson1974). In a departure from Dikaios and Schaeffer, whose debates centered on Greeks and Semites, Masson’s main “ethnic” players were indigenous Cypriots and Anatolians. The influence of Greece is relegated to a bit part, relevant only to the earliest Cypro-Minoan inscriptions, and the influence of “Semites” is restricted to the inscriptions found in Ugarit.

The changing political landscape on the island in the 1970s and the simplistic worldview that equates language, script, and identity provide a context for Masson’s subdivision. Writing mainly in the 1970s, Masson’s theory of the Hurrian origins of the CM 2 Enkomi tablets coincides with the Turkish (Anatolian) occupation, encompassing the northern third of the island including the site of Enkomi. Although the timing might seem overly coincidental, as seen throughout this section, it is often the case that scholarship is influenced by contemporary politics. It can be no coincidence, for instance, that this Element fixates on “collapse” during a moment of heightened global crisis. Hurrian speakers, who we met in Section 1 in relation to the Amarna Letters and the Kingdom of Mitanni, were major players in the Late Bronze Age. In that sense, her focus on Hurrians was not out of place. According to Masson’s theory, Hurrian-speakers were a cohesive ethnic group who, displaced from their homeland in Mitanni by invading Hittites, relocated to Enkomi where they adapted the Cypro-Minoan script to write their language (see Figure 13). Her theory is not corroborated, however, by other archaeological evidence. For Masson, the inscriptions were proof enough: A script’s paleography and the language it recorded revealed the ethnicity of its writers.

Figure 13 Map of Cyprus in context, indicating sites on Cyprus with Cypro-Minoan inscriptions.

Masson’s subdivision ultimately suffers from the faulty assumption that ethnic difference can be reduced to linguistic difference and that linguistic difference is expressed in script choice. One of the takeaways from this Element, I hope, is the lesson that script and language are not the same. Masson’s subdivision is based on both real and perceived differences among Cypro-Minoan inscriptions but flattens those exciting differences in favor of the old colonial and archaeological obsession with ethnicity.

3.3.1 HoChyMin

It was not until 2007 with the publication of Jean-Pierre Olivier’s Édition holistique des textes chypro-minoens, known by its acronym HoChyMin, that an official corpus of Cypro-Minoan inscriptions was published. The Belgian scholar had a long history of publishing the corpora of undeciphered scripts. In partnership with Louis Godart, he had published the corpus of Linear A, which goes by the delightful acronym GORILA, and that of the Cretan hieroglyphic script, the modish CHIC. HoChyMin was a departure from the other corpora, not only in its head-scratching acronym but also in being an édition holistique or “holistic edition.”

Traditionally, Aegean script corpora are compiled by a person or team who organizes and publishes (or republishes) as much information as possible about all recorded inscriptions in a given script. A corpus either is the first publication of new inscriptions from the site or, in the case of subsequent editions, republishes older inscriptions and adds new ones. The case of Cypro-Minoan was different in that inscriptions had already been published, just not assembled, hence “holistic edition.” Olivier understood his remit to be the assembly of a “un patchwork composé avec les membra disjecta” (“a patchwork composed of dismembered parts”), a phrase I quote in the original French for its mélange of French, English, and Latin, a great example of how language identification based on short snippets of text can be misleading (Olivier, Reference Olivier2007, p. 16).

Olivier did not find his job of assembling the Cypro-Minoan patchwork easy. The large number of sites (20+) where Cypro-Minoan writing has been found, the over 600-year timespan in which Cypro-Minoan was in use, and the large number of short or damaged texts whose status as writing is ambiguous posed considerable problems. He suspected even at the time of publication that he had omitted or missed inscriptions due to time and age constraints (Olivier, Reference Olivier2007, p. 16). The unwieldiness of the corpus required him to make difficult choices about which texts to include and exclude. Ultimately, he chose to exclude single-sign texts from the corpus even if he believed many to be writing. There were simply too many, from too many different sites (membra disjecta) to assemble within a reasonable time frame. Excluding them had the benefit of providing uniformity to what was included, namely texts with two or more contiguous script signs written on the same plane. As Olivier suspected, it is now clear that his corpus is missing inscriptions, around eighty, most on vessels, published in obscure publications or in short reports.

HoChyMin benefited Cypro-Minoan studies greatly in that it provided scholars with the first synoptic view of the script since the flood of new inscriptions discovered in the 1950–1970s. Olivier organized the inscriptions according to an ingenious system that encodes an inscription’s writing medium and findspot in its corpus number. For instance, the eighth recovered vessel inscription from Kition was given the corpus number ##137 KITI Avas 008 (see Figure 14). Less helpfully, inscriptions were published usually without accompanying images of the objects they appear on. Olivier also adhered to Masson’s subscript division despite his reservations concerning its utility (HoChyMin, p. 16) and even added a new subscript designation, CM 0, for the earliest “intermediate” Enkomi tablet ##001 ENKO Atab 001, discussed in Section 1.

Figure 14 Olivier’s inscription numbering system. Also adopted by Ferrara in CM II (Reference Ferrara2013).

Drawing by author based on autopsy, courtesy of the Cyprus Museum and the Department of Antiquities, Cyprus.

3.3.2 CM II

A focus on materiality is evident in the second corpus of Cypro-Minoan inscriptions, CM II, published shortly on the heels of HoChyMin in 2013 by the Italian scholar Silvia Ferrara. Following the suggestion of American scholars Nicolle Hirschfeld and Joanna Smith, Ferrara’s catalogue of inscriptions foregrounds archaeological context over paleography and, for the first time, provides photographs of the objects on which inscriptions appear as opposed to only its text (Smith and Hirschfeld, Reference Hirschfeld1999). Thanks to this decision, it is possible for scholars to observe the material aspects of each inscription without having to travel to museum storerooms. In addition to the catalogue of inscriptions, she published a compendium “Analysis” volume, CM I, where in 2012 she analyzed the materiality of Cypro-Minoan inscriptions and contextualized Cypro-Minoan writing practices within the island and outside of it.

Ferrara’s CM II is also innovative for doing away with Masson’s subdivision, a decision which she explains in CM I. In other ways, however, it is overly faithful to HoChyMin and replicates many of its omissions and mistakes. While CM II adds over thirty new inscriptions, around fifty inscriptions remain unpublished or uncollected. Single-sign texts were again excluded, but on different grounds. For Ferrara, it is impossible to judge whether a specific single-sign text represents a writing or marking system.

4.2.1 The Precollapse Decades: Cypro-Minoan Writers at Ugarit

Most mercantile texts date to the LCIIC and LCIIIA periods on both sides of the “Bronze Age collapse.” It is fascinating to wonder whether the traders active in the LCIIC could feel the shifting tides in their overseas networks, or if the traders at Ugarit could sense their city’s impending destruction. Testimonials in letters from Ugarit speak to a level of anxiety, decreased grain supply, hunger in outlying regions, and attacks from the “people of the sea.” Yet other documents record the activities of elite merchants traveling and conducting business unencumbered. One of these merchants, a man named Yabninu, who was also a high administrator (šatammu rabu) in the royal bureaucracy of Ugarit, had such close ties to Cyprus that both mercantile and nonmercantile Cypro-Minoan document forms were found in his home (for a great social history of writing at Ugarit, which puts Yabninu in context, see Boyes, Reference Boyes2021). The ease of mobility implied by Yabninu’s cache of Cypro-Minoan texts speaks to traders’ continued engagement in overseas exchange on the precipice of collapse and reveals the closeness of the networks established by traders.

Yabninu was a senior official in the kingdom of Ugarit who also worked as a merchant and (perhaps) envoy at the end of the 13th century. The more elite the merchant at Ugarit, the freer he was to conduct personal trade in addition to official, royal business (McGeough, Reference McGeough, Eder and Pruzsinszky2015). Yabninu’s house contained ample evidence for close personal and official relationships between himself and people outside of Ugarit. The house contains fine imports from throughout the eastern Mediterranean, including Cyprus, and just under 100 tablets recording the administration of foreign residents at Ugarit and evincing close mercantile ties with Cyprus, Phoenicia, Palestine, and Egypt. Among the tablets, most written in syllabic cuneiform, were two written in Cypro-Minoan. More indicative of close ties to Cyprus than the tablets, perhaps, are the array of mercantile texts scattered throughout the house, single-sign and multisign texts on a mix of local and imported vessels (see Figure 18). Not only do we know through cuneiform texts that Yabninu traded with Cyprus and knew Cypriot merchants, but his house contained a set of vessel texts that would have been right at home on Cyprus.

Figure 18 House of Yabninu with locations of Cypro-Minoan texts indicated.

Drawn by author based on Boyes, Reference Boyes2021.

Indications that traders from Ugarit, as opposed to persons from another walk of life, were likely involved in the transmission of Cypro-Minoan to Ugarit come from the large numbers of single-sign vessel texts on local and imported vessels found throughout the city and its port, Minet el-Beida (Hirschfeld, Reference Hirschfeld, Balensi, Monchambert and Celka2004). Of less clear significance are the findspots of the Cypro-Minoan inscriptions, most of which were found in the homes of traders, since most inscriptions of any kind at Ugarit come from the homes of elite traders. The vessel texts on imported and local vessels indicate the involvement of traders in the production of Cypro-Minoan texts at Ugarit, not simply the importation of Cypro-Minoan texts from Cyprus.

The local adaptation of Cypro-Minoan writing at Ugarit is further evident in the creative adaptation of Cypro-Minoan document forms and the invention of new ones. For instance, four of the twelve Cypro-Minoan texts from Ugarit are tablets, “reflect[ing] experimental use” (Steele, Reference Steele2019, p. 206). The tablets are not only different in document form from the Cypro-Minoan tablets from Enkomi, but they are also different from one another, each its own unique creation, displaying different formatting, punctuation, and techniques and styles of writing. The creativity on display in the Cypro-Minoan texts from Ugarit is reminiscent of the creativity displayed by writers from Ugarit more generally, who invented their own form of alphabetic cuneiform script called “Ugaritic.” Experimentation bridges Cypro-Minoan and alphabetic cuneiform writing at Ugarit, making it hard to say who writers of Cypro-Minoan were. Were they merchant writers of Cypro-Minoan who set up shop at Ugarit, where they were free to reimagine their script? Merchant writers of Ugaritic, excited to learn a new script and experiment with it as they had done with their own? Our evidence is too meager to provide answers, but it raises questions about how writers at Ugarit perceived the relationship between script and identity.

We do not know if, or to what extent, script was seen as an expression of ethnic, political, or national identity in the Late Bronze Age Mediterranean, since we do not know how ethnicity or nationality were constructed or expressed. The choice to use Cypro-Minoan at Ugarit could reflect an expression of mercantile identity, not an ethnic or national one. In experimenting with the Cypro-Minoan script at Ugarit, merchants may have been asserting their identity as traders, and their identity as traders might have superseded or otherwise competed with other modes of identity like ethnicity.

4.2.2 Mercantile Institutions

How representative Yabninu was of his Cypriot counterparts in terms of mobility and a connection to “state-sponsored” interests is difficult to reconstruct. Cypro-Minoan mercantile texts contain some 80+ unique signs that may refer to an equal number of merchants or mercantile organizations (Hirschfeld, Reference Hirschfeld and Smith2002). Are the 80+ signs evidence for 80 individuals as mobile and interconnected as Yabninu was? If so, on whose behalf were they working? Analogies from Ugarit are useful to a limited extent. Mercantile activity at Ugarit was driven by a mix of private and royal interests, with more autonomy to conduct personal trade given to higher-ranked persons (Routledge and McGeough, Reference Routledge, McGeough, Bachhuber and Roberts2009). Exactly how many individuals at a given time were engaged in trade activities or how they were organized is only hinted at in the literary record. Texts from Ugarit refer to aširuma, a word whose translation is debated but which seems to refer to classes or groups of people, perhaps strongly associated with trade activity, under a leader (Monroe, Reference Monroe2009, pp. 123–125). Imagining Cypriot traders divided into corporate bodies of traders each led by a “head” is appealing. It would certainly make sense of the large but fixed groups of signs that appear on mercantile texts over the hundred-year plus period spanning the LCIIC into the LCIIIA. How such groups would obtain access to goods and on whose behalf remains to be understood, given how little is understood about the economic and political organization of Late Bronze Age Cyprus.

Traders organized into groups, conducting trade at least partially on their own behalves and with counterparts similarly arranged overseas, would have been able to transmit knowledge of writing intergenerationally and between different groups. Writing could have been taught “on the job,” so to speak. A learner could have observed how to format texts properly through a mix of direct instruction and incidentally, as traders wrote on objects that entered into wide circulation. Close relationships between traders from different communities, as testified to at the house of Yabninu, would have occasioned the exchange of script knowledge, at least sporadically.

4.2.3 The Evidence from the Sea

The evidence from the sea indicates that the Cypro-Minoan script would have been visible to and used by sailors. The two most famous shipwrecks from the Bronze Age Aegean, Uluburun and Gelidonya, carry different document forms. Arrested midvoyage carrying a cargo crammed with prestige goods and raw materials of royal caliber, the mid 14th-century Uluburun wreck contained over thirty single-sign vessel texts on its transport amphorae, local Cypriot jugs, and other vessels (Pulak and Matheny, Reference Pulak and Matheny2021). In contrast, the late 13th-century Gelidonya wreck, carrying a more quotidian cargo perhaps belonging to a traveling tinsmith, has fewer single-sign vessel texts but an intriguing group of ten copper tools, several already cut for scrap, bearing single-sign texts and a two-sign inscription (see Figure 19). The over 1,000 kilograms of total copper on the ship is believed to represent scrap metal intended for sale or remelting in port. Sailors may not have been aware of the marks and writing on their goods. Some of the texts, however, were centered on tools that had already been cut for scrap, making it likely that traders were responsible for incising signs onto the tool after they had been cut.

Figure 19 Incised multisign text on a socketed pruning hook from the Cape Gelidonya shipwreck, B99.

Photograph by John Littlefield, courtesy of the Institute of Nautical Archaeology.

Clear evidence that traders wrote Cypro-Minoan signs on traded metals is seen in the ingots made from non-Cypriot metals, which have been recovered from ships sunk either en route or in harbor off the coast of modern-day Israel. The cargo of the Hishuley Carmel ship mentioned in Section 4.1, for instance, comprises a copper tool with a two-sign Cypro-Minoan inscription and thirteen tin ingots, five with two-sign inscriptions, the rest with single-sign texts bearing Cypro-Minoan script signs. Texts with Cypro-Minoan script signs are present also on the lead ingots from a newly investigated wreck near modern-day Caesarea (Yahalom-Mack et al., Reference Yahalom-Mack, Finn, Erel, Tirosh, Galili and Yasur-Landau2022), though in this case all are single-sign texts. Neither lead nor tin is indigenous to Cyprus. Traders with knowledge of the Cypro-Minoan script would almost certainly have been intermediaries, ferrying the ingots from their original destinations to Cyprus or elsewhere. As already discussed, the sources of tin in the Late Bronze Age are still debated. Anatolia, Afghanistan, and Cornwall are all plausible sources. The lead of the ingots carried in the Caesarea wreck has been chemically traced to Sardinia. In these cases, the application of Cypro-Minoan texts would almost certainly have been made by middlemen involved in the movement of metals from the edges of the Mediterranean into its interior.

4.2.4 Taprammi and His Bowl

Mobility, interconnectivity, and knowledge of script is evidenced among elites from many different parts of the eastern Mediterranean, not only at Ugarit and Cyprus. The individual relationships developed between traders could have positioned them to survive the collapse of palatial centers with wealth and relationships intact. One reason why Cyprus may have survived the collapse is because of the strong mercantile networks created by its traders that gave them access to goods and traders independent of or at least not whole reliant upon royal networks. An example of a person embedded in such networks is Yabninu. Another similar figure can be found in the person of Taprammi, a Hittite official who traveled widely throughout the eastern Mediterranean and wrote in at least two different scripts (see Figure 20).

Figure 20 Tracing Taprammi’s travels. The map shows places where inscriptions or references to Taprammi have been found. His exact travel routes are not known.

Taprammi’s voyages, tracked through inscriptions he wrote and literary references to him, provide a means through which to imagine the movements of Cypro-Minoan writers (Hawkins, Reference Hawkins and Mellink1993). An official in the Hittite court, Taprammi wore several official titles, including “Scribe,” “Pithos Man,” and “Eunuch” (no one does official titles like the Hittites). Like Yabninu, he seems to have also conducted personal business in addition to carrying out his royal duties. According to a tablet from Ugarit recorded in the syllabic cuneiform script, he was negotiating with the king of Karkemish in northern Syria over a personal business deal involving enslaved persons. Taprammi also left a proverbial paper trail in Anatolia. A stone monument inscribed in the Luwian hieroglyphic script in the capital Hattuša is a dedication from Taprammi dated to the reign of the Hittite king Tudhaliya IV (c. 1245–1215 BCE), and Taprammi’s seal is on a cuneiform document from the same site.

Another of Taprammi’s texts, a bronze bowl also written in Luwian hieroglyphic, reveals him building a network of elites. The inscribed bowl is made of bronze, the appropriate metal for gifts exchanged between persons of equal rank (stone was for gods and silver for kings; Simon, Reference Simon2018). Befitting bronze, the bowl’s inscription is addressed to a scribe of equal rank to himself. The bowl is particularly intriguing for its similarities to inscribed bronze bowls from Cyprus that may even be considered the same document form despite being inscribed in a different script. The formatting of the text on the bowl just below the lip is the same as the inscribed bowls from Cyprus. The biggest difference between the Cypro-Minoan bowls and Taprammi’s gift is that his inscription is on a richly embossed bowl carrying complex pictographic imagery, whereas the Cypro-Minoan bowls are simple and undecorated. If they are the same document form, then the function of the Cypro-Minoan bowls can be inferred from Taprammi’s: Gifts given between elites, perhaps ones similarly well traveled and interconnected with overseas mercantile activity as Taprammi himself.

5.1.1 Exceptional Case 1: The Clay Cylinder Document Form

Two document forms fall in neither the mercantile nor insular camp: clay cylinders and the Enkomi tablets (Masson’s CM 2). What distinguishes these two document forms from the rest is their relatively high degree of standardized material features. In the case of the Enkomi tablets discussed in Section 3, their near-uniform paleography suggests that they were possibly produced in scribal schools. The clay cylinders, for their part, are widely agreed to be administrative document forms.

There are six examples of clay cylinders from two different sites, five examples from Kalavaso-Ayios Dhimitrios (K-AD) and one from Enkomi. The five Kalavaso documents are all contemporary roughly to one another, dating to the LCIIC, found together in Building X, where administrative activity is believed to have taken place. The Enkomi clay cylinder was found in an unclear context and is possibly contemporary to the K-AD cylinders. The clay cylinders are widely regarded as administrative because some contain numerals and lists and were palimpsests (i.e., erased and reused), and because of the administrative setting of K-AD Building X (Smith, Reference Smith and Smith2002).

The clay cylinder writing medium consists of nonperforated, nonhollow clay cylinders approximately 2–5 cm in length and 1.5–4 cm in diameter. Writing on the cylinders runs left to right lengthwise across the long surface of the cylinder, though beyond this similarity there is some variation in formatting among the texts. At least one cylinder, ##101 K-AD Arou 004, contains numerals, and another, Enkomi cylinder ##097, has “many features” suggesting it may “represent a list” (Ferrara Reference Ferrara2012, p. 121). In combination with the administrative findspot of the K-AD cylinders, the presence of numerals and lists on the documents supports an administrative interpretation of the document form. Joanna Smith’s observation that two of the cylinders are palimpsests, erased and reused for the same purpose, also supports their administrative interpretation (Smith, Reference Smith and Smith2002). Narrowing the focus to K-AD, the clay cylinder document form may be the product of a network of writers who learned how to write in administrative contexts at K-AD. The paleography of the cylinders is distinctive vis à vis both other inscriptions found at the site and inscriptions from elsewhere. The internal paleographic uniformity of the K-AD cylinders suggests that their writers received training in how to construct and write the clay cylinder document form while “on the job” there.

5.1.2 Exceptional Case 2: The Enkomi Clay Tablet Document Form

Limited evidence for the presence of scribal schools on Cyprus comes in the form of three tablets from Enkomi dating most likely to the early 12th century, which Émilia Masson believed were written in a separate subscript CM 2. The uniformity in paleography of the tablets is unique among Cypro-Minoan document forms and includes not just uniformity in sign shapes but also in the ductus and height of the signs. Sign height on all three tablets is around 0.3 cm, with the signs adhering to a strict imaginary rule line. The ductus on all three tablets comprises impressed and impressed-and-drawn signs rendered with a round-tipped stylus, which often created a distinctively Cypro-Minoan “tear drop” ductus.

Martina Polig’s 3D models have shown that the paleography of the Enkomi tablets shares features with the inscriptions on other document forms (Polig and Donnelly, Reference Polig and Donnelly2022, p. 49). A similar combination of drawn and impressed elements, for instance, is used in the construction of sign many forms. What is unique to the tablets, however, is that they have the same paleography across all three tablets. In the rest of the insular documents, and certainly in the mercantile ones, sign shapes within a given document form are varied. The uniformity in sign shape and ductus on the Enkomi tablets would seem to indicate that their writers were following the same rules of representation. It suggests they received the same instruction on how to draw sign shapes, what implement to use to draw them, and how to construct and abide by imaginary rule lines. The obedience displayed by the tablet writers is likely the result of the meticulous instruction and sustained practice occasioned by a scribal school curriculum.

5.1.3 Perishable Materials?

The possibility of a scribal school on Cyprus raises the question of why so few document forms seem to have been produced in scribal schools. The six clay cylinders and three Enkomi clay tablets are the exception and not the rule. Their uniform features suggest, in the first case, training in administrative uses of writing and, in the second case, training in a scribal curriculum. If there were administrative and/or scribal uses of writing on Cyprus, in addition to the mercantile and insular uses, why are there so few examples of them?

One possibility is that clay was not the main writing medium used on Cyprus but that most script was written on perishable materials. Roughly contemporary examples of perishable writing materials include inked wooden writing boards in Egypt, animal skins for Aramaic (Radner, Reference Radner and Renger2011) and wax-coated wooden writing boards for cuneiform (Cammarosano et al., Reference Cammarosano, Weirauch, Jendritzki and Kohl2019). Wax-coated wooden boards survive in the eastern Mediterranean in only two examples from the Uluburun shipwreck (à la Figure 21). Use of wax-coated writing boards has been proposed as the possible material of choice for Cyprus, since heavy use of wax could also explain another notable absence from the Cypriot record, that of clay sealings (Smith, Reference Smith and Smith2002). Cylinder seals made from precious and semiprecious stones, not to be confused with the clay cylinders discussed in Section 5.1.1, were used throughout the Aegean and the Ancient Near East in administrative contexts, impressed into lumps of clay, “sealings,” to indicate a seal holder’s ownership, authorization, and/or oversight.

Figure 21 Panels of an ivory writing board dating to the 8th century BCE from Nimrud, Mesopotamia. Wooden boards were more common than ivory.

Courtesy of Metropolitan Museum, 54.117.12a, b.

If wax was used on Cyprus as a writing medium and sealing surface, then the near-total absence of administrative and scribal Cypro-Minoan document forms could be illusory. It still would not explain, however, why Cyprus used wax for writing and sealing when its neighbors to the east and west preferred clay, or why Cyprus had so many distinctive insular and mercantile document forms. Nor is there reason to believe, given the distinctiveness of Cyprus’s surviving writing media, that writers of Cypro-Minoan would have restricted their uses of wax to the anticipated administrative and scribal document forms. Everything we have learned about writers of Cypro-Minoan so far suggests that, if they wrote on wax, they would have written on it in a myriad of ways, displaying the same creativity evident in most Cypro-Minoan document forms.

5.2.1 Cylinder Seals (Not to Be Confused with Clay Cylinders …)

Cylinder seals made from precious stones are one of the few Cypro-Minoan document forms also used as a writing medium in other regional scripts, including syllabic cuneiform, Luwian hieroglyphs, and Linear B, but as is often the case with Cypro-Minoan, the document form is reimagined in the hands of Cypro-Minoan writers. The Cypro-Minoan texts are all extremely brief, never more than one word and often only two signs long, and their formatting is wildly inconsistent (see Figure 22). In comparison, syllabic cuneiform texts on cylinder seals found throughout the Near East often record full phrases, usually expressing ownership or obeisance to a deity. The texts are usually original to the seal’s composition, placed in panels beside or around the seal’s main images. Luwian hieroglyphic seals used by Hittite officials, which include not only cylinder seals but also seals on convex discs, have inscriptions that spread and coil over the entire surface of the seal face. While the content of the syllabic cuneiform and Luwian hieroglyphic texts is not explicitly administrative, the function of cylinder seals in these contexts decidedly is. Likewise, the Linear B texts on cylinder seals, clay seals, and sealings all pertain to the administration of palatial resources, the types of texts that are remarkably absent from Cypriot contexts. In the Aegean, script rarely appears on seals cylinder but is prevalent on sealings, lumps of clay which, when still wet, were impressed with a seal, incised with an administrative text, and then attached to the object being distributed or otherwise administered.

A. ##195 CYPR? Psce 001, h. 2.0 cm;

B. ##199 ENKO Psce 001, h. 3.05 cm, two-part inscription. Both parts of the inscription have unusual, unidentified signs that may be decorative and not script signs.

Photographs and drawings by author. Autopsy courtesy of the British Museum.

Figure 22 Two cylinder seals with different paleography and formatting.

The exact content of Cypro-Minoan texts on cylinder seals is necessarily unknown, but it is highly unlikely that the texts were meant to convey administrative information. Although some scholars attribute an administrative function to cylinder seals on Cyprus, on analogy with neighboring cylinder seal traditions, the Cypro-Minoan document form is not analogous to neighboring ones, and therefore text content is also likely to be different (Donald, Reference Donald2016). The texts on Cypro-Minoan seals are often incredibly hard to read, whether on the surface of the seal or rolled out and impressed onto a surface. It is therefore reasonable to surmise that the mere presence of writing was the message, not its content. In other words, the writing was both decorative and semiotic in the same way that seal motifs were (Aruz, Reference Aruz, Aruz, Graff and Rakic2013). The writing, like the seal images, would convey meaning to the seal’s wearer and to viewers of the seal which was separable from and possibly even superseded the meaning of the phonetic word or abbreviations recorded on the seal.

In favor of the decorative interpretation is the utter lack of order or pattern discernible in the placement, orientation, and paleography of the texts. The chaotic formatting indicates a disregard for the potential reader’s ability to anticipate where the text on a sealing would be and often resulted in texts that are illegible except under close inspection. The placement of texts on cylinder seals is almost as varied as their imagery. Some texts are arranged horizontally, some vertically, and still other texts are discontinuous, squeezed in between the elements of the image. Some texts are upright on the surface, others upside down; some are legible on the surface of the seal, others only on the sealing surface once the seal was rolled out. Still others might never have been clearly legible given the scatter of signs across the seal’s surface. Paleography and ductus are likewise variable. Some texts have a linear ductus. In other cases, the seal carver seems to mimic the shape of signs impressed into wet clay, carving out thick and voluptuous rounded signs (see Figure 22).

Given the willy-nilly features of Cypro-Minoan cylinder seal texts, it might be right to interpret the texts as saying, “here I am, look at me!” and not “name, rank, and title,” more embellishment, less administrative ID. The discovery of many inscribed cylinder seals in tomb contexts also lends support to this interpretation of the texts as adornment rather than information. The texts were certainly not the administrative aids they were on neighboring cylinder seals.

5.2.2 Clay Balls

Small clay balls, about 2 cm in diameter, are perhaps the most distinctive and well-known Cypro-Minoan document form, capturing the imagination of scholars and enthusiasts alike with their simple shape and small size (see Figure 23). In a certain sense, the small size and spherical shape of the balls makes them unsuitable to carry texts. Their small size makes them the type of object that a person could easily lose, and the circle shape makes the texts hard to read in a single glance. Often, a ball must be turned between the fingers for the whole text to become visible (Steele, Reference Steele2014). Yet despite the apparent unsuitability of the clay balls as a writing medium, they are one of the most common Cypro-Minoan document forms, and they seem to have been especially important at the site of Enkomi.

Figure 23 Collection of clay balls of different sizes. A. ##088 ENKO Abou 020 (d. 1.9 cm); B. ##024 ENKO Abou 029 (damaged); C. ##021 ENKO Abou 018 (d. 1.8–2 cm); D. ##023 ENKO Abou 020 (d. 1.3 cm); E. ##022 ENKO Abou 019 (d. 2.3 cm).

Photograph and drawings by author. Autopsy courtesy of the British Museum.

Of the ninety-two recorded clay balls, eighty-seven come from the site of Enkomi, two from Kition, two from Hala Sultan Tekke, and one from Tiryns, Greece. The concentration of balls at Enkomi testifies to their insular character, the peculiarity of their use to Cyprus. At the same time, the lone clay ball from Tiryns raises the possibility that even the most insular of Cypro-Minoan document forms still traveled. In addition, there are two clay balls from Ugarit written in the Ugaritic script, a clear example of a Cypro-Minoan document form adapted to write another script.

Exactly what the balls were used for has attracted much speculation. Suggestions range from weights to gaming marbles to workers’ ID cards to lots, inferred variously from the shape and size of the balls, their findspots, the nature of their texts, or a combination of all three (for an overview of theories, Steele, Reference Steele2014). It is generally assumed that all balls had the same function, that is, they are the same document form, because of their consistent material features. The size of the balls is loosely standardized between 1.7 and 2.3 cm in diameter, the texts are always disposed on the ball’s central axis, and the ball findspots often are associated with metallurgical production, ritual activity, or both. The standardization of the writing material and formatting is contrasted by their stunning paleographic variation, which includes not only a wide variety of sign shapes and drawing and impression techniques but also writing implements (Smith, Reference Smith, Stampolides and Karageorghis2003). Ball writers were given much leeway when putting stylus to clay as long as wrote along the central axis.

Among the various suggestions concerning the function of the balls, lots used in sortition as, suggested by Silvia Ferrara and Miguel Valério (Reference Ferrara and Valério2017) best, accommodates the evidence. Sortition, a method of resolving questions or making decisions through the casting of lots, was practiced throughout the contemporary Ancient Near East. It is amply attested in the Ancient Near Eastern literary record, but not the archaeological one. In nearby Hittite Anatolia, sortition was used to assign individuals priestly roles for specific festivals or for calendar years (Taggar-Cohen, Reference Taggar-Cohen2002). Much later, in the Classical and Hellenistic periods, the use of lots to select jurors, assign plots of land, and resolve political disputes is well attested in the literary and, to a lesser extent, archaeological record.

If the lot interpretation of the balls is correct, then the balls inhabit a space somewhere between administrative and nonadministrative documents, an ambiguity reflected in their material features. The balls are standardized in size, shape, and formatting, but the paleography of the texts is not standardized. Similar to administrative tools, the balls might have been used to assign individuals to priestly roles or plots of land but in a manner which by its very nature invites chance. Like much of Cypro-Minoan, the balls and their function elude straightforward characterization.

5.2.3 Reading the Balls Abroad

There is a tentative consensus that the ball texts contain names, derived in large part from comparison to the two balls from Ugarit that record two local Ugaritian West Semitic names in written the Ugarit script (Ferrara, Reference Ferrara2012, pp. 111–112). Here, decipherment and analysis of document form converge. Because the Ugaritic balls are undoubtedly modeled on the Cypro-Minoan document form and because the Ugaritic script is deciphered, it is possible to infer the content of the undeciphered Cypro-Minoan ball texts from them: They most likely record names. Our ability to read, Ugaritic paired with the concept of document form allows, us to “read” the content of the Cypro-Minoan balls even as the sounds recorded by the signs in the ball inscriptions remain unknown. The balls, like the insular document forms, are at once characteristic of Cyprus but also show the reach of Cypro-Minoan writers outside of the island.

5.2.4 Metal Bowls

The inscribed bowl document form briefly mentioned in Section 4, like the balls, consists of texts inscribed onto relatively small, mobile objects, but they are much less common than the balls (see Figure 24). There are nine Cypro-Minoan inscriptions on metal bowls, seven made from bronze or copper and two from silver (the exact composition of the metals awaits confirmation by testing). Like the balls, the bowls are a largely insular document form that nevertheless hints at their writers’ mobility. All bowls but one, a silver bowl found in a ritual deposit in the acropolis of Ugarit, were found on Cyprus. The bowl from Ugarit, a Cypro-Minoan document form found abroad, has a counterpart on Cyprus in a silver bowl found at Hala Sultan Tekke (Yon, Reference Yon, Calvet, Salles, Yon, Caubet and Fourrier2004). The Hala Sultan Tekke bowl was inscribed in an alphabetic cuneiform script similar to Ugaritic, sometimes called the “short” or “reduced” cuneiform alphabet, fashioned like the Cypro-Minoan document form. It should be regarded as another example of a Cypro-Minoan document form adopted by writers of another script, even if the layout of the text is slightly lower down than on the Cypro-Minoan inscriptions, a difference perhaps attributable to decoration on the bowl’s lip (see Figure 24).

A. ##179 CYPR Mvas 002.

Photograph and drawing by author.

B. ##180 CYPR Mvas 003.

Photograph and drawing by author, courtesy of the Cyprus Museum and the Department of Antiquities, Cyprus.

Figure 24 Two examples of the bowl insular document form.

The precise social function of Cypriot is difficult to determine from archaeological evidence alone since most bowls were not recovered during archaeological excavation but looted and sold on the art market. A pattern is discernible in the small number of inscribed bowls whose findspots are recorded: Bronze bowls come from tombs, and silver bowls from settlements. Written records from the Ancient Near East record bowls made from various metals being given in diplomatic exchange. As mentioned in Section 4, it may have been the case that different metals designated different levels of exchange: Silver bowls were given to kings, and bronze bowls were exchanged as gifts between officials. It is not clear whether the Near Eastern model applies to the Cypriot one, but the differentiation of the findspots of bowls made from different metals could support a distinction in their function. Comparative evidence from the Ancient Near East further suggests that bowls associated with the elite ritualized consumption of liquids, an activity equally associated with settlement and funerary contexts (Feldman, Reference Feldman2014, pp. 111–137).

The bowl texts, like the balls, show standardization in their formatting but not in their paleography. The standard placement of the texts is, evocatively, just below the bowl’s lip. Imagining the bowls were drunk from like a cup, as iconographic evidence suggests, the bowl holder’s lips may have touched or covered the inscription as they drank. Alternately, the bowl holder could have drunk from the opposite side, positioning the text in clear view of their companions or the audience of a ritual performance. Unlike the fixed position of the texts, the paleography and structure of the texts vary from bowl to bowl. Some texts have highly linearized signs, others, like the silver bowl from Enkomi and the alphabetic cuneiform bowl from Hala Sultan Tekke, have signs drawn in the “tear drop” ductus characteristic of Cypro-Minoan. The size and legibility of the signs also differs, as does the apparent methods used to make the inscriptions. Variety is also apparent in the structure of the texts: two texts record one sign sequence, five record multiple sign sequences separated by word dividers, and two record multiple sequences, including sign sequences and numerals, separated by word dividers.

The texts are believed to record names, which could support their interpretation as gifts exchanged between individuals. The belief the texts record names derives from several threads of evidence. First, at least four bowls record short sign sequences that end with the sign CM 023, which, as discussed in Section 3, is generally interpreted as a genitive or dative suffix. Second, later Cypro-Syllabic inscriptions on metal bowls, a document form possibly inherited from Cypro-Minoan, record names. If the texts are the same document form, then it is possible to “read” the Bronze Age inscriptions in light of the Iron Age ones. Third, contemporary metal bowls inscribed in other scripts record names of individuals, some of whom, like Taprammi whom we met in Section 4, bear the official title of scribe. Writers of Cypro-Minoan could have been similar highly mobile individuals.

5.3.1 Continuity and the Metal Bowl Document Form

Unlike those who wrote scripts that perished along with their administrative centers, writers of Cypro-Minoan continued to write far past the collapse period and into the Early Iron Age. Bowls are the only Cypro-Minoan document form apparently adopted wholesale in script transfer from Cypro-Minoan into Cypro-Syllabic. The longevity of the document form and its continued relevance even as the political organization of Cyprus underwent dramatic changes in the transition from the Bronze into the Iron Age speaks either to the survival of the institutional settings in which bowl texts were produced and shared or to the artificial replication of an old practice. The evidence is ambiguous.

A bronze bowl from Palaepaphos is the among the latest dated Cypro-Minoan inscriptions, found in a tomb dating from the mid 11th to 10th centuries (Egetmeyer, Reference Egetmeyer, Karageorghis and Raptou2016). The date of the tomb is not necessarily the date of the bowl or the inscription, nor is there any definite way of determining the bowl or inscription date. Its excavators note that the bowl might have been an heirloom, passed down from generation to generation until it was deposited, in which case the bowl would be older than its context. Heirloom or not, the bowl exhibits strong ties with the Bronze Age. Its text, comprising a 5-sign sequence, a numeral, and a 1 + 1 sequence, records the same 5-sign sequence inscribed onto two bowls with secure Late Bronze Age dates (##179 and ##183). It is also similar to another Bronze Age bowl inscription in recording a numeral (##182).

The bowl exemplifies debates over when Cypro-Minoan ends and Cypro-Syllabic begins. At least part of the ambiguity comes from the small number of both Cypro-Minoan and Cypro-Syllabic inscriptions from the 11th–9th centuries. There is no question that by the 8th century, the Cypro-Syllabic script consists of some fifty-five signs, likely a much smaller number than was used in Cypro-Minoan, and that signs shared between the two scripts change shapes in Cypro-Syllabic. It is also clear that Cypro-Syllabic, by this point, is being used to record the Greek language. Many of the inscriptions dating to the 11th–8th centuries, including the Palaepaphos bowl, record signs whose shapes are intermediary, not exact matches for Cypro-Minoan or later Cypro-Syllabic forms (see, for instance, the debate around the opheltas obelos, as summarized in Duhoux, Reference Duhoux2012). The bowl inscription would be considered such an ambiguous text except for the affinities of its other features to the Bronze Age document form.

When the Palaepaphos bowl was written and by whom raises intriguing questions about continuity in writing practice between the Late Bronze and Early Iron Ages. Were the writers of bowl document forms participating in the same institutional settings from one period to the next? If so, it may have been the same one that mobile, mercantile elites like Taprammi navigated just before the collapse, but reconfigured, severed from its ties to imperial administration and reliant only on relationships between individuals. One piece of evidence in support of this suggestion is the earliest alphabetic inscription found in the Aegean, an inscribed bronze bowl from Knossos, Crete.

5.4.1 The Knossos Tekke Bowl

The bronze Tekke bowl from Knossos on Crete, the center of Minoan and Mycenaean Greek administration on Crete until its destruction in the 14th century, is the oldest evidence for alphabetic writing found in the Aegean (see Figure 25). It was found in a tomb dating to the 9th century, but as with the Palaepaphos bowl some scholars have suggested it was an heirloom of an earlier date. The Phoenician alphabetic inscription, positioned just underneath the bowl’s lip, records a short text, “the bowl of X, son of X …, ” which has been interpreted to be either a declaration of ownership or a dedication (Hoffman, Reference Hoffman1997, pp. 120–124). Its material features broadly fit the Cypro-Minoan document form, with a slight difference in the placement of the text, which is further below the lip than in the Cypro-Minoan examples. The difference in layout is possibly attributable to the fact that the lip contains other decoration, similar to the Hala Sultan Tekke bowl. Even if the layout were identical to the Cypro-Minoan examples, it would be hard to claim that the Knossos bowl was directly inspired by them. Other Phoenician inscribed bowls date to the 10th century, and the bowl is a relatively popular document form in the Iron Age Mediterranean (Feldman, Reference Feldman2014), often found in tombs like their Cypro-Minoan counterparts. It may well be the case that the Cypro-Minoan document form is the genesis for the inscribed bowls that become popular in the Iron Age, but further studies would be necessary to show this is definitely the case.

A. Knossos Tekke bowl with Phoenician inscription.

B. Hala Sultan Tekke bowl with short cuneiform alphabetic inscription.

Drawing by author based on Matthäus, Reference Matthäus1985.

Figure 25 Metal bowl document forms written in different scripts.

The Knossos Tekke bowl (named after the location of the tomb) is one piece of evidence for renewed overseas trade conducted between the Aegean, Cyprus, and Phoenicia after a postcollapse lull (Bourogiannis, Reference Bourogiannis2018). According to some scholars, the bowl is of a Cypriot type and must have been produced in Cyprus. Others say that its features could just as well indicate production in the Levant. In either scenario, the bowl is the product of renewed trade between individuals going between the “Phoenician” Levant, Cyprus, and Crete in its use of a document form long-standing on Cyprus and with a possibly contemporary correlate in the Palaepaphos bowl.

The Tekke bowl embodies long-standing associations between the metal bowl document form and mobile individuals involved in overseas trade and “could suggest multi-way influences in the practice of writing between speakers of local languages and speakers of Greek and Phoenician” (Steele, Reference Steele2019, p. 88). It could well be that the Phoenician alphabet spread in the same type of institutional contexts as Cypro-Minoan itself: networks of overseas traders in which Cyprus played the role of linchpin. On analogy with Cypro-Minoan, the Phoenician alphabet must have been taught and transmitted outside of scribal and centrally administered contexts. Just like with Cypro-Minoan, the script may have also been transmitted in scribal and administrative contexts, but the institutional contexts that facilitated and encouraged the spread of the alphabet throughout the Mediterranean were mercantile.