Ancient writing systems in the Mediterranean

A critical guide to electronic resources

Celtic, gaulish

- 4th c. B.C. - 1st c. A.D.

Examples of writing

Gallo-greek : Inscription on stone (Vaison-la-Romaine)

Gallo-greek : Inscription on stone (Vaison-la-Romaine)
Source: RIG I, G-153.

The Gallo-Greek writing on this stone from Vaison-la-Romaine is very neat:





ειωρου βηλη-

σαμι σοσιν


"Segomaros, Villū's son, citizen of Nîmes, gave to Belesama this sacred enclosure" (translation P.-Y. Lambert).

It is clearly a dedicatory gift, confirmed by the Gaulish verbal form ειωρου. This offering is dedicated to Belesama (in dative), goddess identified as the Gaulish Minerva. The dedicant states Nîmes as his provenance (ναμαυσατις is an ethnic adjective, derived in -ati from Nemausus, the Latin name for Nîmes). τοουτιουϲ seems to be a derivation of the Celtic noun touta "tribe, clan", here translated as "citizen".

The gift is a νεμητον which seems to be a wood or a sacred enclosure (σοσιν is its demonstrative).

The person dedicating this νεμητον is mentioned in the first two lines. His name is Segomaros (name composed with sego- = "victory, strengh" and -maros = "great") followed by a patronymic element in the genitive case.

(Example proposed by Francesca Ciurli)

Gallo-greek : Inscription on stone (Nîmes)

Gallo-greek : Inscription on stone (Nîmes)
Source: RIG I, G-203.

Doric capital dedicated to the Mères Nîmoises discovered in 1740.

[.]αρταρ/ος ι/λλανουιακος δεδε

ματρεβο ναμαυσικαβο βρατουδε[

" (?)artaros Illanus' son gave (it) to the Mères Nîmoises, (?) in recognition, with the tithe, for the realization of the vow" (translation: P.-Y. Lambert).

This inscription is on a pedestal bearing a statue (presently lost).

The writing is regular and meticulous. Other similar inscriptions were found in Gard (near Uzès, Saint-Gilles and Beaucaire).

The dedicant's name is uncertain (the inscription is not complete on the left) but the patronymic component is present.

The verb of dedication δεδε is in the preterit (root: *di, "to offer, to give"). The last term has stimulated many contributions and hypotheses. It is here shortened, but the complete formula should read: βρατουδε[καντεμ. The composition of this syntagm (bratou-dekantem / bratoude-kantem) is still uncertain. 

The entire set (statue + inscription) was dedicated to divinities, mentioned in the dative case (ending in -bo): ματρεβο ναμαυσικαβο, where ναμαυσικαβο is an epiclesis (derived with an -iko- suffixation) derived from the Gaulish name for Nîmes. 

(Example proposed by Francesca Ciurli)

Gallo-latin : Inscription on stone (Alise-Sainte-Reine)

Gallo-latin : Inscription on stone (Alise-Sainte-Reine)
Source: RIG II 1 L-13.

This monumental inscription is one of the few inscriptions written on stone in the Gallo-Latin epigraphic corpus. The stone block (49 x 74 x 13 cm) was found on Mont Auxois in 1839.







"Martialis, Dannotalos'son, offered this building to Ucuetis, and this in accordance with the blacksmiths who honour Ucuetis in Alesia"  (translation: P.-Y. Lambert).

After the dedicant's name, there is the same verb as on the Roman inscription from Vaison-la-Romaine IEVRV, this time using the Latin alphabet.

The theonym is in the dative case VCVETE (*i stem). It may be, according to P.-Y. Lambert, an agent name in -ti, formed on a denominative derivation of the stem *okuo- "sharp, acute" (cf. Latin: acuus) "the sharpener". 

The gift is a CELICNON (here in the accusative case), in all probability a place for banquets. The demonstrative SOSIN is also in the Vaison inscriptions. The building for the god Ucuetis was also found in Alise-Sainte-Reine, at the beginning of the 20th century, after the discovery of this inscription. It is a construction on two levels (perhaps a warehouse and an assembly hall). This place of worship received in particular metallic offerings, which explains the mention of blacksmiths in the inscription.

A second part of the inscription was introduced by ETIC (eti cf. Latin et, Greek ἔτι and the enclitic particle -kwe with apocope).

gobedbi, the term for blacksmith, is in the instrumental case, in the plural form in - bi.  dugiIontiIo  would be a verbal form in the 3rd person plural, whose stem is not yet clearly identified.

Lastly, the locative AlisiIa made this inscription famous because it was the first reference found to the well-known city of Alesia, place of Vercergetorix's final battle and the final conflict of the Gaulish Wars.
(Example proposed by Francesca Ciurli)

Gallo-Latin: magical inscription on lead-plaque (L'Hospitalet-du-Larzac)

Gallo-Latin: magical inscription on lead-plaque (L'Hospitalet-du-Larzac)
Source: La langue gauloise, 1994 (2003²), p. 162-165.

This lead-plaque (RIG II.2, L-98) is the longest document in Gaulish known so far. Discovered in 1983, it is one of a group of "magical" documents containing defixiones; these curse tablets were very common in the Gaulish magical tradition. 

The document is composed of two tablets found one upon another, on the top of a funeral urn. It contains around 60 lines and 170 words (or fragmentary words).

It was written by two distinct hands. The first text apparently re-enacted a witch battle. The second use of the tablet seemed to have sought to lessen the magic of the former; the second hand also seems to have been less Latinized than the first one.

P.-Y. Lambert's translation:

1a, 1-7: “Send this women's charm against the names here below; this (is) a witch charm bewitching witches. O Adsagsona, look twice Severa Tertionicna, their thread (that which ties?) witch and their writing (that which writes?) witch, let it release the one whom they will have struck with a curse; with a bad spell against their names, make the bewitchment against this group [+ a dozen feminine names].”

1b, 6-7: “let these women mentioned here above, once bewitched, be powerless in front of him.”

2a, 3-10: “Any man acting as a judge, that these women would have struck with a curse, let her (Severa Tertionicna) delete this man's curse; let there not be any witch by writing, witch by thread, witch by giving among these women, who solicit Servera witch by writing, witch by thread, the foreigner.”

Gallo-Latin: Calendar from Coligny (Ain)

Gallo-Latin: Calendar from Coligny (Ain)

This bronze plaque is conserved in the Gallo-Roman Museum in Lyon-Fourvière. It was found in 1897 in Ain.

It is a Gaulish calendar written using the Latin alphabet in the 2nd c. AD. This exemplar is most probably a late copy of an older calendar.

It counts five consecutive years. Months are lunar and months of 30 days (MAT "good") alternate with months of 29 days (ANMAT "bad"). The lunar calendar is spread over the solar year, which is why this calendar has leap months (one at the beginning of the first year and one in the middle of the third year).

Written in 16 columns, it has 4 months per column. There are no phrases (perhaps only one in the heading of the first leap month).

The names of the months can be read as follows (with many graphic variants):













Gallo-Latin: Potter's accounts (La Graufesenque)

Gallo-Latin: Potter's accounts (La Graufesenque)
Source: Musée Fenaille

The archaeological site of La Graufesenque (Millau, Aveyron, France) is exceptional for Gaulish epigraphy. At the beginning of the 20th century, thousands of terra sigillata ceramic shards coming from potters' kilns were found.

For economic reasons, kilns were shared between many potters, who could in this way bake their products together. An “accounts list” was also put into the kiln during baking; this list, which recapitulated the distinct lots placed in the kiln, was a simple clay plate baked with the rest of the batch.

These documents are very useful because they provide a large amount of anthroponymic elements as well as numeric data about how workshops operated. On the other hand, the linguistic material is very poor, with only rare phrases and few common nouns. Most of these latter concern bookkeeping and are technical terms referring to ceramics works.

Gallo-Etruscan : bilingual Inscription from Vercelli

Gallo-Etruscan : bilingual Inscription from Vercelli
Source: RIG II 1, E-2. Musée Camillo Leone (Vercelli)

This bilingual inscription from the 2nd c. BC is written on a schist stone, discovered in 1960 in Piedmont (height: 149 cm; width: 70 cm).

It is a boundary stone which delimitated, with three other stones now lost, the land of a man named Acisius "Argantomaterecus".

One text is written in Latin, using the Latin alphabet and a second one is written in Gaulish, using the Gallo-Etruscan alphabet.

The Latin text is as follows:

Finis / campo . quem / dedit . Acisius / argantomater / ecus . communem / deis . et . hominib / us . ita . uti . lapides / II II . statuti sunt

Translation: P.-Y. Lambert: "End of the land that Acisius Arganto-materecus gave to be in common amongst gods and men, within the area where the four stones were erected."

Gaulish text is as follows:

aKisios. arKaToKo{K} / maTereKos . To[ś]o / KoT[e (.) a]Tom/ś Teuoχ / Toni[o]n EV

Translation: P.-Y. Lambert: "Acisios Argantomaterekos (gave) the ATOS that belongs to gods and men."

The owner's name (in nominative, Akisios) is followed by a disputed term, with multiple components. P.-Y. Lambert proposes the following construction: preposition kom- “with” +  ater-, “father” + suffix -eko-.  This syntagm -atereko- would be equivalent to the Latin patricius “patrician”, according to P.-Y. Lambert. Following the same idea, the sequence komatereko- should be compared to the expression “patres conscripti”. Arganto-komatereko would then indicate the fee requested by the treasurer (the Roman Quaestor).
The Latin text is more specific and seems more detailed than the Gaulish one.

The word ATOŠ or ATOM from the Gaulish text cannot be translated. It is commented in the Latin version of the text by the word campus. What seems to be the verbal form in the Gaulish sentence, TO[.]O / KOT[..] could be read TOŠOKOTE , the preterit of the verb "to give", perhaps with an infix -so-.

The adjective TEUOXTONION, “common to gods and men” is a compound form which M. Lejeune analysed as "divine" (*deiwo- ) and “terrestrial, mortal” (*gdhon-io-).

The last two letters could be the abbreviation for EX VOTO: in this area Romanization was already very advanced, as the necessity of a bilingual inscription indicates.

(Example proposed by Francesca Ciurli)

Gallo-Etruscan: bilingual Inscription from Briona

Gallo-Etruscan: bilingual Inscription from Briona
Source: RIG II 1, E-1.

This irregular stone (140 x 95 cm) was found in 1859 10 km northeast of Briona. The inscription, in the so-called "Lugano alphabet" can be dated to the 2nd-1st c. BC.

It features four wheels drawn vertically. The principal text (9 lines) contains almost exclusively personal names. A secondary text (a) is carved on the superior part of the stone, in the same direction as the principal text.

Lastly, a second text (b) can be read vertically, near the wheels. The order of reading is controversial.

a) ]n[.]k[..]esasoioikan[










b) takos·toutas·[

" a) ?" "Dannotalos'sons, Quintos, legate, Andocombogios, Setubogios, and Essandecot(t)os'(sons), Andareuiseos, Dannotalos, have raised this cairn" "b) ? decision of the tribe" (translation: P.-Y. Lambert).

TANOTALIKNOI (which cannot be separated from TANOTALOS) shows a patronymic suffixation -ikno-. It is followed by three names in the nominative: KUITOS = Quintos, borrowing from Latin, as well as the attributive adjective in Latin LEKATOS = legatus.

The anthroponyms ANOKOPOKIOS and SETUPOKIOS, are Celtic composed names in -bogios = "who strikes".

A second section is composed of ESANEKOTI (sons), in the genitive case: they are ANNAREUISEOS and TANOTALOS.

The final verb KARNITUS , derived from *karno "cairn", is the only indication for the function of the inscription, which is the dedication of a cairn.

The (b) text is also disputed. Two noteworthy terms are TAKOS (perhaps meaning "order, decision", in parallel with Old Welsh and Oscan) and the word TOUTAS , which is the name for a Gaulish political unit.

For a recent article and a new reading of this inscription see Estarán 2015.

(Example proposed by Francesca Ciurli)