ARCHAEOLINGUA

Edited by
ERZSÉBET JEREM and WOLFGANG MEID

Series Minor
1
The Author examining the Larzac tablet
Millau Museum, June 1, 1992
 WOLFGANG MEID

GAULISH INSCRIPTIONS
Their interpretation in the light of archaeological
evidence and their value as a source of linguistic
and sociological information

BUDAPEST 2014
 Third edition,
revised and enlarged 2014

The Cover photograph shows the Menhir of Vieux-Poitiers,

with Gaulish inscription (see p. 36)
Ratin briuatiom Frontu Tarbetisonios ieuru

ISBN 978-963-9911-61-1
HU-ISSN 1216-6847

© The Author and Archaeolingua Foundation

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system,
or transmitted in any form or by any means, electronic, mechanical, digitised, photocopying,
recording or otherwise without the prior permission of the publisher.

2014
ARCHAEOLINGUA ALAPÍTVÁNY
H-1014 Budapest, Úri u. 49.
Word processing by the author
Desktop editing and layout by Rita Kovács
Printed by Prime Rate Kft., Budapest
 Preface

This short survey of Gaulish epigraphical “literature” was first published in 1992,
then reprinted with minor revisions in 1994. It has been out of print for many years
since, but being still in demand, a renewed edition was called for. Considering
the progress made in the field of Gaulish epigraphics in the meantime, this new
edition has been updated to account for more recent scholarship and also has
been extended in order to deal with, among others, the one major inscription
discovered in 1997 at Châteaubleau, which has received considerable attention
in the meantime.
Among the dubious or pseudo-Gaulish inscriptions stands out the one
discovered as early as 1887 at Rom (Deux-Sèvres), believed to be the longest
of Gaulish inscriptions known at that time, which had received various fanciful
interpretations since, Gaulish as well as Latin, and of which even the latest
treatment in vol. II,2 of the grand Receuil des Inscriptions Gauloises is at a loss
to extract some meaning. I have added my opinion, for whatever it is worth, of
this strange document which I consider a sort of “love story” in a slave milieu,
composed in an extraordinary linguistic medley of rather low-grade Latin, Greek
and Gaulish.

Innsbruck, November 2014 Wolfgang Meid

 Gaulish Inscriptions

From historical and archaeological records and from the distribution of place-
names we know, or at least can infer, that the ancient Celts, at the height of their
expansion in the latter half of the first millennium BC, occupied, or controlled,
vast territories in South Central, Western and Eastern Europe. They were present
in southern Germany, France and Belgium, the British Isles, Spain and Portugal,
Italy, Bohemia, Austria and Hungary; they invaded the Balkans, Greece and
Minor Asia where some of them settled permanently – the Galatians of the New
Testament were of Celtic descent (Γáëάτáé being one of the general names under
which the Celts were known in antiquity1). Compared to this huge expansion
of what we assume were Celtic peoples speaking varieties of an ancient Celtic
language, the actual linguistic remains of that era, documented in the form of
texts, are very meagre indeed.

Fig. 1. Prehistoric expansion of the Celts and historic migrations since

the 6th century BC. Map adapted from R. and V.S. Megaw, Celtic Art, p. 11.

1
Êåëôïß. -áß / Celtae, Gallī being the others.
8

The Celts, being heirs to a culture which depended on oral tradition, were not
given to writing and adopted this habit comparatively late in the course of their
cultural contacts with Greeks, Romans and other peoples of the Mediterranean
world. Since, after the Roman conquest of Gaul and of other territories inhabited
by Celts, Latin became the normal medium of public communications, and since,
under the impact of Latin, Celtic speech declined rapidly, there was neither much
need nor much opportunity for writing in the native tongue.
All the same, the surviving records of Ancient Celtic, although scanty
compared with the mass of Latin or other inscriptions, are by no means negligible.
Apart from the names and glosses transmitted in other contexts the inscriptions
constitute the only direct linguistic evidence for that period. At the same time they
supply valuable sociological and cultural information. In the last few decades
many new inscriptions have been discovered, some of them longer texts which
seem to indicate that the practice of writing in the vernacular was perhaps not
quite so unusual after all. It is worthwhile then to survey this material and to ask
ourselves what information – linguistic or otherwise – it yields.
Inscriptions in Celtic speech have been forthcoming mainly from three areas:
(a) from the north-central part of the Iberian Peninsula – “Celtiberian” with
mostly very brief texts, but also some longer and two rather long inscriptions,
both found at Botorrita (the one in 1970, the other – much longer – in 1992;2
(b) from the district of the lake of Lugano – “Lepontic” with about 40 brief
or very brief texts;
(c) from Gaul – both Cisalpine and especially Transalpine Gaul.
These areas also represent dialect areas with distinct linguistic features.
The Celtiberian area is particularly complex and archaic; it represents an
archaic q-Celtic area. Lepontic is p-Celtic, and so is Gaulish, apart from a few
archaisms3 which still show reflexes of the original Indo-European labiovelar.

2
The adjoining Lusitanian inscriptions have also been taken to represent a Celtic dialect,
so by J. Untermann, but a different Indo-European idiom by most other scholars, and
quite recently the so-called “Tartessian” inscriptions in the Southwest of the Peninsula,
hitherto considered non-Celtic or even non-Indo-European, have also been claimed
for Celtic, so by John T. Koch (who even believes this area to represent the nucleus
of Celtic expansion), but this matter is still sub iudice. Cf. John T. Koch, Tartessian.
Celtic in the South-west at the Dawn of History. Second edition, revised & expanded.
Aberystwyth: Centre for Advanced Welsh and Celtic Studies. National Library of
Wales. 2013.
3
Such as Equos, name of a month in the Coligny Calendar (see fn. 124), as against
*epos in ordinary Gaulish.
 9

Names with preserved Indo-European *p can also be found in Gaul which raises
similar questions as in Spain. I am inclined to see in them traces of other Indo-
European dialects in sub-, ad- or peristratic relation to the Celtic idioms which
in turn became the dominating language of these areas.4 While Lepontic may be
considered an archaic side-dialect of Gaulish, Celtiberian appears to be a rather
different Celtic language altogether.
In what follows I shall concentrate on Gaulish. Gaulish inscriptions are by no
means a uniform corpus of documents. They span several centuries, are attested
in three geographically distinct areas and comprise diverse subject matter. By
being written in three different alphabets they exhibit cultural affinities to the
Italic, Greek and Roman worlds.
Italy presents three inscriptions of importance (Todi, Briona, Vercelli); they
are in the North Italic alphabet which does not distinguish between voiceless and
voiced occlusives; two of these inscriptions are bilingual. Gallia Narbonensis
presents about 60 inscriptions in the Greek alphabet, most of them funerary and
votive inscriptions. Eastern and Central Gaul have more than 100 inscriptions
in the Latin alphabet which, of course, are chronologically later than the bulk
of the Gallo-Greek and the Gallo-Italic inscriptions. The great pottery centres
of Lezoux and La Graufesenque in Aquitania present us with a great number
of graffiti; for the greater part they are in Latin or meant to be so, but many
contain Gaulish words or even short sentences. The La Graufesenque graffiti5,
for instance, apart from some technical vocabulary6, give us the Gaulish ordinal
numerals from one to ten.7
4
This has parallels in the Germanic linguistic territory where (particularly in the
Northwest) we also find traces of an integrated substratum language with retention
of Indo-European *p and other stops not subjected to the Germanic sound-shift. See
W. Meid, “Hans Kuhns ‘Nordwestblock’-Hypothese. Zur Problematik der ‘Völker
zwischen Germanen und Kelten’.” In: Germanenprobleme in heutiger Sicht (ed. H.
Beck), Berlin 1986, pp. 183–212.
5
Masterly new edition by Robert Marichal, Les graffites de la Graufesenque, Éditions
du CNRS, Paris 1988. Regrettably the Gaulish sentences are not part of this edition,
but they have been edited in vol. II, fasc. 2 of RIG published in 2002.
6
Gaulish words attested are tuq(q)os (tuððos) “filling (of the furnace)” (always with
ordinal numeral: first, second etc., up to ten), luχtos (gen.) “charge (of the furnace)”,
autagis “detachment, (sub)division”, uχsedia (equivalent to Latin summa) “grand
total”. The names of the various vessels, though, are in (Vulgar) Latin.
7
cintuχ, allos, tr[ (= tritios, cf. Welsh trydydd), petuar(ios), pinpetos, sueχos, seχtametos,
oχtumetos, namet(os), decametos. These formations are, for the greater part, Common
Celtic (cf., for instance, seχtametos = Old Irish sechtmad, Welsh seithfed). It may be
10

Fig. 2. Graffito from La Graufesenque; the first words are tuqos suexos
“sixth charge (of the furnace)”. Musée de Millau.

The above numbers mean inscriptions of any substance. To determine the

total number of Gaulish inscriptions is quite impossible because many are
fragmentary, consisting only of a few letters, others consist only of a single word
or name; therefore one cannot be always sure if they were meant to be in the
noted, however, that on a recently discovered lead plate fron Rezé (estuary of the
Loire) recording commercial transactions, some alternative ordinals of more archaic
formation seem to be attested, such as [pi]χte, representing *piχto- < *kw§kw-to- (cf.
Latin qu¥ntus) as against pinpetos in La Graufesenque (= Old Irish cóiced, Middle
Welsh pymhet). Cf. P.-Y. Lambert – D. Stifter, Études Celtiques 38 (2012), 146–150.
 11

Gaulish language (the hypothetical context might have been Greek or Latin).
There are other inscriptions which cannot be classified with certainty because
their status is unclear.8 Others appear in a sort of mixed language or show code-
switching. Such texts are of course particularly interesting because they testify
to the decline of Gaulish and point to hybrid forms of speech as an intermediate
stage before the total disappearance of Gaulish.
The normal practice in dealing with inscriptions is to list them regionally
according to place of provenance. While this approach is all right for purely
descriptive and cataloguing purposes and for regional statistics, it neglects the
relevant semantic and pragmatic features of the texts themselves which do not come
out sufficiently well by this method. The cataloguing of inscriptions according
to place of provenance should be complemented, therefore, by a classification
according to semantic and pragmatic criteria and their corresponding linguistic
expression. Relevant criteria are, e.g., carrier of the inscription (stone, metal,
other materials; type of the object) and type of script used, character (public,
monumental, private, intimate, magic, religious), function (funerary, votive,
dedicatory), indication of manufacture or ownership, commercial, etc.), special
intention (charms and incantations, texts of playful nature or erotic significance,
etc.). The various types of inscriptions usually have linguistic properties in
common which emerge clearly only through such comparison; these typological
similarities may facilitate the interpretation of difficult or obscure texts. The
temporal and local distribution of types and features tends to show certain

8
This applies to a rather large body of apparent votive inscriptions found in what is
believed to have been an ancient sanctuary near present-day Glozel, about 70 km
south of Vichy. At least part of them are genuine, written (like Lepontic) in a sort
of North-Italic alphabet, dating from the 3rd century BC until the Roman era. Their
reading however, and still less their interpretation, cannot be taken as assured, and
therefore they are left out of consideration here, especially as falsifications must have
been fabricated in later times, for souvenir or other purposes. As far as they can be
deciphered they show similarities to Gaulish, but the forms of the names or lexemes
are strangely syn- or apocopated.
Still, this material would deserve serious investigation, but so far no celticist of repute
has taken the pains to deal with it. It was left to a Swiss amateur researcher, Hans-
Rudolf Hitz, to make the elucidation of this dubious matter his life interest. His last
two major publications, with ample bibliography, are (published privately): Ein Corpus
der altkeltischen Inschriften von Glozel (2009), and Die Hintergründe der Inschriften
‘im heiligen Hain von Glozel’ nemu Chlausei. Von den lepontisch-etruskischen zu den
gallisch-keltischen Einflüssen (2011). See also my note in Kratylos 43 (1998), 26.
12

patterns, which usefully links this approach with the purely geographic listing,
giving it profile and accentuation.
Concentrating on those inscriptions which have a certain body of text and
which can tell us something, linguistically or otherwise, we may divide them into
two major groups:
(a) lapidary inscriptions (i.e. cut in stone),
(b) inscriptions on other carriers.
The first group is, by its very nature (the durability of the material) monumental
in character: for the most part, it consists of funerary or votive inscriptions.
The second group consists of inscriptions on objects of everyday use –
pottery and other household implements, toilet articles etc. –, a wide range of
objects such as the editors of the Corpus Inscriptionum Latinarum would call
the Instrumentum domesticum. Inscriptions on such objects are, as a rule, of a
personal, private nature; in many cases they simply indicate the owner or the
maker of the object. In some cases we get votive inscriptions, too, if the object
in question was given, as a votive gift, to a god (or to a number of divinities
in common). Pottery and metal objects are the main carriers; wood, cloth and
other possible materials, being subject to decay, have not survived as bearers of
inscriptions.
A special group is represented by little lead plates carrying inscriptions
(usually in a Latin cursive script) which clearly served a magical purpose. Several
of these have been found in recent years, and – being longer texts – they are the
most important addition to our knowlege of Gaulish, although they are also the
texts which present the greatest difficulties. The Chamalières text found in 1971
contains about 60 words, the inscription of Larzac found in 1983 has about 160
words and is thus the longest continuous Gaulish text which we now possess,
the most recent addition, the Châteaubleau tile discovered in 1997 with a text of
intimate character has 60 or more words (word division often being uncertain).
The problem is that being confronted now with longer texts we realize that
we still know very little of Gaulish. Up to the seventies of the last century we
had to do only with rather brief texts of a stereotyped nature – short sentences the
meaning of which one could more or less correctly guess. Being able to understand
such texts and to identify their limited grammar and vocabulary one could deceive
oneself into believing that Gaulish was an easy thing, at least as far as the evidence
went. The emergence of longer texts, with a different subject matter, has brought
home to us that Gaulish is a vast field of which we previously had just caught
a glimpse and that our knowledge acquired so far is quite insufficient when we
 13

have to deal with such new matter.9 In order to overcome such difficulties and to
progress towards an understanding of such texts it will be necessary to develop,
and apply more subtle techniques in dealing with them, making use not only of
the analytical, philological and comparative methods which are at the linguist’s
disposal but taking advantage also of all extralinguistic factors which may shed
light on the semiotics and semantics of these texts. Semantics: the question of what
it is all about is the main issue and the key to the matter, previously known and
identifiable linguistic elements serving only to guide our steps and prevent us from
errors. The methodological crux in the deciphering of such texts is that although we
may be able to identify certain words and linguistic forms, they may not be vital to
the issue, they may be peripheral, and any theories built upon peripheral elements
while ignoring the essential ones are almost certainly doomed to failure. To have a
fair chance of success it will be necessary to find out the central arguments of the
text and to interpret them correctly; the rest will then fall into place. We are still
quite a distance away from such success, but good progress has been made with
the Chamalières inscription which is perhaps 80 % clear whereas with the Larzac
inscription this rate is considerably lower, which means that this text is only partly
understood and, in some essential parts, not understood at all. Most of what one can
see is that in Larzac women play a prominent part, two groups of women magicians
(mnas brictas) apparently fighting each other with magical means. This, in itself,
however, is interesting enough.
But let me not be tempted to start at the difficult end; let us rather start from
the easy end, and let me try to present a sort of typological approach to Gaulish
inscriptions, by which common as well as distinguishing features will come out
more clearly.
The most simple type of a meaningful inscription is that which – on any
material – consists of just one word. In most cases this will be a personal name.
Thus, the objects of the instrumentum domesticum will, often enough, carry their
possessor’s name (in the same way as we have our names inscribed on personal
objects). Thus one finds names like
Máñïò, Åβïõñïò, Ëïõãïõò, Ìáãá, Ìáãåóéëëá, Êñéîéá,
Êáìïõëá, Äïííéáò
on objects of pottery. They are not high-sounding compound names of noble people
but uncompounded names or extensions of simple word stems; these names, we
9
The insular Celtic languages are of no great help; they are attested much later, have
changed considerably in the meantime, and the subject matter of their texts is quite
different.
14

can infer, are, for the greater part,

the names of ordinary people; Måros
means “great”, Maga the same,
Kriksia (Welsh crych) means “curly”.
Donniås ist the genitive of an å-stem
which may be masculine or feminine
(donna “brown”). Perhaps Eburos
“yew” (Old Irish ibar) is the short
form of a compound name, perhaps
also Lugus10 which in a way falls out
of place. Lugus, being the name of
the highest god, would be unusual as
a man’s name in that form, but could
be the short form of a compound.11
Fig. 3. Graffito on the bottom of a vase
from Eyguières (Bouches-du-Rhône); Lapidary inscriptions which
the reading is Êρéîéá. Musée de Salon. consist of personal names are
memorial in character. The simplest
way to commemorate a dead person is
to have his or her name inscribed on an epitaph. The setting-up of stones, of course,
goes back to times immemorial, before the practice of writing was known. The
stone itself, bare or with lines of
ornament, or with the outline of
a human person, symbolized the
dead person and perhaps was
considered a repository of his
soul. Examples of such stones
are known, among other places,
from Württemberg, an early
Celtic area.
The next step would be then
to write the individual’s name on
to the stone. The cases where we
find just a single name are rare, Fig. 4. Bifacial stele on top of a funerary mound
Êïìá, Aτéλá, ]ruondu being at Tübingen-Kilchberg.

10
RIG I, G-159
11
Llywarch, name of a famous Welsh poet, could be such a name, going back to *Lugu-
markos “the stallion of Lugus”.
 15

instances.12 These names are in the nominative,

but Êρåéτå /krītē/, on a beautifully carved
stone from Nîmes13, probably is in the dative:
“(monument) for Krītis”. Definitely in the dative
is Aδãåííïõé , also from Nîmes14. The identity
of these individuals probably was well known or
could be inferred from the environment in which
the stones stood, constituting perhaps a family
burial-ground.
The more common practice, however, is
that a person is properly identified by reference
to his or her father. That is, the name occurs in
what may be considered its official form, with
patronym attached. There are two main types
of the patronymic formula. The first type gives
the father’s name in the genitive, e.g. Doiros
Segomari “Doiros son of Segomāros”. This
formula occurs hardly at all on epitaphs which
give nothing but the name of the deceased;
it occurs, however, in larger contexts.15 It is
not found in southern Gaul. No doubt it is an
archaism which had partly gone out of use, for
it is the usual way of expressing the patronym
in Celtiberian, and it occurs also in Cisalpine
Gaul and in Lepontic. In contrast to the practice
in Ireland and Wales, the word for “son”, which
would be *mapos, is never explicitly mentioned Fig. 5. Epitaph from Nîmes,
in Gaulish inscriptions. with inscription Κρειτε
“for Krītis”.
The other way of expressing the patronym
Nîmes, Musée Archéologique.
is by means of a patronymic adjective. Of the
two main formations, the one by means of a -kn-

12
RIG I, G-114, G-218, RIG II,1, L-5.
13
RIG I, G-213.
14
RIG I, G-208.
15
Doiros Segomari is the subject in the dedication to the god Alisanos on a bronze pan
from Couchey (Côte-d’Or): Doiros Segomari ieuru Alisanu, DAG, p. 492, RIG II,2,
L-133. On the ieuru inscriptions see below, pp. 28, 33 ff.
16

Fig. 6. Inscription on the handle of a bronze patera from Couchey (Côte-d’Or).

Dijon, Musée Archéologique.

suffix, type Ategnatos Drutiknos – “Ategnātos son of Drūtos”16, again is not found
in southern Gaul, only in the North and in Cisalpine Gaul, and it hardly occurs out
of context. Southern Gaul, i.e. the Gallo-Greek inscriptions, exclusively have a
form with suffix -ºo-, appearing as -éï- or -åï-; e.g.:17
Ìåδïõρåéî Ëé[τïõ]ìáñåïò
Aτåóqáò [Σ]ìåρτïõñåéãéïò
Εóκéããïñåéî Êïíδéëëåïò
fem., in the dative,
Εóκåããáé Bλáíδïïõéκïõíéáé
There are quite a number of examples of this type.
The next step would be to have more elaborate funerary inscriptions, giving
perhaps the name of the caretaking person, referring to the burial itself by means
of a verbal form, or giving other particulars.
A small group of inscriptions is characterized by the verb karnitu (3rd sing.)
or (3rd pl.) karnitus. Most explicit of all is the bilingual inscription found at Todi
(Umbria), outside the actual Celtic settlement area, preserved in the Vatican
Museum.18 It is also a duplicate inscription, with almost identical text on both
sides. The Gaulish text is in the North Italic alphabet. The inscription marks the
burial of one Ategnātos son of Drūtos by his youngest brother Coisis, and the
16
See on p. 19.
17
RIG I, G-71, G-3, G-207, G-146.
18
RIG II,1, E-5 (with ample bibliography).
 17

Fig. 7. Epitaph from Nîmes reading Εóκéããïñåéî Êïíδéëëåïò.

Nîmes, Musée Archéologique.

Latin text, as far as it is preserved, says in reference to the burial-place that Coisis
Druti f. frater eius minimus locavit et statuit.
18

Fig. 8. Epitaph from Coudoux (Bouches-du-Rhône),

with inscription Aτåóqáò [Σ]ìåρτïõñåéãéïò. Private.

Fig. 9. Woman’s epitaph from Gargas (Vaucluse); horizontal inscription on a stone

block: Εóκåããáé Bλáíδïïõéκïõíéáé. Avignon, Musée Calvet.
 19

The Gaulish text, with variant readings on sides A and B, is:

ateknati trutikni karnitu lokan / artuaś koisis trutiknos
In phonetic transcription, and with names capitalized, it would read:
Ategnati Drutikni. karnitu logan / artuaś Koisis Drutiknos.
To the mentioning of the name of the deceased, in the onomastic formula
(Drutiknos = Drūtī fīlius), there is appended a sentence which refers to the setting-up
of the burial site, loga apparently being the grave-bed and artua referring to the
covering or superstructure. The latter word stands in the accusative plural (-ś <
*-ns). It is apparently related to Old Irish art “stone slab”, cf. in Cormac’s glossary
art .i. cloch no leac lige19, lige being here the semantic equivalent of Gaulish loga,
and also etymologically related. The verbal stem karni- apparently is derived from
a noun which is well known from Irish and Welsh, carn “heap of stones, stone
memorial, ‘cairn’ ”. It was common, from times immemorial, to cover the dead
with stones (for which there probably were practical as well as religious reasons).
From this there arose the practice of laying out, or mantling the burial-place with
stones. Stones by themselves may be a memorial. From Ireland there is attested
the custom that, when warriors went to battle, everyone would pick up a stone and
throw it onto a heap. Returning from battle each of the survivors would take up
‘his’ stone again, the remaining stones thus symbolizing the number of the dead
and at the same time being their monument. From such a collective memorial it
is not a big step to an individual memorial. Here, in the civilized Mediterranean
world, we are a further stage away from the simple heaping-up of stones; karni-, a
verb with archaic content, has become a technical term for the erection of a grave.
κáñíéτïõ occurs once again in a defective inscription from southern Gaul.20 The
3rd pl. form karnitus occurs after a list of plural subjects introduced by tanotaliknoi
“descendants of Tanotalos” (= /Dannotalos/)21 in the inscription from Briona (near
Novara).22 Part of this inscription is illegible, but there is mention of takos toutas,
apparently the magistrate of the community (takos = /tagos/ rather than /tankos/).
One of the sons of Tanotalos carries the title lekatos (= /legātos/) which shows the
beginning integration of these Gauls into the Roman system of administration.

19
“art, this is ‘stone’ or ‘stone plate of a tomb’.” Cormac was bishop of Caisel and
king of Munster, a learned man. The glossary attributed to him dates from the late 9th
century.
20
RIG I, G-151.
21
For the phonetic form of this name, see the inscription from Alesia (p. 34 f.).
22
RIG II,1, E-1 (with ample bibliography).
20

Fig. 10–11. Bilingual funerary inscription from Todi (Umbria), on the front and reverse
sides of a cippus. Latin and Gaulish. The Gaulish text is in the North Italic alphabet.
Rome, Vatican Museum.
21
22

Fig. 12. Funerary inscription from Briona (Novara),

written in the North Italic alphabet. Novara, Museo Lapidario.
 23

We find a similar case in southern Gaul where in a fragmentary inscription23 a

person carries the title ðñáéôùñ (= Latin praetor).
The relation 3rd sing. : 3rd plur. which we find in karnitu : karnitus is a rather
strange one, the plural being marked simply by an addition of -s to the singular
form. We should have expected something like a t : nt relation in the endings, but
apparently t is not part of the ending but is a preterital suffix similar to the one
of the Germanic dental preterite. Another plural form of the same kind, lubitus,
was seen in a sentence from a La Graufesenque graffito the meaning of which is
obscure but where a quantity of three thirds plays a role:
aricani lubitus ris tecuandoedo tidres trianis.
But this inscription is no longer a valid testimonial, since aricani is not a plural
subject (as originally supposed) but the genitive of an otherwise well-attested
potter’s name Aricanos (or -us), and lubitus (whose reading was always in doubt)
is now being read as lubitías, so this inscription must be seen now in another
context (see p. 66 below).

Fig. 13. Graffito from La Graufesenque with a short sentence in Gaulish.

Collection Hermet, Musée de Rodez.

The same singular : plural relation applies also in the verbal form for the act
of dedication, 3rd sing. ieuru : 3rd plur. *ieurus (actually attested iourus).
We have seen that karnitu is a derivative of *karno- (or *karnā-). In the same
way logitoe24 in the inscription of Néris-les-Bains
23
RIG I, G-108.
24
If this is the correct reading; see the following note.
24

Bratronos Nantonicn(os) Epaðate÷torigi Leucullo suiorebe logitoe25

is a derivative of loga which occurs in the Todi inscription in the meaning “grave”.
But here the verb may just mean “placed it” and refer to a dedication, like the
form legasit on a vase with the inscription
Buscilla sosio legasit in Alixie Magalu26

Fig. 14. Vase with dedication Buscilla sosio legasit in Alixie Magalu from Séraucourt
(Bourges). Musée des Antiquitées Nationales, St.-Germain-en-Laye.

Another inscription which appears to have to do with funeral rites is on an

urn27; it reads
vercobretos readdas

25
The word-division and the reading are partly in doubt; Lejeune, RIG II,2, p. 94 ff.
proposes Leucutio and togitoi. logitoe (or logitoi) could be interpreted as a 3rd sing.
preterite form *logito + pronominal object *-ed (or *-id). suiorebe probably means
“with his sisters” (instrumental form, -be < *-bi), continuing (with internal loss of -s-)
the Indo-European lexeme *s}esor-.
26
Séraucourt (Bourges); CIL XIII, 10017,70; Dottin no. 47; DAG p. 354; RIG II,2, L-79.
27
Found 1978 at Argenton-sur-Creuse (the ancient Argentomagus); published by L.
Fleuriot, Études Celtiques 18 (1981), 93–97. Now in RIG II,2, L-78.
 25

Fig. 15. Urn from Argenton-sur-Creuse with inscription vercobretos readdas.

readdas no doubt is a verbal form; the meaning probably is that the magistrate
(vercobretos28) provided for the funeral. re-, a weakened form of ro- (< *pro), is
a preverb, and addas apparently contains another preverb, ad-, combining with a
form either of the Indo-European root *dhē-, zero grade Celtic *da-, or of *stā-,
zero grade Celtic *sta-, both in the meaning of “to place, to set up”. As to a
semantic equivalent, cf. German ‘beisetzen’.
If we leave funerary inscriptions and pass on to dedications, we may at once
draw attention to another important bilingual inscription found in 1966 near
Vercelli in northern Italy and apparently dating from the first century BC, some
time after the famous battle against the Germanic Cimbri.29 The Latin text, which
is more explicit and well legible, tells us that a certain Acisius Argantocomaterecus
28
A word known from Caesar, Bellum Gallicum 1,16,5 and from coins. Its etymological
meaning may be paraphrased by “commissioner of public works”.
29
First publication by M.G. Tibiletti Bruno, “La nuova iscrizione epicorica di Vercelli”,
Rendiconti della Academia Nazionale dei Lincei, XXXI 5–6 (1976), pp. 355–376;
subsequent discussion by M. Lejeune, “Une bilingue gauloise-latine à Verceil”,
Académie des Inscriptions et Belles-Lettres, Comptes rendus 1977, pp. 582–610, and
edition in RIG II,1, E-2. For further discussion, based on an autopsy of the stone, see
Meid 1989, 7–16.
26

Fig. 16. Stele with bilingual insccription in Latin and Gaulish, found near Vercelli.
The Gaulish text is in the North Italic alphabet. Vercelli, Museo Leone.
 27

(a dealer in money apparently, a banker, and, therefore, a rich man) donated these
grounds, a campus, both to gods and men:
… comunem deis et hominibus …
The Gaulish text, which is shorter and badly legible in parts, contains what
seems to be the semantic equivalent of the Latin phrase, in the form of a dvandva
compound consisting of the etymological cognates of Latin deus and homo:30
tevoχtonion
/dēvogdonion/
“of gods and men”

akisios · arkatoto[k]
materekos · tośo
kote · atoś · teuoχ
tonion · eu

Fig. 17. Facsimile of the Gaulish text of the Vercelli inscription.

The dvandva type is rare and archaic, and this compound shows an additional
archaic feature – the word for “man”, which in Irish and Welsh is duine or
dyn respectively, from Celtic *donios, occurs here with the older and original
consonant cluster gd-, protected here by compound juncture. *(g)donios means
“terrestrial” (as opposed to the celestial gods) and is a derivative of the ancient
Indo-European word for “earth” preserved in Greek χqών, Celtic *gdonios being

30
This striking interpretation is owed to M. Lejeune.
28

consequently formally equivalent to Greek χqόνéïò. The initial cluster has been
simplified in most languages.
As for dedications or tributes to gods there are two main types:
(a) One group of inscriptions (which are in the Greek alphabet and occur only
in the South) contains the phrase δεδεβρατουδεκαíτεμ.31
(b) Another type (written in either alphabet, Latin or Greek, and occurring
both in the South and in the North) contains the verb ieuru.32
In group (a) the recipient is always a god or a triad of mātrēs; no specific
object is mentioned apart from the phrase δεδεβρατουδεκαíτεμ which therefore
must express the dedication.
In group (b) specific objects are mentioned, except in the case where the
object is self-evident. The recipient is in most cases a god, but the particular
object may also be for the benefit of humans; for this reason, ieuru expresses a
solemn gift, normally in a religious context but not necessarily so.
Unfortunately the verbal form ieuru has so far resisted all attempts at formal
and etymological analysis; it may be a disguised reduplicated form.33 The form
δεδε, however, is formally transparent; it is a reduplicated perfect, possibly equal
to Latin dedit, as one would be inclined to think at first glance, but more likely
a form of the Indo-European root *dhē- (in Greek ôßqçìé, hqçκα) in the sense
of “dedicate”. âρατου, which is formally an instrumental, is cognate with Latin
grātus and therefore means “gratefully” or “in gratitude”. We can see then that
we have to do here with a vote of thanks. The rest of the formula had always been
misunderstood. Scholars had favoured a division βρατουδε καíτεμ (-ν), assuming
that äå was a postposition (= Latin dē). They were left then with an enigmatic
lexeme καíτεμ (-ν), apparently the object, but resisting all attempts at semantic
or etymological interpretation. The matter remained undecided until 1974 when
O. Szemerényi with a stroke of genius cut the Gordian knot, showing a solution

31
Complete dossier in M. Lejeune, “Inscriptions lapidaires de Narbonnaise, I-VII”, Études
Celtiques 12 (1968), 21–91. The formula may be shortened, or show abbreviations.
32
M. Lejeune, “Le dossier gaulois ieuru”, in: Recherches de Linguistique. Hommages à
M. Leroy (Bruxelles, 1980), pp. 110–118.
33
There may be loss of Indo-European *p involved which would account for the lack
of transparence. There are attempts to connect the form with Old Irish ro-ír “granted”
(present ernaid, a p-root): P.-Y. Lambert, Zeitschrift für celtische Philologie 37 (1979),
207–213, K.H. Schmidt, Zeitschrift für celtische Philologie 41 (1986), 373. See also
M. Lejeune, Hommages Leroy, p. 110 and Delamarre 2003, 188 f. (with additional
bibliography).
 29

which was self-evident in its simplicity.34 By drawing attention to Greek and

Latin dedications which contained äåêÜôç, or decuma respectively (“tenth part,
tithe”), like
MåóóÜνéïé êαr NαυðÜêôéïé PíÝqåí Äér EÏëυìðßùé
äεêÜôáí Pð’ τ§í ðïëåìßùí
or
Hercolei sacrum. Caesius C. l(ibertus) Tertius
decuma facta dedit donum mereto
he was able to show that the word division was äεäε âñáôïõ äεêáíôåì, the
phrase meaning “dedicated in gratitude the tithe”. Example35:
Ïõçâñïõìáñïò äεäε Tαñáíïïõ âñáôïõ äεêáíôåì
“Vebrumāros dedicated to (the god) Taranus gratefully the tithe”.
The lesson to be learned from this is that there is intercultural influence at
work, and that the awareness of such intercultural features will be of assistance in
the interpretation of particular Gaulish texts.36
The remarkable thing linguistically is that äεêáíôåì, accusative of *dekantā
or *dekantī, a feminine formation of Indo-European *de"»tos, is of archaic
formation, because the common form of the ordinal “the tenth” in Gaulish is
decametos which, with its thematic suffix, clearly is an innovation – one which
is, however, already Common Celtic (Old Irish dechmad, Welsh degfed, also
Celtiberian dekameta37 which occurs in the same technical meaning as Gaulish
äεêáíôåì).38 From a semantic point, though, the question remains what in the
individual case this special “gift” consisted in.
Some of these dedications are to the Mother Goddesses (usually a triad):

34
“A Gaulish dedicatory formula”, Zeitschrift für vergleichende Sprachforschung
88 (1974), 246–286. But see also M. Lejeune, “Quel celtique dans
ÄÅÄÅÂÑÁÔÏÕÄÅÊÁÍÔÅÌ?”, in: Studies in Greek, Latin, and Indo-European
Linguistics Offered to L.R. Palmer, ed. A. Morpurgo-Davies and W. Meid (Innsbruck,
1976), pp. 135–151, and again O. Szemerényi, Zeitschrift für celtische Philologie 44
(1991), 303–310. The superiority of Szemerényi’s interpretation has been stressed by
E. Campanile in Le lingue indoeuropee di frammentaria attestazione, ed. E. Vineis
(Pisa, 1983), p. 211 f.
35
Orgon (Bouches-du-Rhône); DAG no. 44, p. 90 f.; RIG I, G-27, pp. 52–56.
36
See also E. Campanile, loc. cit., p. 212.
37
Acc. TeKameTam in the first Botorrita inscription.
38
Cf. also fn. 7 above (end) for another possible case of archaic formation.
30

Fig. 18. Dedication to the god Taranus from Orgon (Bouches-du-Rhône).

Avignon, Musée Calvet.

Ìáôñåâï Íáìáõóéêáâï “to the Mothers of Nemausos” (Nîmes)39

Ìáôñåâï Ãëáíåéêáâï “to the Mothers of Glanum” (St.-Rémy)40
Ñïêëïéóéáâï “to the Very Renowned Ones” (also St.-Rémy)41

39
RIG I, G-203; beautifully carved inscription on a capital. Full text: [Ê]áôáñ[ïò
É]ëëáíïõéáêïò äεäε Ìáôñåâï Íáìáõóéêáâï âñáôïõ äå[êá.
40
RIG I, G-64; on the socket of a little altar. Full text: Ìáôñåâï Ãëáíåéêáâï âñáôïõ
äåêáíôåí. This name of the Glanian Mothers occurs also in the exactly corresponding
Latin form in a Latin dedication to the divinity of the source, from which Glanum
derives its name, to the Glanian Mothers and to Fortuna Redux: Glani et Glanicabus
et Fortunae Reduci ... (Glanum, in situ; see the photograph in F. Salviat, Glanum, St.-
Rémy-de-Provence 1980, p. 36). – The lexical element glan-, well attested in Celtic
river-names, means “clear, pure”.
41
RIG I, G-65; likewise on the socket of a small altar. Full text: Êïñíçëéá Ñïêëïéóéáâï
âñáôïõ äåêáíô. The epithet of the divinities (no doubt also the Glanian Mothers) is
based on the well-known Indo-European lexeme *"le}es- “fame”; for a formal analysis
and for the sound development see K.H. Schmidt, Studia Celtica 14/15 (1979/80),
285 f., who improved on the treatment by M. Lejeune, Études Celtiques 15 (1976–78),
95 f.; additional remark by E.P. Hamp, ibid. 23 (1986), 47. See also M. Lejeune, Études
Celtiques 16 (1979), 101 f. and RIG I, G-65 (interpretation “les Écoutantes”, with
reference to Auribus in Latin inscriptions). The corrrect interpretation is no doubt “to
the Very Renowned Ones”; it finds formal, if not semantic support in two formations
 31

Fig. 19. Dedication to the “Nemausian Mothers”. Nîmes, Musée Archéologique.

Fig. 20. Dedication to the “Glanian Mothers”. St.-Rémy-de-Provence, Hôtel de Sade.

32

Fig. 21. Dedication to the “Very Renowned” Mothers.

St.-Rémy-de-Provence, Hôtel de Sade.
 33

By the way, the new inscription from Larzac has come up with the nominative
sing. of the word for “mother” which is matir (and also with “daughter”, duχtir).42
An interesting fact is that we not only have dedications to Mothers but in
one instance also to Fathers: In an
inscription discovered on a stone in the
churchyard of Plumergat (Brittany)43 we
read … Atrebo Aganntobo …, the latter
word being apparently an adjective
from a local (tribal?) name, in the dative
plur. masc. (-obo as against fem. -ābo).
This may surprise at first glance, but
we must remind ourselves that ‘father’
is well known as an attribute of gods,
particularly in the Roman world (Mars
pater, Dis pater, Juppiter) but also
elsewhere; it is a trait of Indo-European
religion wheras the cult of the Mothers
is pre-Indo-European in origin.
Let us now cast an eye upon the
ieuru inscriptions.
In one well-known inscription44 a
citizen of Nemausos (Nîmes) dedicates
a íåìçôïí, a consecrated area, to the
goddess Belisama, in another case45 the Fig. 22. Stone in the churchyard of
object is a canecosedlon, an upholstered Plumergat (Bretagne), dedicated to the
seat (?), in still another case it is a “Agantian Fathers”.

from Vedic Indic adduced by E. Campanile, loc. cit., p. 215, one with pra- (= Gaulish
ro-) directly on the basis of the s-stem śrávas-: prá-śravas- (“weit berühmt” oder
“laut tönend”: H. Graßmann, Wörterbuch zum Rig-Veda, col. 883), one with suffix
-(i)ya- from the same base, but without prefixed pra-: śravasíya- (“rühmenswert,
preiserstrebend”: ibid., col. 1421).
42
Also the genitive plur. Ìáôñoí “(Sanctuary) of the Mothers” is known from a rock
inscription on a hill near Istres (Bouches-du-Rhône); see M. Lejeune, Études Celtiques
25 (1988), 97–101.
43
First edited by G. Bernier, Annales de Bretagne 77 (1970), 655–667, with a note by M.
Lejeune, 669–672; subsequently in RIG II,1, L-15, pp. 177–181.
44
Vaison-La-Romaine (Vaucluse); RIG I, G-153.
45
Autun; RIG II,1, L-10.
34

celicnon, some kind of building. This inscription is from Alesia46, and it is famous
for its relative clause with its relative verbal form (up to recently the only example
known from Gaul):
Martialis Dannotali ieuru Ucuete sosin celicnon
etic gobedbi dugiiontiio Ucuetin in Alisiia
“Martialis son of Dannotalos dedicated this building to Ucuetis,
together also with the metal-workers47 who serve Ucuetis in Alesia”
The beneficiaries
are the god Ucuetis,
and doubtless also
the gobed-, the metal-
workers. Alesia, as
the archaeological
evidence shows, was a
centre of metal works,
and the remains of the
building, the centre
part of which must
have been two-storied,
are still to be seen.48
The word celicnon has
passed into Gothic,
perhaps from Galatian
Fig. 23. Dedication of a nemeton by a citizen of Celtic; Gothic kelikn
Nemausos. Avignon, Musée Calvet.
denotes the upper part
of a building.49
46
Alise-Sainte-Reine; RIG II,1, L-13 (with ample documentation).
47
gobedbi which formerly was taken as dative plur. (“to the metal-workers”) is now
better interpreted as instrumental, in accordance with the inherited function of the
suffix, since more forms with instrumental force are now attested (e.g., mesamobi “by
the worst”, see p. 63 below). The word itself is related to Old Irish gobann “smith”.
48
The plate with the inscription was found in its immediate vicinity.
49
This fits the meaning of celicnon here, but unfortunately (if one may say so) there
exists another form celicnu, in a different context, on a graffito from Banassac: lubi
rutenica onobíía, tíedi ulano celicnu, where it is uncertain what it means, and so far
it has been impossible to make the two words compatible in meaning. The text of
this inscription is an advertising slogan of the buy me type. The first part seems clear:
“prefer Rutenian aquavit”, but for the second part anything goes. See in particular
 35

Fig. 24. Alesia. Remains of a building near which the inscribed stone plate,
shown below, was found and to which it doubtless refers.

Fig. 25. Dedication of a celicnon at Alesia in honour of the god Ucuetis.

Musée de Alise-Sainte-Reine.
36

Fig. 26. The inscription on the menhir of Vieux-Poitiers.

In another example, on a menhir (Vieux-Poitiers, still in situ50), we find ratin

briuatiom, “a bridge ramp”, as the object of ieuru. There is a river nearby, and
there are indications that there had been a bridge; no doubt the dedication was for
the public benefit.
Most remarkable, at first sight, is the inscription discovered 1953, together
with the upper part of a human statue, from St.-Germain-Sources-Seine:51
Aresequani Ariios iourus Luciio(n) Nertecoma(ri)
Äáγïëéôïõò áõïωõô
Here we have a plural subject Aresequani Ariios “the riverains of the Seine
(and) Arios”, and the plural form of the verb, iourus. But the interesting part of
it is the fact that – judging from the linguistic expression – the dedication would
appear to be not an object but a person: Luciio(n) Nertecoma(ri). But how is this
to be understood?
The sources of the Seine (today the property of the City of Paris, and still an
idyllic place) had been the place of a sanctuary, and many votive objects dedicated
to the dea Sequana have been found there, including wooden sculptures of human

L. Fleuriot, Études Celtiques 14 (1975), 443–450 and, more soberly, M. Lejeune,

Revue des Études Anciennes 81 (1979), 260. See below p. 66 on the lubi inscriptions.
50
Naintré (Vienne); RIG II,1, L-3, pp. 69–82 (with ample documentation).
51
RIG II,1, L-12; Meid 1989, 32–35.
 37

Fig. 27. Area of the sources of the Seine. It was the site of a sanctuary dedicated to the
dea Sequana, from which many votive gifts have been recovered.

Fig. 28. Modern classicistic representation of the dea Sequana in an artificial grove in
the source area of the Seine.
38

Fig. 29. Upper part of a stele with human statue and votive inscription,
excavated from the sanctuary of the Seine. Dijon, Musée Archéologique.
 39

Fig. 30. Votive gifts from the sanctuary of the Seine, representing human body parts.
Dijon, Musée Archéologique.
40

bodies and body parts, put there by individuals seeking cures.52 The dedication of
the man Lucios is of a different nature; it is honorific, but materially it is on similar
lines: the verb iourus apparently does not refer to the person but to the statue
which was erected in his honour, and dedicated to the divinity of the sanctuary.
Linguistically it is a clumsy formulation. The lower part of the inscription is in
Greek letters and constitutes the signature of the artist: “Dagolitus made it”. auot
which occurs several times and which is another mysterious verbal form53 thus
has the meaning of “fecit, dðïßåé”.
I ought to mention that apart from 3rd sing. ieuru and 3rd plur. iourus we now
also have an instance of what appears to be the 1st sing., ieuri (-ī < *-ai):54
… ieuri Rigani Rosmertiac
“I dedicated (this) to the Queen and to Rosmerta”,
the ‘Queen’ apparently being the Great Mother (cg. Welsh Rhiannon <
*Rīgantonā).55
Dedications or votive gifts are not the only cases where a divinity is mentioned
by name. Cult representations of the divinities, effigies, altar reliefs might carry
their names, as a few instances show. Thus on the famous altars from Notre Dame
de Paris56 we see several pictorial representations, one of a horned god with
superscription, now already very damaged and only partly legible, Cernunnos;
another one of a god with an axe who is cutting a tree, superscribed Esus; a third
one of a mythological configuration consisting of a bull standing behind a tree,
with three birds standing on the bull’s head and back showing through the foliage

52
See Meid 1989, 29 f.
53
The suggestion of P.-Y. Lambert in Mélanges J.-B. Colbert de Beaulieu (1987), p. 527 f.
that au(u)ot(e) is to be segmented into a preverb au- and a verbal stem }ot- from an
Indo-European root *}edh- “to conduct” would appear plausible on semantic grounds
(cf. German ‘ausführen’ in the sense of “to execute”), but does in no way account for the
phonetic and morphological problems and is therefore no more than a faint possibility.
54
On an inscription in Latin cursive script found at Lezoux, published by M. Lejeune and
R. Marichal in Études Celtiques 15 (1978), 151–156. The interpretation of ieuri as 1st
pers. sing. is by M. Lejeune; see also Hommages Leroy, p. 113, RIG I, pp. 448, 451.
It has been contested by O. Szemerényi who concludes that the form must be 3rd pers.
dual (subjects Rigani and Rosmertia): Zeitschrift für celtische Philologie 36 (1979),
295 fn. 11, ibid. 44 (1991), 305, fn. 4. A form ειωραι subsequently discovered and
interpreted as 3rd pers. sing. has resulted in a new discussion of the entire paradigmatic
complex; see RIG II, 2, L-67, p. 182.
55
About these connections see Meid 1991, 40–45, also 2010, 170.
56
Now in the Cluny Museum.
 41

Fig. 31. Representations of the

“Woodcutter God” Esus and of
the Tarvos Trigaranus, “the Bull
with the Three Cranes”, on a
Paris altar stone.
Paris, Hôtel de Cluny.
42

Fig. 32. The Woodcutter God and the Bull on a representation from Trier.
Trier, Rheinisches Landesmuseum.
 43

of the tree. The superscribed legend is Tarvos Trigaranus, “the Bull with the
Three Cranes”. The Woodcutter God and the Tarvos Trigaranus obviously are
part of one and the same mythological scenarium57, because they occur together
in a somewhat differently arranged representation from Trier.58
Let us now turn to texts of a magical character, and this is also where some of
the recently-found longer inscriptions come in. I shall treat in some detail (though
not exhaustively) the two most important ones, the one from Chamalières (near
Clermont-Ferrand), and the other from Larzac (near Millau). Both are written in
Latin cursive script on pieces of lead which, for several reasons, is the preferred
metal for writings addressed to the deities of the underworld, as the numerous
examples of defixiones in the Latin language show.
The already famous inscription from Chamalières, found in 1971 together
with many votive objects when the building which belonged to the Source du
Rocher was demolished and cleared for a new building site, has been well treated
already, and there is a good deal of agreement.59 The text is a plea to Maponos
(the youthful god and Gaulish Apollo) by a group of elderly men to provide a cure
to their various ailments – rheumatism, failing eyesight, failing potency – and it
is interesting, how the material finds from holy watering-places like Chamalières
itself (which is a renowned spa even today) and Sources-Seine (votive objects
consisting of sculptures of human bodies or parts thereof, mainly from wood)
bear out the interpretation by their pictorial language. The interpretation of these
inscriptions calls therefore for interdisciplinary research.60
In the beginning the Arvernian Maponos is being invoked (by his name, the
“youthful” god and in Celtic mythology the son of the Magna Mater Matrona61),

57
This may have had to do with Destruction (Death) and Regeneration (Rebirth) as
cosmic forces.
58
This, however, carries no inscription.
59
First edition by M. Lejeune and R. Marichal, Études Celtiques 15 (1976/77), 156–168;
see also L. Fleuriot, Études Celtiques 15 (1976/77),173–190, 16 (1979), 135–139, 17
(1980), 145–159; P.-Y. Lambert, Études Celtiques 16 (1979), 141–169, idem, Bulletin
of the Board of Celtic Studies 34 (1987), 10–17; K.H. Schmidt, Bulletin of the Board
of Celtic Studies 29 (1980/82), 256–268; P.L. Henry, Études Celtiques 21 (1984),
141–150, Meid 1987, 48–53, Meid 1989, 27–31. There have also been observations
by various scholars on single points which, however, are too numerous to be recorded
here. See RIG II,2, L-100, pp. 269–280.
60
Cf. Meid 1987, 27–31 (with further references).
61
Cf. Mabon uab Modron, a remote character in medieval Welsh saga; see Meid 1991,
42.
44

Fig. 33. Lead plate with magical inscription excavated, together with many wooden
sculptures of human bodies or body parts, from the site of the Source du Rocher at
Chamalières (Puy-de-Dôme). Clermont-Ferrand, Musée Bargoin.

Fig. 34. Facsimile of the Chamalières inscription, by R. Marichal.

 45

who with the help of brixtia anderon is supposed to effect (lit.: “to speed up”62)
something on behalf of a group of male persons one of whom is their “speaker”.
brixtia is related to Old Irish bricht “magic formula”, and brixtia anderon,
according to M. Lejeune, may mean “by magic of the subterraneans” (the dī
inferī), anderon in this case being genitive plur. of a word anderos which would
be the Gaulish equivalent of Latin inferus and Sanskrit adhara- “nether”, all three
from Indo-European *§d[eros. There is another possibility, however (advocated
by P.L. Henry), that brixtia anderon may mean “by magic of women”, anderon in
this case being connected with Old Irish ainder (from *anderā) “(young) woman”.
This interpretation would find support (apart from a famous passage in Old Irish
literature63) in the occurrence of brictas “magically powerful” qualifying mnas
“women” in the Larzac inscription in which women magicians play a dominant
role.
Another word with magical significance in the Chamalières text is naritu in
the instrumental phrase risu naritu “with magically powerful inscription”. This
word derives in last instance from the Indo-European root *(ə)ner-64, the basic
meaning of which is “creative force”, hence “magical power”, with derivations in
form of Celtic *nerto- “power, strength”, Germanic Nerthus, name of a goddess
glossed by Tacitus as “Terra mater”. The immediate base of naritu- seems to be
a verbal stem with causative force and long o-grade ablaut *nōr-e~- or *nōr-ī-
(Indo-European *ō > Celt. ā) “to strengthen magically”, from which we would
get *nārito- “magically strengthened”.65 After a list of the names of the persons
involved (one of which is their “speaker” or advocate, adgarios66), and after the
mention of a special group “who will swear”: toncsiíontío67, there follows the
central part of the text in which in somewhat cryptical terms three magical effects
are formulated which can be understood as transformations from a negative to a

62
lotites; lōt- cf. Old Irish lúath “swift”.
63
In the Lorica of St. Patrick, which provides protection, among other things, “against
the spells of women, smiths and druids”: fri brichta ban ocus gobann ocus druad;
Thesaurus Palaeohibernicus II 357.
64
IEW 765, in laryngealistic notation *h2ner-.
65
For the long o-grade in this root cf. Hom. Greek í§ñïø in íþñïðé ÷áëê² and íùñås
· díåñãås Hesych.
66
Root gar- “to call” (Old Irish gairid). Since gar- is semantically equivalent to Latin
vōcāre, Gaulish adgarios may be a calque on Latin ad-vōcātus.
67
Cf. Old Irish tongid.
46

Fig. 35. Votive gifts from the sanctuary of the Source du Rocher at Chamalières.
Clermont-Ferrand, Musée Bargoin.
 47

positive state and which are linguistically expressed by three pairs of semantic
oppositions:
meion ponc sesit buetid ollon
“Small … shall become big”
reguccambion
“I stretch (what is) crooked”
exops pissíiumi
“(As one) deprived of eye-sight I shall see”
Interesting verbal forms in these phrases are buet-id, equatable with Old Indic
(Vedic) bhuvat, and pissíiu-mi, 1st sing. future of pis- “to see”, see also p. 67 on
appisetu (< *ad-pis-), cf. Old Irish ad-cí “sees” (Indo-European root *kwis-).
Apparently these are effects desired by the persons in question, clothed in
speech forms which purport to anticipate the result and are thus, in this context,
magically performative.68 In the final part of the text this magical procedure is put
into operation by going through the appropriate ritual.
The inscription found in 1983 at Larzac (in the vicinity of Millau, about
15 km south of the great pottery centre La Graufesenque) contains about 160
words and is thus the longest Gaulish text which we now possess. It was found in
a grave chamber. The text was inscribed on a lead plate which had been broken
into two pieces; these served to cover an urn which contained the remains of a
female body. The four sides of the broken tablet were originally inscribed with
a continuous text by one and the same hand; subsequently, however, a different
person on one side erased several lines of the original text and substituted another,
different text. The original text therefore is no longer continuous but is interrupted
by this superimposed short text. Apart from the loss of some letters at the margins
through decay the main body of the text is well preserved.
The text of the Larzac inscription is of extraordinary interest since, unlike
Chamalières, where men are the protagonists, the dramatis personae in this piece
are women. Not ordinary women, though, but members of a rather formidable
breed of mnas brictas “women endowed with magic”. Apparently there are two
rival groups. One is represented by two females, Severa and Tertionicna, together
with their unnamed indigenous and non-indigenous followers. This group had
apparently practised harmful magic upon another group, and it is this other group
which, with the help of a “wise woman”, uidlua, tries to counter this attack,

68
See Meid 1987, 50–52.
48

Fig. 36. Fragment of the longest Gaulish inscription, found at Larzac (Aveyron),
near Millau. The text is a counter-charm against a group of women magicians.
Here the beginning of the text is shown. Musée de Millau.

reduce its effects, render Severa and Tertionicna innocuous, and even proposes
some kind of a non-aggression pact.
The Larzac text was first edited in 1985 by a French team under the direction
of Michel Lejeune.69 On that occasion, two members of the team, L. Fleuriot and
P.-Y. Lambert, offered complete interpretations of the whole text which, however,
are divergent or even contradictory in many parts. On the other hand, M. Lejeune
exercised remarkable restraint, prudently limiting himself to analyzing the structure
of the text and discussing some of its salient features. All this tends to show that the

69
“Le plomb du Larzac”, Études Celtiques 22 (1985), 88–177 (also separately: Le plomb
magique du Larzac et les sorcières gauloises, Paris 1985).
 49

Fig. 37. Facsimile of this fragment of the Larzac inscription by A. Vernhet,

the excavator of the site.

text, at that time, was still far from being fully understood. In the meantime some
progress has been made in particular by K.H. Schmidt who interpreted several
passages and commented on some of the linguistic forms of the text70, with the
result that a better, though by no means complete understanding has been arrived at.
For reasons of time and space I will not enter here into a detailed discussion of the
text71, but restrict myself to pointing out some of its interesting features.
70
“Zum plomb du Larzac”, in: Celtic Language, Celtic Culture, Festschrift for E.P.
Hamp (Van Nuys, Cal. 1990), 16–25, also in Linguistique Balkanique 31 (1988),
25–29, finally and comprehensively in Meid – Anreiter 1996, 23–36.
71
For such a discussion see W. Meid in Meid – Anreiter 1996, 41–50 as well as the
important contribution by K.H. Schmidt in the same volume, pp. 23–36. The whole
text is now re-edited and discussed in RIG II, 2 under L-98, pp. 251–266. Most recent
50

The subject is women: mnas (nom. pl.), mnanom, bnanom (gen. pl.); they
are characterized as brictas “possessing magic” or by other attributes. A certain
number of them are mentioned by their names, and in this list they are defined
by reference to another women who may be either the mother or the daughter of
the person in question. In a few cases this reference is mutual. The word duxtir
“daughter” is new in this context72, and so is the nominative matir. Another term
of relationship is dona, of uncertain meaning and provenance: it hardly represents
Latin domina (through domna, donna), more likely it is a genuinely Gaulish
word, connected with Old Irish duine, Welsh dyn “person”.73 dona is followed, as
far as one can make out, by a form in -us or -ius which Lejeune would interpret
as instrumental plural of an o-stem: dona paullius “female person in relationship
with the Paullians”. In connection with Severa and Tertionicna74 andogna
“indigenous” and anandogna “non-indigenous” occur, also acolut[ “followers”
(or verb “to follow”), ultimately of Greek origin.
Severa and Tertionicna are to be rendered lissata, liciata; these words apparently
express specific effects of magic: perhaps “spell-bound” and “fettered with bonds”
(‘fascinated’, in its literal sense), respectively; related nouns are lissina, licina.75
Their dreaded practice is ni-tig- “to stick (stab, prick) into”, hence “perform harmful
magic”: ponc nitixsintor sies “when they should perform harmful magic”.
This verb has the same technical meaning as Latin defigere (defixus; defixio)
of which it seems to be a semantic calque, though from a formal point infigere
is nearer76; its etymological connection, however, is with Latin in-stīg-ā-re
“instigate” (originally by pricking into) , Sanskrit tig-má- “sharp, pointed”, Greek
óôßãìá, English stick, stitch, German stechen.77 The root sag- “to pursue” (Old
Irish saigid) is present in a participial formation sagitiont- and in an agent noun
discussion of the text (with special regard to typology, structure and rhetorical features)
by B. Mees, “The Women of Larzac”, Keltische Forschungen 3 (2008), 169–188.
72
It occurs in the form typical for the Western Indo-European languages, without
intermedial vowel (“laryngeal” reflex).
73
The insular Celtic words, however, like the /dēvo-gdonion/ of the Vercelli
inscription, show a different formation, with suffix -ºo-, whereas dona implies a
thematic formation *don-o-s/-ā. The derivation from Latin domina is defended by G.
Neumann in Indogermanica et Italica. Festschrift für H. Rix, Innsbruck 1993, 340.
74
It is not quite clear whether Severa Tertionicna (so in most instances) represents one or
two persons (Lejeune is in favour of two).
75
Cf. Greek ëßóóïìáé “entreat” (?). Latin licium “thread”. The Gaulish forms probably
stand in some loan-relationship to these words.
76
Cf. gladium hosti in pectus (Cic.), sagitta infigitur arbore mali (Verg.).
77
Indo-European root *(s)teig-/(s)tig-, IEW 1016 f.
 51

formation adsagsona “persecutrix”. Another verb is peti- “to spare”, connected

with Welsh ar-bedu, Old Irish ar-cessi in the same sense: suet petidsiont sies
peti sagitiontias seu[er]im tertio[nicnim] “inasmuch as they shall spare, spare
the ones persecuting Severa (and) Tertionicna”. Noteworthy are the 3rd person
imperatives biietutu, bi(i)ontutu “shall be” or “shall strike”(?)78 on account of
their doubly marked endings, though it is not clear in whatever way these are to be
explained.79 In the short but rather obscure text written by the second scribe there
occurs what seems to be the term for the “underworld”, antumnos (for andumnos)
corresponding to Welsh annwfn.80 Other words of interest are anatia “soul” (cf.
Welsh eneid), anuana “names” (which shows vocalization of internal m through
lenition; cf. Old Irish anmann, Middle Welsh enwain), barnauno- “standing trial,
being judged”, a participial formation with *-mno- (likewise with vocalization
of m) from a verbal stem barna- “to judge” (cf. Old Irish barn “judge”, Welsh
barn “judgement”), ratet, a 3rd sing. verbal form in the sense of “pledge, promise,
guarantee” (cf. Old Irish ráth “pledge”).
All in all, this text offers profound insights into the practice of sorcery,
the belief and superstitions connected with it, and into the status and social
organization of the persons involved.
Another text of apparently magical character was found in 1973 at Lezoux.
It is inscribed on a metal lamina folded together around a coin and pierced,
obviously to be worn around the neck as an amulet. The text was subsequently
edited by L. Fleuriot81; the reading is partly in doubt and the text itself for the
most part obscure.82 One can only make vague guesses as to the meaning of the
text and its function. It seems to be a protective charm, apparently designed to
protect a person named Dagilos (spelt dagilox) on a particular journey, or more
generally, on his journey through life. This type of protective charm, where it
occurs in a literary form, is called lōrīca (which is Latin and means literally

78
In the latter meaning probably forms of the root bi- in Old Irish benaid “strikes”, subj.
-bia; equally possible, even preferable, from a formal point would be “shall be” (root
of Old Irish biid, Latin fiō), but this seems to be ruled out by the syntax of biontutu
which appears to be governing an accusative object.
79
Inherited imperative ending *-tu + affixation of the adverbial particle Indo-European
*-tōd found also in the “future” imperatives of Latin, Greek and Sanskrit (Old Latin
datōd, Sanskrit bharatād)?
80
The original Celtic form perhaps was *ande-dubnos “very deep”, see Meid 1991, 50
with fn. 6.
81
Études Celtiques 23 (1986), 63–70.
82
Now re-edited in RIG II, 2, L-101 with alternative readings also by R. Marichal.
52

Fig. 38. Amulet found at Lezoux with Gaulish inscription. Drawing by L. Fleuriot.
 53

“breastplate, cuirass”). There are two famous literary loricae in Old Irish,
one of them a charm which St. Patrick is supposed to have sung when being
ambushed by his adversaries.83 The other lorica, although in Christian guise, is
in essence pagan. In our case we would thus have an example of a Gaulish lorica
designed to guide its wearer through all dangers on a journey; for instance he
should chase away begging women: mendicas soniti. Interestingly mendica is
the well-known Latin word (feminine of mendicus “beggar”), whereas soniti
is connected with Old Irish sennid “chases, hunts”. The second Irish lorica
mentions vagabond women in the same breath with robbers and armed bands
which shows that they were, in their special way, potentially dangerous to the
lone traveller. 84
Other inscriptions serving a magical or a similar purpose are known or have
been discovered recently, but most are of limited interest, being partly or wholly
unintelligible, and may be passed over here.85
Of great interest, however, is another rather long Gaulish text, eleven lines
inscribed on a tile, which came to light in 1997 during excavations at a Roman
site near Châteaubleau, a small village in the North of France (dep. Seine-et-
Marne). This text, evidently composed by a man, who speaks several times of
himself in verbal forms of the 1st person singular and addresses himself in the
first place to a woman, but referring also to her family background, was first
edited, with ample commentary, in 1998 by P.-Y. Lambert in Études Celtiques,
then again in 2002 in the latest volume of Recueil des Inscriptions Gauloises,
also by Lambert.86 Lambert regards this rather personal text as some sort of
marriage proposal, other commentators seem to agree, though, considering its
many uncertainties, with certain reservations,87 but most recently B. Mees has

83
See also fn. 63.
84
Nám millither téol ná cuire ban ná cuire buiden “May no thief attack me, nor a
company of women, nor a company of warriors”: D. Greene – F. O’Connor, A Golden
Treasury of Irish Poetry A.D. 600 to 1200, London 1967, p. 34 f.
85
Some are not really Gaulish inscriptions as their basic language is Latin but containing
Gaulish words or phrases, and of course native Gaulish onomastics. The opposite
(Gaulish texts containing Latin elements) also occurs, as we have seen.
86
P.-Y. Lambert, “La tuile gauloise de Châteaubleau (Seine-et-Marne)”, Études Celtiques
34 (1998–2000), 57–115 and RIG II, 2, L-93.
87
P. Schrijver, “The Châteaubleau Tile as a link between Latin and French and between
Gaulish and Brittonic”, also in Études Celtiques 34, 135–142 (stressing in particular
the linguistically late character of the text); D. Stifter, “Notes on Châteaubleau (L-93)”,
Keltische Forschungen 4 (2008), 229–244.
54

Fig. 39. The Châteaubleau tile, photo and facsimile by P.-Y. Lambert from RIG II,2,
L-93 p. 239; also in Études Celtiques 34 (1998–2000), p. 64 f.
 55

tried to interpret it as a defixio, a curse against the woman in question, hence also
as a magical text.88 His interpretions, however, which he supports by typological
arguments, seem to me rather forced and, on the whole, do not carry conviction.
I think that Lambert is basically right, though his suggestions need not always be
correct. Considerable parts of the text remain unclear, but the general meaning
can be inferred from many indications. The text falls into three parts. In the first
part the man, who does not name himself, states his interest in a certain woman
whom he also does not name, probably in order not to compromise her. But he
will not formally propose before he knows whether she or her family will accept
him as a suitor. The central part seems to be about getting positive signals which
would encourage him to propose formally. If she would have him, and tell him
so – so the final part –, she would get a very good husband in him, and he would
be glad to take her as his wife.
In what follows I give an unedited transliteration of the text:89
1 nemnaliíumi beni. ueíonna incorobouido
2 neíanmanbe gniíou apeni temeuelle íexsetesi
3 sueregeniatu o quprinnopetamebissi íeteta
4 miíi íegumi. suante ueíommi petamassi papissone
5 suirexetesi íegiíinna anmanbe íeguisini
6 siaxsiou. beíiassunebiti mot upiíummiateri
7 xsi índore core. nuana íegumisini. beíassusete
8 sue cluiou sedagisamo cele uiro íonoue
9 ííobiíe beíiassusete rega íexstumisendi
10 me. setingi papissonebeíiassusetemetingise
11 tingibeíiassuseteregarise íexstumisendi

88
B. Mees, “Words from the well at Gallo-Roman Châteaubleau”, Zeitschrift für celtische
Philologie 58 (2011), 87–108.
89
Mainly after Lambert, with some minor alterations, barring errors. This transliteration,
however, does not make the morphological and syntactic structure of the text
transparent. Word or morpheme boundaries are often unmarked or in doubt, spacings
often wrong or misplaced. In an edited text word boundaries should be marked, and in
subsequent quotations I have marked them according to my understanding. B. Mees
in the article cited above provides an edited text (p. 91) with which I can agree to
some extent, but my interpretation would be different. – As in other transcriptions, í
(with accent) renders I (i longa); x denotes the guttural spirant ÷. Double ss in certain
forms is barred: ss, representing the characteristic Gaulish double dental spirant (the
so-called Tau gallicum), finally resulting in ss.
56

In his opening remark the man, rather surprisingly, states that he has nothing
against women: nemnaliíumi (= ne mnā(s) līºū-mī) “I do not accuse women”
(cf. Old Irish líid “accuses, violates”)90, but since he obviously wants one for
wife (which the following beni ueíonna seems to indicate) he thinks he needs
to be careful going about this matter, not publicizing his intentions by means
of contracts and mentioning of names: in corobo uido neí anmanbe gniíou “in
contracts publicly (= “wittingly”) I do not do it by name”91, continuing “so that
they (lit. “you”, plur.) could not say that you want me”92, ape ni te me uelle
íexsetesi, taking ape as a conjunction (< *at-kwe), ni as the negation, íexsetesi
as the verb (2nd plur., s-subjunctive; root *ºek- “speak, declare”, cf. Old High
German iehhan, Middle Welsh ieith “speech, language”), and te me uelle as a
Latin accusativus cum infinitivo. Since many Gaulish texts contain Latin words
or phrases this may not be surprising.93
In the middle part in which only few words stand out clearly, the speaker
apparently mentions the family of the woman (sue-regenia, cf. Welsh rhieini
“parents”), wondering whether they would accept him as a suitor. In that case
he would proceed (siaxsiou “I shall follow up”, corresponding to Old Irish
sïass-, future of saigid) and declare his intent, make known the names and see
(piíummi94) her father: ateri-xsi95, with a view of making a contract (ín ... core).
He then expects the woman to signal her consent: se te sue cluiou “if I hear
you so”, se dagisamo cele, uiro íono ueííobííe “if you desire a very good husband,
a right man”96, beíiassu “then I would like to be it”. The end seems to convey the
idea “If you are willing I am willing too”: me se tingi papissone beíiassu “if you
will accept me as your husband I shall be it”, se te me tingi, se tingi, beíiassu “if
90
It is a cautionary statement; the meaning may not be so drastic, perhaps “I do not wish
to say anything disfavourable about women” (or “to offend women”).
91
anmanbe (-be < *-bi) instrumental plur.; the plural in “by name” may refer to the duo
or tria nomina to identify a person, or perhaps to the names of both persons involved.
corobo I take as dative plur. of the word corresponding to Old Irish cor “contract”;
gniíou corresponds exactly to Old Irish gníu “I do, make”.
92
Alternatively “that I want you”.
93
It would be possible to understand this phrase also as Gaulish by reading ueííe (for
ueííe(s)) instead of uelle: “that you want me” (the root being *}ī-, cf. Latin vīs (in the
paradigm of velle: volō, vīs, vult).
94
pis- “see” (cf. p. 47 above), with intervocalic loss of s.
95
Possibly for ateri(n) (i)xsi, the latter Gallo-Latin = ipse.
96
cele = Old Irish céile “companion, partner, husband”, Welsh cilydd; uiro I take to be
the word for “man” (*}iros), but it could also be the adjective *}īros “true” (Old Irish
fír, Welsh gwir); íono = Welsh iawn “just”.
 57

you will accept me, if you will, I shall be it”97 – but “you should tell me this”:
íexstu-mi sendi.98.
As regards the form tingi which I have translated here with “accept”, I do
not think one can impute it the strongly negative, “stigmatic” meaning of tig-
in the Larzac plate (ni-tixsintor; see p. 50 above); the situation and the context
demand a positive meaning: the wooing must be concluded by mutual acceptance.
Therefore I posit “take, accept” as the meaning of tingi, connecting it, though
hesitantly, with the root *tek-99 which is attested in Old Irish in the sense of “take
hold, take possession” (techt- in techtaim “have”, techtaigid “takes possession”,
con·tetaig “has in common, shares”) and in Germanic (Old Norse þiggja, Old
Saxon thiggian in the meaning “accept, receive”) and which, in Lithuanian tenkù,
tèkti, shows also nasalization in the present stem. Gaulish ting- thus could be
explained as a nasalized form of *tek- with raising of e > i before the nasal and
lenition of the final consonant.
What I have tentatively translated as “husband”, papissone, which previous
commentators have taken as a personal name (a divine name: Lambert, the
woman’s name: Mees), I rather take as a noun, the obvious base of which is the
familiar papa word. The (seemingly) hypocoristic formation *papissō (or already
the Vulgar Latin general form papissone) may denote, rather jokingly, the pater
familias, as the suitor already sees himself (so to speak as “the future father of
your children”).
The text of this inscription which may be dated into the late 2nd or the 3rd
century represents a variant of northern Gaulish with remarkable late dialect
features. Most remarkable of these is the loss of final consonants, in particular of
s and n, the original markers of nominative and accusative, which is in line with
the contemporary development in spoken Latin. There is weakening of unstressed
vowels, curious phonetic breaking (like -ei- for *iº or *ī) or diphthongization,
so in the final *-ū of 1st person singular forms (gniíou, cluiou, siaxsiou), which
however is preserved as such before a suffix (as in íegu-mi). The root *ºek- “speak,
declare” appears in numerous verbal or nominal forms which exhibit lenition of
k > g in intervocalic position, whereas lenition elsewhere is not regularly marked

97
beíiassu is a form of *biºa- (< *bh}iºå-), cf. Old Irish bíu, Welsh byddyaf, but the
Gaulish form apparently is extended by a dental suffix plus futuric or desiderative s
resulting in the characteristic double spirant (Tau gallicum or ss).
98
I accept here the analysis of this form by P. Schrijver in the article cited, p. 138, but
with *ºek- as the verb.
99
IEW 1057 f.
58

in writing (which, being by nature conservative, need not always represent the
actual pronounciation).
All in all, the Châteaubleau tile is a very interesting document which invites
further elucidation.
The text I am going to comment on next, is, in my opinion, also some sort
of a “love story”, though coming from quite a different milieu. It is not exactly a
Gaulish text, though it was considered as such in the beginning. It dates from about
the same period (2nd to 3rd century). It is written on a lead tablet found already
in 1887 during excavations at the site of ancient Rauranum, now Rom (Deux-
Sèvres), situated at the Roman road between Saintes (Mediolanum Santonum)
and Poitiers (Linovum Pictonum) in Aquitania. This text, inscribed on both sides
of the tablet in scriptio continua, deciphered and edited, with facsimile, in 1898
by Camille Jullian100 and believed by him to be in the Gaulish language, has
for quite a while passed for the longest Gaulish text extant at the time, so still
in Dottin.101 But the knowledge of Gaulish was very restricted at that time, and
since the text was incomprehensible there was no safe basis for that attribution,

Fig. 40. Inscription of Rom (Deux-Sèvres). Facsimile by C. Jullian

100
Revue Celtique 19 (1898), 168 ff.
101
Dottin no. 52.
 59

especially as some passages appeared to be in Vulgar Latin, as noted in particular

by Whatmough and Pokorny.102 This, however, did not deter some scholars from
producing Gaulish interpretations103, but these, being largely fanciful, need not
be taken seriously. In fact the text was never really understood, and celticists
could not make much of it. So it came as something of a relief when in 1962 it
seemingly was demonstrated that the text was in plain Latin, representing a defixio
in a milieu of rival stage actors. This apparent demonstration was produced by
the Austrian archaeologist Rudolf Egger in collaboration with the Breton celticist
Christian Guyonvarc’h who had provided photographs, and it was published in
the memoirs of the Austrian Academy of Sciences.104
But this new reading of the text, differing considerably from the previous
ones, and its novel interpretation seemed too good to be true, and raised serious
doubts. An autopsy of the object which I made in the Museum of St.-Germain-en-
Laye in 1987 and another one in 1997 convinced me that the new readings had no
factual basis, and were in fact chimerical, hence the interpretation based on them
was illusionary, the result of wishful thinking. The consequence was that one had
to return to the older readings, eventually correcting them where necessary. As the
surface of the object is rather deteriorated and the writing badly legible in parts,
also of irregular ductus and with many ligatures, it will be difficult to provide
incontestable readings throughout, alternative readings being possible in not a
few places.105 But the frequent repetitions or variations of certain phrases will be
of help in establishing the text. The latest treatment of the object in vol. II,2 of the
Recueil provides revised readings by R. Marichal; some of them, however, are
doubtful or improbable for linguistic reasons. As for the supposed meaning of the
text, no opinion is provided.106

102
Whatmough, DAG p. 391 f., J. Pokorny, Celtica 3 (1956), 306.
103
O. Haas, Zeitschrift für celtische Philologie 23 (1943), 285–297, G.S. Olmsted,
Journal of Indo-European Studies 19 (1991), 283–286.
104
R. Egger, Die Fluchtafel von Rom (Deux-Sèvres). Ihre Entzifferung und Sprache. Wien
1962 (Sitzungsberichte der Österreichischen Akademie der Wissenschaften).
105
I have refrained from marking uncertain letters, because there are so many of them. To
discuss all potential readings would lead too far here and would be the task rather of a
critical edition of the text.
106
RIG II,2, L-103, pp. 285–296, with ample bibliography.
60

I have dealt with this text twice, in publications of 1996 and 2007, to which
I refer for more detailed information and opinion.107 Here I shall only give a
summary.
The texts on both sides of the tablet are conventionally marked A and B; in my
opinion the sequence is B – A, since the first word on A is a subordinate Gaulish
conjunction: ape (possibly < *at-kwe).108 My provisional reading is as follows:
B 1 teuoraiimo A 1 apeciallicarti
2 ehzaatantotehon 2 etiheiontcaticnato
3 zoatantatecom 3 nademtisseclotu
4 priatososioderti 4 cilasedemtitiont
5 noipommioateho 5 eticartaontdibo
6 tissepoteatepri 6 nasosiodeuipia
7 auimoatantateh 7 sosiopurasosio
8 ontezatimezo 8 gouisa[--]ehotisse
9 ziateuoraiimo 9 sosiopuraheoti[
10 apesosioderti 10 suademtaapo
11 imonademtisse 11 dunnauoliset
12 ueie[-----------]

The base language is Latin, even if this is not obvious at first sight. What kind
of Latin is it then? Apparently a Vulgar Latin on a very low level, on the way
already towards Romance; so priauimo (B 6–7) < precavimus is already near to
French (nous) priâmes.109 This Vulgar Latin is mixed with elements which may
be Greek, and with Gaulish ones such as the pronoun sosio.
As regards the script, the occurrence of several instances of h, and moreover
of z, is worthy of note. Of course h occurs also in late Latin texts, but it is no
longer pronounced and is therefore often written as a mute letter (before vowel,
or in hiatus). z is not an original Latin grapheme, nor a Gaulish one; it has passed
into Latin from Greek, and it may be taken here as an indication of possible Greek

107
„Die Inschrift von Rom (Deux-Sèvres)“, in: Meid – Anreiter 1996, pp. 118–123;
„Die pseudo-gallische Inschrift von Rom (Deux-Sèvres). Text und Interpretation“, in:
Gaulois et Celtique Continental (ed. P-Y. Lambert et G.-J. Pinault), Droz 2007, pp.
277–284.
108
This occurs also in B 10 and in the Châteaubleau tile (p. 56 above).
109
Of course, this verb being originally deponens (precārī), we would never get a form
like this in Classical Latin (but precati sumus instead).
 61

(or Latinized Greek) elements in the text.110 These elements apparently refer to
the religious sphere. So in B 7–8 we find atanta tehon which I would interpret
as áθáíáôá θåωí “immortal one of gods”,111 while zia (B 9) and contracted za
(B 2, 8) may be < Greek δsá, δºá (> æá) “divine” or perhaps < θåßá (of θåsïò),
also “divine”.112
Another example of Greek is timezo (B 8) “I shall honour” = ôéìÞóω, future
of Greek ôéìÜω, and, if I am not mistaken, we can find the verbal expression of
male sexual activity, Greek ïnöω, in B 5 oipommio which may be a Gallicized
relative verbal form *oipomi-ºo “quam futuo”.
As one can already guess, love is involved here. Perhaps the loving couple
belongs to the rank of slaves, which would account for the low linguistic register
and the mixed language of the text. The following scenario may be imagined. A
young man, Catignatus, is in love with a young woman, Clotucilla, who is his
concubine. He refers to her as his derti (literal meaning “skin”, related to Greek
δέρω, δέρìá – a figurative term for a person of female sex, comparable to Latin
scortum)113: compriato sosio dertin (B 3–5) “this beloved darling”, sosio derti(n)
imo(n) (B 10–11)114 “this my darling”. He fears that she could be taken away from
him (na demtisse A 3, B 11) and implores the goddess Divona (dibona, A 5–6)
that, should she be taken from him (se demtitiont, A 4), she (sua demta, A 10) be
restored to him. The basis of these extraordinary verbal forms is of course Latin
dēmo, demptus, but in apodunna (A 10–11) we have a possible reflex also of
Greek Pðïäï™íáé “give back”.
110
The occurrence of Greek elements in Gallo-Latin and then also in Gaulish speech need
not surprise on account of the vicinity of Marseille, a Greek colony by origin, and the
use of Greek letters in early writings. The Gallo-Latin charm on a silver plaquette of
Poitiers, beginning bis gontaurion analabis ... bis gontaurios catalages (Meid 1980, 9
f., 29, RIG II,2, L-110) features, apart from the name of the medicinal herb itself, two
Latinized Greek verbal forms (Píáλάβηò, κáταλλάγηò), and we find Greek elements
also in the Gaulish healing charms recorded by Marcellus of Bordeau (Meid 1980, 10,
more in Heilpflanzen und Heilsprüche. Zeugnisse gallischer Sprache bei Marcellus
von Bordeaux, Innsbruck 1996, passim).
111
It should be noted, however, that Greek PθÜíáôïò is an adjective of two endings (-ïò
m. f.), and therefore should not have a feminine form in -á. The feminine form atanta
(also in B 3) < *atanata is therefore analogical, possibly the product of Latinization.
112
As to the phonetic representation by z cf. Italian zio, zia „uncle, aunt“ from the
homonym θåsïò, θåßá.
113
Comparable also French la peau, jargon for a prostitute, German (without sexual
connotation) arme Haut.
114
Possibly derti-mo.
62

The proposed restoration of the text, in semantic units, with word separation, is
as follows:
te uoraiimo, eh, za, atanto tehon, zo(a), atanta te,
compriato sosin dertin oipommio atehotisse potea(t).
te priauimo, atanta tehon, te, za, timezo, zia,
te uoraiimo, ape sosio derti(n) imo(n) na demtisse [ueie...?]
ape ci alli carti eti heiont Caticnato
na demtisse Clotucil(l)a.
se demtitiont eti cartaont, Dibona, sosio, deui, pia,
sosio pura, sosio gouisa [at]ehoti[sse],
sosio pura heoti[sse]
sua demta apodunna uolis(s)et.
Paraphrasing interpretation:
“We implore you (oravimus)115, divine one, immortal one of gods,
living one, immortal one, you, that this beloved darling, whom I
fuck, might be left to me.
We pray to you (precavimus)116, immortal one of gods, you, divine
one, I shall honour, we implore you, that this darling of mine may
not be taken away, that any other lovers leave her to Catignatus,
that Clotucilla is not taken away.
If they take her away and befriend her, o Divona (we implore you
that) she, o goddess, remain faithful, pure and joyful, be left pure,
(and that) he (= the other lover) would be willing to give back the
one taken away.”
I am aware that, due to the uncertainties of the readings and the extraordinary
linguistic forms and meanings, my interpretation is subjective as well as
speculative, but it is coherent and makes sense. Thus, it may serve as a basis for
further discussions of this remarkable document, which offers us an invaluable
insight into the linguistic usage of the lower ranks of society, which we would
never get in literary texts.
Leaving now the “magical” inscriptions we move on to texts which are in
various other ways of human interest. Partly they have a socializing function,
partly they are personal or even intimate in character. The first of these texts
which I shall mention still falls into the category of longer texts.

115
The initial u (= v) in uoraiimo seems to be an on-glide.
116
In both cases the sense is rather that of present tense oramus, precamus (class. -mur)
 63

It was also found at Lezoux, and it was on a terracotta plate of which, however,
only a fragment of about one third of its size has survived, so that the greater part
of the text is missing. As some intelligible phrases show, it was moralizing in
character, giving advice to a young person about how to conduct himself properly
in life:
mesamobi molatus certiognu sueticon
“praise by the worst (is) self-damaging to the righteous”117
nu gnate ne dama gussou
“now, my boy, do not yield to violence” (?)
batoron ueia suebreto
“one should go one’s way by one’s own judgement”.

Fig. 41. Fragment of a terracotta plate found at Lezoux,

inscribed with a moralizing text.

As one can judge from even these short phrases, this text, had it survived in
its full length, would have to be regarded as one of the most important and most
interesting of Gaulish texts, as well as for its quasi-philosophical content as for
117
mesamobi is instrumental plur., equatable with Old Irish messam “worst” (superlative
of olc “bad”); molatus = Old Irish molad “praise”.
64

its sophisticated language. As many other texts, also this one shows a certain
admixture of Latin elements, e.g. uero ne curri, ambito (= -u).
For curiosity’s sake it may be mentioned that its first (anonymus) editor
considered this text a cooking recipe (for making some sort of a pasta), apparently
because the repeated occurrence of the lexeme pap- suggested to him a meaning
like “pap” (Latin pappa, etc.) Even L. Fleuriot, its second editor118, although well
aware that pap- was nothing but the pronominal adjective *pāpos “every” (Welsh
pawb), could not free himself from the impression that this text had something to
do with eating (perhaps because it was written on a plate). As a matter of fact, the
text has nothing to do with the nature of the object which is just a suitable carrier,
in the same way as in the modern souvenir industry ceramic plates are inscribed
with sentimental or “funny” texts (often parodies of popular wisdom and morals).

Fig. 42. Graffito on a drinking cup from Banassac (Lozère).

Musée des Antiquitées Nationales, St.-Germain-en-Laye.

118
Études Celtiques 17 (1980), 127–147. Now in RIG II, 2, L-66. A “military context” is
assumed by K. McCone in Meid – Anreiter 1996, 107–117.
 65

On the other hand, drinking vessels may carry inscriptions which do refer to
their special purpose. Noteworthy is the repeated reference to communal drinking
or its socializing effect:119
neddamon delgu linda
“I contain the drinks of the nearest”120
ibetis uciu, andecari biiete
“Drink from this, (and) you will be very amiable”121
ïõåíéêïé ìåδïõ
“Friendly through mead”.122

Fig. 43. Inscription on the foot of a bottle from Limé (Aisne), now lost.

119
Cf. CIL XIII 10016,4 accipe me [si]tie(n)s et trade sodali, on a drinking vessel from
Mainz.
120
Drinking cup from Banassac (Lozère); RIG II.2. L-50. First correctly read and
interpreted by J. Vendryes, Études Celtiques 7 (1955), 9–15 (“proximorum teneo
potus”); neddamon (dd = ðð) cf. Old Irish nessam “nearest”.
121
According to the reading and interpretation by L. Fleuriot, Études Celtiques 18 (1981),
89–93. This inscription on a bottle from Limé (Aisne) was formerly believed to be in
Latin (cf. CIL XIII 10025,188).
122
Goblet from Vallauris (Alpes-Maritimes); RIG I, pp. 414–419 (with different
interpretation).
66

Several graffiti on drinking vessels seem to contain an “advertising” message

suggesting to take pleasure in (and thereby to buy) wines or other potables of a
particular region, such as the already mentioned
lubi rutenica onobíía
“prefer Rutenian aquavit”123
or
lubi caunonnas sincera
“prefer the full-bodied (wines) of Caunonna”124
Here we find the 2nd sing. imperative of the verbal stem lubi- “to love, desire”,
of which also the 2nd sing. subjunctive, lubiias, is attested in a fragmentary
graffito125 where a certain potter Aricanos apparently suggests that “you may
love” his ware: [ari]cani lubiias ... (rest unintelligible).
The same Aricanos, again in the genitive Aricani, figures in the graffito already
cited on p. 23, in which the supposed reading lubitus must be replaced now by
lubitías, apparently a noun depending of Aricani and referring to an order of an
unspecified speciality of his in popular demand: “his fancied (specialities)”.126
aricaní lubitías
ris tecuandoedo
tidres tríanís
The strange word in the second line (so far unexplained) consists, in my
opinion, of the pronoun te (dependent on ris), and the Latin phrase quando edo
“when I take (them) out”.
Among personal objects rings have a special significance because a ring is
more than a mere ornament or object of value: it is above all a binding symbol.
Inscriptions on rings therefore are of an intimately personal character and express
attachment to a partner.
A fine example of this is offered by a gold ring from Thiaucourt (Belgica).127
In the past, owing to wrong word-division, the continuous inscription around this
123
Banassac; RIG II, 2, L-51 (listing all sorts of dubious interpretations as regards the
second part of the inscription, ommitted here; see fn. 49 above). Another graffito from
Banassac, also with lubi, but for the rest unintelligible, is L-53.
124
La Graufesenque; RIG II, 2, L-37. sincera is taken from Latin (sincērus “genuine”),
meaning here, as neuter plur., unblended, undiluted wines.
125
La Graufesenque; RIG II, 2, L-36.
126
La Graufesenque; RIG II, 2, L-35.1; with full bibliography.
127
See L. Fleuriot, Études Celtiques 16 (1979), 123–134 whose interpretation, however,
cannot be accepted. Now in RIG II,2, L-127, but with equally unsatisfactory treatment.
 67

octagonal ring had always been misread and consequently misunderstood. With
correct word-division a very moving text emerges in which the wearer (a woman)
assures her partner that she will never ‘turn away’ from him:
Adiantunne, ni exuertinin appisetu
“Adiantunnos (voc.), (this ring) shall not see a disloyal one”,
literally “one who turns away”. In slight modification of K.H. Schmidt’s analysis
and interpretation128 I take exuertinin to be accusative of a nominative exuertina
(or -inis)129; appisetu is 3rd person sing. imperative of ad-pis- “to see”.130
Finally I should like to draw attention to a curious category of texts all of
which come from Eastern or East-Central France (several from Autun).131 Their
outward characteristic is that they are inscribed on so-called spindle-whorls.132
However, they have nothing to do with the practice of spinning (as J. Whatmough
erroneously thought133) but rather reflect spinning-room amusements. These short
texts are, either in a subtle or in a more direct way, erotically suggestive. They
are, as a rule, addressed to young women; the implied speakers are, of course,
young men. Some of the texts are in straight Latin and as such do not interest us
here, but they may give us an idea of what to expect in the Gaulish texts:
accede, urbana
aue, uale, bella tu
salue, soror
salue tu, puella
aue, domina, sitiio

128
In Le lingue indoeuropee di frammentaria attestazione, ed. E. Vineis (Pisa 1983) 83,
and somewhat differently in Zeitschrift für celtische Philologie 41 (1986), 178, 373
fn. 5.
129
A derivation of a noun *exuertis (or exuerta) “turning away, separation”. The
etymological sense of exuertina (-is) comes out better in German translation: “eine
Abtrünnige”.
130
Cf. Old Irish ad-cí; the simplex pis- occurs in the Chamalières inscription (pissíiumi).
131
I have discussed these in Gallisch oder Lateinisch? (Meid 1980); see also M. Lejeune,
Études Celtiques 15 (1976–78), 96–104 and for the references DAG pp. 359, 495 f.,
499 f. and now RIG II,2, pp. 317–335 (L-111 to 122).
132
A spindle-whorl (French peson de fuseau, German Spinnwirtel) is a ring of a conical
shape, usually made out of schist, which was stuck onto a spindle in order to give it
weight and momentum.
133
Language 25 (1949), 388–391.
68

In the last phrase, sitiio is a good example of the allusive sugestiveness and
ambiguousness of these texts, since the speaker is most likely not thirsting for
beer …
Some texts are in pure Gaulish:
moni, gnatha, gabi buððutton imon134
“Come here, girl, take my little kiss”
tionouimpi morucin135
“Divinely-beautiful maiden”136
The greater part, however, is couched in a kind of mixed language, a colloquial
jargon composed of elements of Latin and Gaulish, the basic grammatical
structure of which is Latin while the vocabulary is to a large extent Gaulish
or is ambiguous in the sense that its etymological roots are identical for both
languages. This vocabulary common to, or similar in both languages may be
considered as one of the reasons why Gaulish gave way so easily to Latin. Our
texts testify an intermediate stage – a Gallo-Latin strongly coloured by the native
Gaulish idiom:137
nata uimpi, curmi da
“Pretty girl, give beer”
geneta, uis, cara
“Dear girl, are you willing?”
taurina uimpi
“Pretty bull-girl”138

134
St.-Réverien (Nièvre); RIG II,2, L-119. See also C. Watkins in Studia Celtica et Indo-
Europea (Archaeolingua vol. 10, Budapest 1999), 541 f. who thinks that buððutton
calls for a stronger meaning and opts for the male organ, comparing Old Irish bot
“penis” (and interpreting also the divine epithet Bussu-mārus in the same sense).
This is taken up by Delamarre 93, but is rejected by Stifter, Zeitschrift für celtische
Philologie 58 (2011), 174, fn. 20.
135
Gièvres (Loir-et-Cher); RIG II,2, L-111.
136
For morucin cf. Welsh morwyn “girl”; tiono- represents *dēvono- (through *dīvono-,
*diono-).
137
All are from Autun; RIG II,2, L-112–115, 117.
138
From a grammatical point of view taurina is nothing else but a motion feminine of
taurinus “young bull”, formed automatically in disregard of semantic inhibitions. This
grammatico-semantic monstrosity is nevertheless erotically suggestive: a heifer fit for
the bull …
 69

marcosior maternia
“I should like to ride …”139
matta dagomota, baline enata
“Silly-girl, good-to-fuck140, engendered by the phallos”
Superficially one may experience doubt as to whether to class these phrases
as Latin or Gaulish, but in fact they are gallicized Latin. Later this Gallo-Latin
replete with Gaulish words was gradually “purified”, to the extent that in modern
French only a small number of words of Gaulish origin have survived.

Fig. 44. Examples of inscribed spindle-whorls from the Museum of Autun.

This ends our survey of Gaulish “literature” which – though it illuminates

certain facets of Gaulish public and private life – is in no way representative of
the material and especially of the spiritual culture of the Gauls. There are two
main reasons for this, representing seemingly unsurmountable obstacles to our
full understanding of the Gaulish language and of what used to be expressed in it.
The first, and principal, reason is that all “druidical” learning and wisdom
(which must have been immense) and all practice connected to it, by religious
interdiction was never committed to writing but used to be only orally transmitted

139
On marcosior cf. M. Peters, also in Studia Celtica et Indogermanica, p. 305.
140
Cf. Old Irish moth “membrum virile”.
70

Fig. 45. Fragment of the Gaulish Calendar found in 1887 near Coligny, showing the
months Equos and Samon(ios). Lyon, Musée de la Civilisation Gallo-Romaine.
 71

to a select class of persons, so that one important source (comparable no doubt

to the total corpus of the Vedic literature of Ancient India) falls out altogether.141
This source, if it existed, would have covered the language fully.
The second reason (which is partly the consequence of the first, but mainly
due to the conquest and colonization of Gaul) is that – in the absence of a tradition
of writing in the Gaulish language – any writing which was in fact done, was
done at first in the Greek, then in the Latin language which, as time progressed,
became the univeral language in Gaul.
Thus any habit of writing in the Gaulish language which had developed in
the meantime came to an end. In view of this unfavourable situation one must be
content with what in fact did survive of Gaulish texts, and make the most of it. The
results of recent excavations which brought a considerable increase in Gaulish
texts let us hope, though, that this constant flux of newly found inscriptions may
continue also in the future, leading to a gradual increase of our knowledge of
Gaulish. Although at present it may seem unrealistic that Gaulish will ever leave
the state of a fragmentarily attested language (a so-called “Trümmersprache”) and
attain the status of a sufficiently documented “corpus” language142, the increase
of new texts will result in the filling of more and more gaps, so that the grammar
and the lexicon of Gaulish will have to be constantly revised and rewritten.
141
The only (fortuitous) exception concerns Astronomy and the Calendar of which a
fragment (the so-called Coligny Calendar) has come down to us – evidence of the
supreme standard of astronomical knowledge. It must be borne in mind, however,
that the bronze plate on which this calendar was engraved was deliberately destroyed
before it was hidden in the earth. The most recent research on the Coligny Calendar
is by Garrett S. Olmsted who has attempted to reconstruct its run over its full period:
Garrett Olmsted, The Gaulish Calendar. A Reconstruction from the Bronze Fragments
from Coligny with an Analysis of its Function as a Highly Accurate Lunar/Solar
Predictor as well as an Explanation of its Terminology and Development. Bonn 1992,
followed by A Definitive Reconstructed Text of the Coligny Calendar, Washington
D.C. 2001 (Journal of Indo-European Studies Monograph vol. 39). These monographs
follow upon the edition contained in Vol. III: Les Calendriers, par P.-M. Duval et
Georges Pinault, Paris 1986, of Recueil des Inscriptions Gauloises. From a linguistic
point of view the Coligny Calendar is important inasmuch as it contains the Gaulish
names of the months apart from other technical vocabulary much of which, however,
occurs in abbreviated form.
142
There is always hope that the unexpected may yet occur, as is shown by the example of
Celtiberian where the sensational discovery at Botorrita in October 1992 of a very long
inscription (several times longer than the one found there in 1970) and subsequently
of several other substantial texts has changed the overall situation of that language, in
spite of the uncertainties remaining, very much for the better.
 References

Pertinent literature is, as a rule, quoted in full in the text or in the footnotes, with
the exception of some works for which the following abbreviations are used:

CIL Corpus Inscriptionum Latinarum

DAG J. Whatmough, The Dialects of Ancient Gaul. Cambridge,
Mass., 1970.
Delamarre 2003 X. Delamarre, Dictionnaire de la langue gauloise. Une
approche linguistique du vieux-celtique continental.
2e edition revue et augmentée. Paris 2003.
Dottin G. Dottin, La langue gauloise. Paris 1920.
IEW J. Pokorny, Indogermanisches etymologisches Wörterbuch.
Bern 1959.
Meid 1980 W. Meid, Gallisch oder Lateinisch? Soziolinguistische
und andere Bemerkungen zu populären gallo-lateinischen
Inschriften. Innsbruck 1980.
Meid 1987 W. Meid, Die Interpretation gallischer Inschriften. Wien
1987.
Meid 1989 W. Meid, Zur Lesung und Deutung gallischer Inschriften.
Innsbruck 1989.
Meid 1991 W. Meid, Aspekte der keltischen und germanischen
Religion im Zeugnis der Sprache. Innsbruck 1991.
Meid 2010 W. Meid, The Celts. Innsbruck 2010.
Meid – Anreiter 1996 W. Meid – P. Anreiter, Die größeren altkeltischen
Sprachdenkmäler. Innsbruck 1996.
RIG I Recueil des Inscriptions Gauloises, Vol. I: Textes gallo-
grecs, par M. Lejeune. Paris 1985.
RIG II,1 Recueil des Inscriptions Gauloises, Vol. II, fasc. 1: Textes
gallo-étrusques. Textes gallo-latins sur pierre, par M.
Lejeune. Paris 1988.
RIG II,2 Recueil des Inscriptions Gauloises, Vol. II, fasc. 2: Textes
gallo-latins sur instrumentum, par P.-Y. Lambert. Paris
2002.
 Edited by
ERZSÉBET JEREM and WOLFGANG MEID

Main Series
2. Stefan Schumacher: Die rätischen Inschriften. Geschichte und heutiger
Stand der Forschung. 1992. 2. vermehrte Auflage 2004. 375 pp. € 38.-. ISBN
963 8046 53 8
8. Man and the Animal World. Studies in Archaeozoology, Archaeology,
Anthropology and Palaeolinguistics in memoriam Sándor Bökönyi.
Edited by Peter Anreiter, László Bartosiewicz, Erzsébet Jerem and Wolfgang
Meid. 1998. 720 pp. € 60.-. ISBN 963 8046 15 5
10. Studia Celtica et Indogermanica. Festschrift für Wolfgang Meid. Heraus-
gegeben von Peter Anreiter und Erzsébet Jerem. 1999. 572 pp. € 64.-. ISBN
963 8046 28 7
18. Raimund Karl: Altkeltische Sozialstrukturen. 2006. 609 pp. € 62.-. ISBN 963
8046 69 4
19. Martin Hannes Graf: Schaf und Ziege im frühgeschichtlichen Mitteleuropa.
Sprach- und kulturgeschichtliche Studien. 2006. 320 pp. € 48.-. ISBN 963
8046 70 8
24. The Master of Animals in Old World Iconography. Edited by Derek B.
Counts and Bettina Arnold. 2010. 256 pp. € 56.-. ISBN 978 963 9911 14 7
25. Archaeological, Cultural and Linguistic Heritage. Festschrift for
Erzsébet Jerem in Honour of her 70th Birthday. Edited by Peter Anreiter,
Eszter Bánffy, László Bartosiewicz, Wolfgang Meid and Carola Metzner-
Nebelsick. 2012. 640 pp. € 76.-. ISBN 978 963 9911 28 4
 Series Minor
1. Wolfgang Meid: Gaulish Inscriptions. Their interpretation in the light
of archaeological evidence and their value as a source of linguistic and
sociological information. 2014. 3rd revised edition. 74 pp. € 20.-. ISBN 978
963 9911 61 1
6. Marija Gimbutas: Das Ende Alteuropas. Der Einfall von Steppennomaden
aus Südrußland und die Indogermanisierung Mitteleuropas. 1994.
2. Aufl. 2000. 135 pp. € 30.-. ISBN 963 8046 09 0
8. Wolfgang Meid: Die keltischen Sprachen und Literaturen. Ein Überblick.
1997. 2. Aufl. 2005. 94 pp. € 18.-. ISBN 963 8046 65 1
12. Francisco Marco Simón: Die Religion im keltischen Hispanien. 1998.
168 pp. € 26.-. ISBN 963 8046 24 4
13. Peter Raulwing: Horses, Chariots and Indo-Europeans. Problems of
Chariotry Research from the Viewpoint of Indo-European Linguistics.
2000. 210 pp. € 34.-. ISBN 963 8046 26 0
16. Peter Anreiter: Die vorrömischen Namen Pannoniens. 2001. 316 pp. € 30.-.
ISBN 963 8046 39 2
17. Paul Gaechter: Die Gedächtniskultur in Irland. 2003. 116 pp. € 20.-. ISBN
963 8046 45 7
19. Karin Stüber: Schmied und Frau. Studien zur gallischen Epigraphik und
Onomastik. 2004. 125 pp. € 22.-. ISBN 963 8046 55 4
20. Wolfgang Meid: Keltische Personennamen in Pannonien. 2005. 350 pp.
€ 30.- ISBN 963 8046 56 2
31. Archaeological Imaginations of Religion. Edited by Thomas Meier and
Petra Tillessen. 2014. 404 pp. € 38.-. ISBN 978 963 9911 24 6
34. Alexander Falileyev: In Search of the Eastern Celts. Studies in
Geographical Names, their Distribution and Morphology. 2014. 173 pp.
€ 28.-. ISBN 978 963 9911 56 7

Please address orders to:

ARCHAEOLINGUA
H-1250 Budapest, Pf. 41.
Fax: (+361) 3758939
e-mail: kovacsr@archaeolingua.hu
www.archaeolingua.hu www.hungarianarchaeology.hu
 INNSBRUCKER BEITRÄGE ZUR KULTURWISSENSCHAFT
INNSBRUCKER BEITRÄGE ZUR SPRACHWISSENSCHAFT
Publications in Indo-European and Celtic Studies
A Greek Man in the Iberian Street. Papers in Linguistics and Epigraphy in Honour
of Javier de Hoz. 2011. 434 pp. € 88.00. ISBN 978-3-85124-726-8. (IBS 140)
Michael Janda: Eleusis. Das indogermanische Erbe der Mysterien. 2000. 351 S.
€ 64.00. ISBN 978-3-85124-675-9. (IBS 96)
Michael Janda: Elysion. Entstehung und Entwicklung der griechischen Religion.
2005. 427 S. € 80.00. ISBN 978-3-85124-702-2. (IBS 119)
Michael Janda: Die Musik nach dem Chaos. Der Schöpfungsmythos der europäischen
Vorzeit. 2010. 412 S. € 64.00. ISBN 978-3-85124-227-0. (IBK NF 1)
Michael Janda: Purpurnes Meer. Sprache und Kultur der homerischen Welt. 2014.
727 S. € 96.00. ISBN 978-3-85124-233-1. (IBK NF 7)
Michael Janda: Morgenröte über Mykene. Ein indogermanischer Mythos. 2014. 167 S.
€ 36.00. ISBN 978-3-85124-235-5. (IBK NF 9)
Kim McCone: The Indo-European Origins of the Old Irish Nasal Presents, Subjunctives
and Futures. 1991. 210 S. € 36.00. ISBN 978-3-85124-617-9. (IBS 66)
Wolfgang Meid: Die erste Botorrita-Inschrift. Interpretation eines keltiberischen
Sprachdenkmals. 1993. 140 S. € 32.00. ISBN 978-3-85124-639-1. (IBS 76)
Wolfgang Meid: Kleinere keltiberische Sprachdenkmäler. 1996. 61 S., mit Abbild.
€ 12.00. ISBN 978-3-85124-656-8. (IBS V 64)
Wolfgang Meid: Die Romanze von Froech und Findabair. Táin Bó Froích. Altirischer
Text mit Einleitung, deutscher Übersetzung, ausführlichem philologisch-linguistischem
Kommentar und Glossar. 2., neubearbeitete Auflage. 2009. 274 S. € 32.00. ISBN 978-3-
85124-226-3. (IBK SH 1030)
Wolfgang Meid: The Celts. 2010. 182 pp. € 24.00. ISBN 978-3-85124-228-7. (IBK NF 2)
Wolfgang Meid: Ausgewählte Schriften zum Indogermanischen, Keltischen und
Germanischen. 2012. 448 S. € 80.00. ISBN 978-3-85124-730-5. (IBS 144)
Angelo Mercado: Italic Verse. A Study of the Poetic Remains of Old Latin, Faliscan, and
Sabellic. 2012. 464 pp. € 80.00. ISBN 978-3-85124-731-2. (IBS 145)
Bestellungen werden erbeten an:
Universität Innsbruck
Institut für Sprachen und Literaturen – Bereich Sprachwissenschaft
A-6020 Innsbruck, Innrain 52
Telefax: (+43-512) 507-2837 E-Mail: wolfgang.meid@uibk.ac.at