Ancient writing systems in the Mediterranean

A critical guide to electronic resources


- 6th-2nd century BC

edited by: Simona Marchesini     DOI: 10.25429/   (translation revised by Melanie Rockenhaus)
Last update: 12/2022

  • Introduction
  • Index
  • Further information

Block of “leccese stone” from Ceglie (3rd century BC)

Introduction and definition
The word “Messapic” refers to the language and alphabet of the pre-Roman Augustan II region Apulia et Calabria, corresponding to modern Puglia. The epigraphs range chronologically from the middle of the 6th century to the end of the 2nd century B.C. In a narrow sense, the term refers only to the language and alphabet attested in the Salentine peninsula. It is not yet clear if the spoken language was the same in the whole of Puglia, from the Gargano in the north to Leuca in the South, because of the scarcity of the inscriptions from the northern Daunia and Peucetia, in northern Apulia (moreover coming from a later period, i.e. the 4th century B.C. on). The incomplete alphabet coming from the Daunia also presents some variance in the letter sets compared to the Salentine alphabet, and is more similar to Hellenistic alphabetic types.

The whole collection of Messapic inscriptions (Monumenta Linguae Messapicae = MLM), published in 2002, registered 545 epigraphic items. Forty-eight other texts in the MLM have been considered doubtful (dubiae vel alienae), written in mixed or scarcely identifiable writing. In these cases, the analysis of the letters did not lead to a precise attribution. There is also a group of about 30 coins.

Since 2002, almost a hundred new inscriptions have been unveiled and partly published. Some of them are in print. They were discovered in Vaste, Castrum Minervae, and Ostuni. In particular, the inscriptions recently found in Vaste (Mastronuzzi, Ciuchini 2011), Castro (D'Andria, Lombardo 2009; D'Andria 2021) and Ostuni (Poccetti 2008) have overturned the statistics of attestations compared to 2002.

Most of the inscriptions come from the sacred area of the Grotta di Santa Maria di Agnano in Ostuni, followed by the sanctuary of Athena in Castro, Roca (in particular Grotta Poesia with few inscriptions published), then Lecce, Alezio, Oria, Mesagne, Ceglie Messapica, Rudiae, Gnathia, Valesio and Vaste. 

Go to the online resources.

Online resources



The chronology of Messapic epigraphy has been confirmed by a re-analysis of all alphabetic types, their revision and attribution to distinct phases, by means of computer processing (MLM 2002; Marchesini 2009).

A first result is the definition of the lower chronological limit of epigraphic data, not later than the 2nd century B.C. The end of the production of written texts seems to be linked to the wars against Hannibal, which brought profound cultural and political transformations in the region. We assist after this time to a progressive blending of the Italiote scenario, which finds its climax in the lex Iulia Municipalis of 45 B.C., with a complete Romanization of Apulia.

Seven epigraphic phases have been identified, inclusive of two transition periods. In each phase, the vanishing or the strong reduction of several letter types is to be found, along with the contemporary introduction of new types, which become common in the later phase. Concerning the modifications of the alphabet set, some boundaries can be set between the phases. Then each phase can be dated thanks to some chronological markers, i.e., inscriptions which can be dated with certainty, independent from the epigraphic information itself and in a more objective way, thanks to archaeological elements.

The epigraphic periods are divided and dated as follows:


I (experimentation) mid 6th-first half 5th c. BCE
II (transition) second half of the 5th c. BCE
III (stabilisation) 4th c. BCE
IV (transition) 4th-3rd c. BCE
V (archaic graphemes disappear) 3rd c. BCE
VI (stabilisation) 3rd c. BCE
VII (calligraphic phase) second half 3rd-end 2nd c BCE

Alphabetical models

The nearest and most direct model for the Messapic Alphabet is the Laconic from Taranto. Nevertheless, in the first epigraphic phase (Phase I), represented by the most ancient inscriptions from Veretum (Patù, near Lecce), Oria (near Brindisi) and Cavallino (also near Lecce), some letters seem to be suggested from other models such as Corinthian or Euboic. In many cases it is not possible here to distinguish the local from the Greek inscriptions.

Some inscribed fragments from the Monte Papalucio sanctuary near Oria, or some from Grotta Porcinara by Veretum, which seem to have been produced locally, show Greek letters. This fact can probably be explained by the presence of resident Greek artisans, who lived and worked together with the local population and adopted some epichoric cultural aspects.

In the indigenous texts, certain features such as either the uncertainty of the ductus, sometimes right-handed, sometimes retrograde, with cases of boustrophedon writing, or the poor connotation of the letters used, are to be interpreted as typical of a phase of alphabetic experimentation.


The oldest inscriptions attested in Messapia (at Veretum, Oria and Cavallino) have a religious-cultural connotation and come from urban or proto-urban contexts. This suggests that one of the most probable causes for the introduction of writing in the Messapic world was the need to enshrine in some way, with concrete acts of ethnic awareness, the spaces of an urban reality, as observed for instance in Cavallino. On the other hand, at Veretum and Uria the cultural contexts themselves offer some of the oldest written documents. The content of the texts, where identifiable, seems substantially referable to the cultic-votive sphere but also there is no lack of personal names among the earliest epigraphic attestations, particularly from contexts that we can define as funerary-celebratory, such as those of the inscribed cippus or semata.

Text classes

Most of the Messapic epigraphic corpus belongs in a funerary context, as often is the case in the epigraphic tradition of Pre-Roman Italy. A second, smaller group belongs to the religious and cultic sphere. There are also inscribed stones from public buildings: even if numerically scarce, these architectonic elements offer the longest texts. Among the public inscriptions, there are also some cippi, stelae or columns with the names of the remembered or praised people. Coins are considered an independent text class.

Unfortunately, a substantial number (almost 19% of the total) of inscriptions have been lost since the first partial publication or cannot be attributed to any of the categories listed above. The other best-represented classes, that of loom weights and that of fine pottery, both tableware and votive, fit with difficulty into a well-defined sphere: for fine pottery, it is often difficult to identify the destination sphere, which fluctuates between votive-religious and personal. There are few inscriptions on pottery from tombs, for example, which can be observed in many other Italic contexts (I am thinking of Etruscan), a fact that can be traced back to the particular funerary customs of the Messapi, who, especially in the Archaic period, preferred to mark the burial, albeit rarely, using anepigraphic semata, placed outside the tomb. Lastly, loom weights, a numerically conspicuous class, mostly fall into commercial inscriptions, like trademarks, although there is no lack of specimens engraved with divine names (Marchesini 1995).

According to our documentation, it can be stated that the context most commonly associated with writing is the funerary context. It must be pointed out, however, that the numerical preponderance of grave inscriptions may also be due to the fact that for many years, tombs were the main object of research in the area of the Messapic settlements and that our data may therefore be partially distorted by selective archaeological exploration. In recent times, sanctuary inscriptions, with the cases of Ostuni and Castro, are increasingly highlighting the class of fine tableware pottery used for dedications to divinity.

Alphabetic features

With respect to the Laconic, Tarentine model – a 'red' alphabet in Kirchhoff's classification –the following adjustments are to be observed/recorded:


- Loss of the phi sign, replaced in most cases with <p> (as in Aprodita).


- Introduction of a trident with an angular basis (simbolo) and of its variants (simbolo, simbolo), probably to represent a phoneme /h/. The grapheme had two different values: /kh/ in western Greek alphabets and /ps/in the eastern ones. In archaic Messapic inscriptions it may have been used instead of the grapheme <h>. The most plausible hypothesis is that it is the graphic rendering of a fricative sound determined by the vowel+j sequence. This is suggested by attestations such as Haivasimboloias Zaras (MLM 12 Bal, first half 6th - first half 5th century BC) and Aviθas Balesimboloias Zaras (MLM 15 Bal, bronze disc, same chronology). The synchronic use of the two signs can perhaps be explained by the fact that, while h (<*s) possessed an aspiration stroke, the transition sound between the vowel and the sonorant /j/ was a fricative sound that was attempted to be rendered with a special grapheme.


The graphemes <simbolo> and <H> must have seemed so close in phonological correspondence that one was excluded by redundancy (the trident sign). In recent inscriptions, the simbolo sign (with variants) was omitted, and the grapheme <H> was generalised: in the 3rd century we find Haivahias (MLM 3 Car);


- Introduction of the cross sign <+> (also in the variant of S. Andrew’s cross: <X>) to realise a palatal sibilant /š/, while in the western Greek alphabets the grapheme represented the phoneme /ks/


- Use of a sign theta (Θ, in Laconic = /s/) for the palatalised dental plosive: Blatθes < *Blat-yo-s.


- Addition of a trident with a square basis simbolo. The letter, already present in the earliest inscriptions (Phase I), peaked between the 4th and 2nd century B.C., only to be abandoned in the 2nd/1st century B.C. Its probable phonological value is uncertain, as it alternates with T;


- Use of a grapheme <z>, alternating with <+>, as the phoneme /z/ is rendered as /s/ in contact with /t/: Dazet, (PN m.) Daštas.

Ductus e divisio verborum

The Messapic inscriptions are generally written from left to right. Some archaic cases show a 'boustrophedon' script: the text is written without interruption, thus changing the direction of the script only in the new line. Cases of "false boustrophedon" are also known, where each new line is written after turning the stone by 180 degrees: in this way the writing always runs in the same direction.

The Messapic script is continuous (scriptio continua), without interruption between the words. In some cases, however, a blank space is left between words. Many inscriptions show the trend to let a word end coincide with the end of the line.

Bibliographical references

For an in-depth analysis of the literary sources on the Messapians, we can consider the founding contributions of M. Lombardo (1991 and 1992) and J.-L. Lamboley 1996. The volumes of BTCG can be useful to get information on single sites. The latest overview is offered in Marchesini 2021.

Information on the statistical grouping of epigraphic classes can be found in the introduction to the sites in the MLM, vol I. Marchesini 2020 offers an update on the alphabetic situation. The first edition of Messapic texts was undertaken by F. Ribezzo. His contribution, the Corpus Inscriptionum Messapicarum (CIM) was published in several issues of the journal Rivista Italo-Greca-Italica di Filologia, Lingua, Antichità (RIGI).

An update to the CIM was later offered by the same Ribezzo with the name Nuove Ricerche per il Corpus Inscriptionum Messapicarum (NRCIM) in 1994. At the same time, J. Whatmough edited in the Prae-Italic Dialects of Italy the part devoted to the Messapic epigraphy. In 1965 O. Parlangèli started publishing Nuove Iscrizioni Messapiche (NIM), which, more than a corpus, offers a linguistic comment to the Messapic words he identified. He used mostly drawings, adding photos only rarely. In 1964 C. de Simone published all the Messapic inscriptions known at that time in the second volume of Hans Krahe's Die Sprache der Illyries (de Simone 1964), entitled Die Messapiche Inschriften. He offered a chronology of all the inscriptions based on the associations of letter types.

Many new Messapic texts unveiled after Parlangeli's edition have been edited in three volumes by C. Santoro Nuovi Studi Messapici (Santoro 1976, 1982 and 1984). In the Monumenta Inscriptionum Raeticarum (MLM, 2002) all known inscriptions have been published with chronological framing after seriation and all identified lemmas were gathered in a separate volume (vol II). An update of new findings is in Marchesini 1999, 2015 and 2020.