Silvia Ferrara, ‘Writing in Cypro-Minoan: Beyond decipherment’

Paper presented at British Epigraphy Society Autumn Meeting. (Brief report by Gabriel Bodard)

Writing in Cypro-Minoan: Beyond Decipherment.

Silvia Ferrara, Oxford,  November 21, 2009

In this paper, Ferrara introduced the audience to problems in the decipherment and interpretation of inscriptions in the Cypro-Minoan script. Since there are only some 217 documents (comprising 4000 signs) in this script, decipherment is difficult if not impossible, and so identification of the language and context of the texts will depend more upon quantitative elements, the objects themselves, their distribution and other archaeological information.


The traditional classification of Cypro-Minoan by Emilia Masson identifies four versions of the script/language, but this classification is not reflected in the art, style of objects, or geographical features. The majority of the texts are in Masson’s Cypro-Minoan 1 (CM1) script, with a wide geographical distibution and range of supports (including many tiny clay boules); this is assumed to be the original form of the language. The corpus of CM2 is made up entirely of three tablets, all from the city of Enkomi and of similar date. CM3 is attested in nine inscriptions from the mainland city of Ugarit. Both Enkomi and Ugarit also give CM1 texts–and the CM1 texts from Ugarit do not seem, as sometimes suggested, to have been imported from Cyrpus since their form is similar to that of Cuneiform tablets from the mainland. The differences, identified by Masson, were based on the assumption that some characters were unique in each script, but since none of the scripts are identified, this is a difficult and to some degree subjective identification.

In 2007 Jean-Pierre Olivier published a revised classification of the CM1 script, in which he pointed out that out of 96 distinct signs, CM1 and CM2 share only 42; there are 19 unique signs in CM1, 15 unique signs in CM2 (a quarter of the total signs in that script) and 7 unique signs in CM3 (12% of the total signs in that script). Unique signs mostly occur only once in each script, raising the possibility that in at least some cases these are the result of scribal incompetence rather than deliberate distinction. Nevertheless, if these very signficant differences are meaningful, then the introduction of new signs in each script must represent sounds or other real distinctions not recorded in the other subgroups. Leaving aside the incompletely attested CM3, one may either assume that CM1 and CM2 are different languages, or that the subject matter between the two collections of texts was sufficiently diverse that the vocabulary would differ significantly between the two.

Ferrara considered four possibile interpretations of the classification of Cypro-Minoan from these texts: (1) the texts represent a single script and a single language, the language spoken throughout the island of Cyprus; the apparent differences between the scripts makes this interpretation problematic. (2) The texts represent a single script used to represent multiple languages; again the script variants are not explained. (3) The texts represent multiple scripts, all used to represent a single language; this is Ferrara’s prefered interpretation. The representation of new sounds in a script does not have to mean the introduction of new sounds to the language, as they may simply be sounds that were not previously recorded in the written form of the language, or new complex syllables such as CVC groups. (4) The texts represent three separate scripts used to record three different languages. This possibility explains the palaeography well, but is problematic archaeologically. Is one of the languages a form of Cypriot koine? Were the writers of CM2 at Enkomi perhaps not newcomers but an established population with their own language? Ferrara discussed the issues of multilingualism to examine this final possibility a little further. Looking for parallels to the use of multiple languages and scripts in a single geographical and chronological context (as at Enkomi), she looked at the case of Cretan Hieroglyphic and Linear A, but the archaeological contexts of these two scripts were fairly distinct and so they may have been different ways to record a single language.

A better comparison may be the case of the written archives at Ugarit, where seven languages are recorded in five scripts. The Akkadian lingua franca is written in Cuneiform using a syllabic system. Hurrian also uses Cuneiform script, but with an alphabetic system; the syllabic values of Cuneiform are intertwined with these alphabetic values with no apparent difficulty. In this parallel, biliteracy within a single language functions perfectly well.

The linguistic diversity of Cyprus is unknown. If CM2 was derived from CM1 to record a new language, there is no archaeological record of the people whose language this was (but nor is there for Hurrian). Given the limited literacy of the general population at this period, the idea that there would be two separate administrations in a single city, sometimes in identical contexts, using mutally unintelligible scripts and languages, is bizarre and counter-intuitive.

Ferrara concluded that there is no secure reason to consider Cypro-Minoan to be made up of more than one language. She ended her paper with an appeal to epigraphers to make a special effort to integrate contextual and archaeological features in the interpretation of text-bearing objects, along with the epigraphic staples of palaeography and philology.

This entry was posted in BES, events, report. Bookmark the permalink.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.