B) Celtic epithets: general approach

The Mother Goddesses are thus known with a variety of epithets. As we are going to see, it must be borne in mind that many of them are not Celtic but Germanic, while others are undeniably Latin, such as Parcae, Fatae, Junones, Domesticae or Campestres. This is the reason why Christoph Rüger’s analysis of the Mothers, in his article entitled ‘Beobachtungen zu den epigraphischen Belegen der Muttergottheiten in den lateinischen Provinzen des Imperium Romanum’ [‘Notes on the Inscriptions dedicated to the Mother Goddesses in the Latin Provinces of the Roman Empire’], published in Matronen und verwandte Gottheiten [‘Matronae and related Goddesses’] in 1987, is beyond the scope of this study.173 Indeed, he does not distinguish the Latin and Germanic names from the Celtic epithets. Moreover, his list seems to be drawn principally according to the Latin divine epithets rather than the Celtic and Germanic ones. All attributive byname groups taken into account, he classifies the cult of the Mothers into eight categories: geographic deities, for instance the Ambioreneses, Montes and Campestres ; roadway deities, for instance the Triviae and Quadruviae; spring deities, for instance the Nymphae, Fontes and Suleviae; animal deities, for instance the Cervae and Gantunae; ancestral deities, for instance Proxsumae and Veteres;174 tutelary deities, for instance, Dominae and Virgines; functional deities, for instance Nutrices, Parcae and Medicinae ; and deities whose names are the plural form of a single deity, such as Cereres, Junones and Dianae.175

Classifying deities according to the meaning of their epithets, as did Rüger, is problematic, for the ambiguity of etymological evidence means that a byname can have various etymologies and refer to diverse attributions. Those different possible significations cultivate ambiguity. It also appeals to the supernatural and to the mystic, mysterious and complex nature of deities, who are multi-faceted, for they have the ability to possess various kinds of functions and attributes within a single personality. Furthermore, it is clear that a certain amount of attributive bynames* still remain obscure or hypothetical to the scholar. It is nonetheless possible to establish a broad outline, keeping in mind that an epithet can fall into several categories.

Despite these difficulties of classification, it is clear that some of the goddess bynames are geographic or toponymic*. This means that they refer to a place which they personify and protect. The Matres Glanicae for instance are clearly ‘the Mothers of Glanum’ and the Matres Nemausicae, ‘The Mothers of Nîmes’ (see above). In addition, epithets can be ethnonymic*, referring to names of tribes. This means that either the sept* took its name from the goddess they believed in, or they gave their ethnic name to the goddess they held in high respect and esteem. In any case, the goddess is eponymous of the tribe she represents, nourishes and protects. Such is the case of the Matres Treverae, venerated in Birten (Germany), who are ‘the Mothers of the Treveri’,176 and the Matronae Vediantiae, honoured in Cimiez (Nice, Alpes-Maritimes), who are ‘the Mothers of the Vediantii’. 177

Others seem to refer to natural elements, such as the Matronae Dervonnae (‘Mothers of the Oak’) in Milano and Brescia (Italy),178 or the Matres Eburnicae (‘Mothers of the Yew’) in Yvours-sur-le-Rhône (Rhône).179 Finally, some bynames* tend to be descriptive of what the goddesses incarnate or the functions and attributes they fulfill. For instance, the Matres Mogontiones, venerated in Agonès (Hérault), must have embodied ‘Youth’, on account of the meaning of their appelation.180 As for the Matronae Lubicae, honouredin Cologne, they might have personified ‘Love’ as well as ‘Affection’ given to the people, for their name is possibly derived from the root *lub-, *lob- ‘to like’, ‘to love’. The verbal forms lubi, lubiias and lubitias, ‘love’, ‘that you love’ and ‘loved’ are attested on various inscriptions from Gaul (< IE *leubh-, ‘to love’, ‘to desire’), and Lubos and Lubus are common male proper names in Celti-Iberia.181 Schmidt and Delamarre propose to gloss their name as ‘The Loving, Affectionate Mothers’ or ‘The Endearing Mothers’.182 Similarly, the Rocloisiabo, ‘the Listening Goddesses’, honoured in Saint-Rémy-de-Provence, have the ability of listening to the prayers of their pilgrims.183

The Matres Eburnicae is a good example of the issue of categorization. Their epithet relates to a natural element - the yew tree – and might have come from an ethnonym*, for it refers to the tribe of the Eburones, but this hypothesis implies a linguistic transformation Eburonikā > Eburnikā (see Chapters 2 and 3). Moreover, they might have been the personification of some intoxicating cults attached to yew, as will be shown in Chapter 5.

Notes
173.

Rüger, 1987, pp. 1-30.

174.

Despite their Latin name which means ‘Kinswomen’, the Proxsumae do not seem to be Roman goddesses. Their cult is attested in Gaul only, which tends to prove that they were Celtic goddesses in origin (see below). As for the Veteres, they are not Roman either, for their name is also written Vheteres in the inscriptions, which proves that their name is not related to the Latin adjective vetus meaning ‘old’. Thus, Veteres does not mean ‘the Ancient’. It is actually a Germanic name designating the ram.

175.

Rüger, 1987, pp. 2-3.

176.

CIL XIII, 8634.

177.

CIL V, 7872, 7873 ; Barruol, 1999, p. 366. See Chapters 3 and 5 for more details about those goddesses.

178.

CIL V, 5791, 4208.

179.

There may be a homonymy between the place-name Yvours (*Eburnicum, ‘Place planted with Yew Trees’) and the divine epithet Eburnicae. CIL XIII, 1765 ; Vendryes, 1997, p. 46. The inscription was discovered near the wall of the garden of the Castle of Yvourt, near Lyon. It had been re-used* in the wall of the castle.

180.

AE 1986, 471. See the section on Mogontia in Chapter 4 for more details.

181.

CIL XIII, 8220 ; Delamarre, 2003, p. 209 indicates that the word ‘love’ is unknown in insular Celtic.

182.

Schmidt, 1987, pp. 143, 149 ; Delamarre, 2007, pp. 120, 225.

183.

RIG I, n° G-65. See Chapter 5 for more details.