- The Earliest Evidence
- Roman Era and Later Evidence
- Lug: A Pan-Celtic Deity?
- Female Deities
The apparent lack of religious unity across the Celtic world can distress those who expect some degree of consistency: the symbolism of Iron Age artifacts seems to bear little resemblance to the much richer and more plentiful evidence from the Roman era, and these materials often sit uneasily alongside texts produced by the medieval Celts of Britain and Ireland. A recent tally of divine Celtic names from inscriptions carved within the bounds of the Roman Empire alone stands at about 800, and this does not include the theonyms which survive in medieval Goidelic and Brythonic texts. Assembling a simple, consistent and universal pantheon of Celtic gods and goddesses is an impossible task: not only are there major disparities across time periods and regions, but thinking in terms of anthropomorphic deities – humanoid beings which personify certain functions, qualities or concerns – limits our understanding of religious phenomena too much. Instead, we must be open to all sorts of manifestations of power in all sorts of forms.
How do we account for and synthesize this motley collection of characters and material? Can we detect continuity in Celtic communities across space and time, or periods of significant disruption and reintegration in the conceptualization of superhuman beings and forces? Was there any conceptualization of the divine common to all Celts, or were there major divisions and differences across regions?
The Earliest Evidence
The Indo-European root *dyeu means “to shine” and the Indo-European word for “god” *deiw-os is derived from this by association with the bright sky. These roots can be seen in English words such as “deity” and “divine.” What is interesting, however, is that the Celtic conception of the location of the divine seems to have moved away from the Indo-European focus on the heavens and towards the earth, even though the ancestral word for “god” was inherited.
La Tène art gives us the impression that no firm boundaries were believed to exist between the sacred powers, humankind, plants and animals. We see stylized human faces and animal bodies emerging from foliage and animals with human heads on stone sculptures and metalwork. Sometimes these representations are realistic, but often they are surreal or abstract. Do such images represent deities, divine ancestors, scenarios from myths, powers embedded in the objects themselves, or something else? Given the cultural preoccupations of the Iron Age Celts, we might expect there to have been religious significance invested in metalwork, mining, and warfare. On the continent and in southern Britain, we would also expect to see religious symbolism associated with the divine patronage and protection of mercantile trade and urban development.
Mirrored leaves on either side of male heads, interpreted by some as double mistletoe leaves, begin to appear on stone carvings in northern Gaul shortly before the La Tène period (and similar symmetrical serpentine patterns appear on many other objects, sometimes in a more abstracted manner, throughout La Tène art). Do these male figures and leaf-shapes around their heads have any special significance? Do the faces represent men or gods?
The statues from Glauberg (see left), dated to c.400 BCE, provide an interesting case study. All of the items depicted as worn by the one intact statue (fragments of three others remain) – shield, sword, and torc – were actually found as artifacts in one of the tombs at the site. A leaf-shaped wireframe was found in one of the tombs, possibly once covered in leather, which is the only evidence of such leaf-shaped head-gear yet found. It is most likely, then, that the statues depict the people buried in the mound who must have had some special status for the community who wanted to keep their memory, and symbolic presence, alive.
The leaf shapes around male heads may represent the physical manifestation of their invisible powers or the projection of their sacred essence. Could this be related to the descriptions of heroes in early Irish tales, who are described as having a lúan láith “warrior’s light” emanating from their foreheads when they go into battle mode? An alternative name for this enigmatic effect in the Irish tales is én gaile “bird of war-fury,” and it is not hard to see these protuberances as wings. Very similar motifs appear in Greek, Latin and Iranian tales, suggesting that this is an ancient Indo-European concept.
The hero is conceptualized in many cultures as mediators between humanity and the gods, often having divine parentage and bringing special benefits to their communities, especially at times of crisis. Some of the iconography and ritual activity of the Celtic Iron Age has been interpreted as a manifestation of cults of ancestors and/or heroes which helped to strengthen the identity and cohesion of growing communities. In southern Gaul in particular, conspicuous cult sites, often including human heads (real or symbolic) and seated human figures, were built in the public spaces of proto-urban settlements (see right). The heads displayed on pillars or depicted in sculpture may be those of ancestors – guardian spirits – whose stature increased with each passing generation. This is not unlike the case of ancient Greece, where city-states cultivated hero cults to facilitate identification with the polity. Similar structures and practices are visible at some sacred sites in southern Britain in the late Iron Age and Roman period, where high-status burials can form the focus for ritual activity.
Similarly, many Iron Age oppida of northwest Iberia featured stone statues that “represent idealized warrior aristocrats (heroes) that symbolically protected the entrance to the settlements, while at the same time making explicit the power of the ruling elite.” (See the statue from Lezenho hillfort, Vila Real, Portugal, above.) There are rudimentary echoes of the cult of the hero-patron-protector in medieval Irish and Welsh tales, some of which depict the burial of heroes at strategic locations where they can continue to defend the interests of their community even after death.
Rather than indicating continuity, however, the creation and promotion of these hero-ancestor cults may indicate social stress and communal crisis: the transition from kin-based to urban, state-based society, the encroaching power of mediterranean empires, and the conflicts between rival Celtic dynasties competing for control over territories. Over time, the identity of these hero-ancestors could have been replaced by other deities; such parochial cults may help to account for the great diversity of Celtic theonyms.
Some sites seem to have been felt to contain a special power or quality which made them so sacred as to be the abode of deities or perhaps deities themselves. There are a great many Celtic theonyms formed from toponyms (e.g., names of god(desse)s based on names of places) and qualifying epithets based on toponyms. Most survive in Roman-era inscriptions but it is possible that some predate Roman conquest; this is suggested by similar patterns among the Insular Celts who personified the essence and sovereignty of entire territories in goddesses, such as Anu, the divine manifestation of the province of Munster in Ireland. Although many of these deities were considered female, especially those associated with waters (see below), some of them were considered male.
La Tène-era Celtic art contains many representations of animals which seem to have some religious significance. Animals are often represented simply as animals, such as ravens or boars on the armor and weaponry of warriors, probably to gain some of their power. There are also human-animal hybrids, such as the cross-legged figure from Late Iron Age Gaul who has hooves for feet (see left). Does this depict an intermediary between the human and animal worlds who has characteristics of both? Is this a human who has been reincarnated as an animal? Is this a shaman-priest who can inhabit the bodies of animals for certain purposes? Although we have no way to recover the original intentions of such art, it is evocative of the motif of shape-shifting so common in Insular Celtic literature.
In an animate universe, even objects can be endowed with spiritual power and function as a kind of being, especially if the blacksmith was understood as a kind of magic worker. Mirrored serpentine figures, called “dragon pairs” by many Celtic art historians, appear on swords and scabbards as early as the mid-5th century and recur with variation throughout La Tène material culture. Sometimes these dragon figures are traced in gold. Did they represent the power of the sword, or invoke some special protection on its owner? In early Irish literature, weapons were said to be alive: they sometimes spoke or fought on their own accord.
Roman Era and Later Evidence
One of the major problems with the Roman sources which mention Celtic gods is that they usually equate them with particular Greek or Roman gods, thus coloring information by filtering it through Roman ways of thinking. The main Greek and Roman gods named by authors such as Caesar are:
- Apollo: Greek and Roman god of the sun, light, poetry, prophecy and healing; leader of the Muses and defender of livestock.
- Jupiter: Supreme Roman god (father of divine family); god of the sky and thunder; associated with the oak tree.
- Mars: Roman god of war; divine ancestor of the Romans; birth of Mars coincided with new year.
- Mercury: Roman god of commerce, travelers, luck, and eloquence; escort of the dead to the Otherworld; his Greek equivalent, Hermes, was also the god of oaths and contracts.
Very few Celtic gods had their native names recorded in Greek or Roman documents. We have already seen Ogmios in Lucian’s text Heracles and the Irish god with a variant of this name, Ogma, associated with the ogam writing script. Ogmios is also mentioned on two tablets containing magical spells found at Bregenz, high in the Alps of western Austria. He seems to have the role of a conveyer of the message to be given to other divinities in these spells, but there is no further evidence of his cult.
Three Celtic gods are mentioned in Pharsalia by Lucan which names Teutates, Esus, and Taranis. The root of the name Teutates is *teutā “nation, tribe,” suggesting that he was the sacred patron of the interests and concerns of the nation. His name occurs as an epithet in connection with Roman gods on several inscriptions: one with Apollo, three with Jupiter, two with Mars, and two with Mercury.
The name Esus may be derived from the term *aisus “divine,” simply indicating a deity. One of the few appearances of his name is on the Pilier des Nautes (“Pillar of the Boatmen,” shown above), which was carved for a Gallo-Roman temple in the civitas which later became Paris and dedicated to Tiberius Caesar Augustus, who became emperor in 14 CE. The panel with which his name is associated depicts him cutting down a willow tree with an axe. A statue of Mercury from Lezoux in central France also seems to equate him with Esus, and he may be mentioned in an inscription from Noricum (in central Europe). He also seems to be invoked in a magic spell in Gaulish.
The name Taranis is clearly and directly derived from the word for thunder (and therefore closely related to the Scandinavian thunder-god Thor). There is a Gallo-Greek dedication to him at Orgon near the ancient colony of Massalia from the 1st century BCE and a probable Latin dedication to him at Amien. There are a number of cult sites in southern France commemorating where lightning struck the ground, which was believed to impregnate it, and these sites have a strong geographical correlation to Jupiter altars which feature wheel symbolism. These two symbols are united in a bronze statue from Châtelet in Romanized Gaul. Silius Italicus Punica (§3.344) implies that the druids read divine messages from lightning strikes. Taranis is among several gods named in a magical text found at a 2nd-century Gallo-Roman temple written in a mixture of Gaulish and Latin.
The god Belenus is mentioned in a late-4th-century Roman text as a god that had been worshipped by the druids of western France; he is also claimed by Tertullian (†225), an early Christian scholar, to have been the main god of the people of Noricum (corresponding to much of modern Austria). The root Bel appears in Old Brythonic personal names as well, such as the historically-attested Cunobelinos, king of the Britons, and the mythical Welsh figure Beli Mawr. The root Bel may mean “bright” and may also form the first part of the name of the festival of Beltain, which has strong associations with fire. There are some 51 inscriptions to Belenus (or Belinos) across the Celtic world from as early as the 1st century BCE and he is usually paired with Apollo in Romanized dedications.
What are we to make of this “fertile chaos” of names and images, many of which only appear less than a handful of times? It is not likely that the names, functions and characteristics of the Iron Age Celtic deities resembled those of the Roman world very much, but as Roman religious norms became increasingly prominent and the primary framework for defining and identifying divinities, localized understandings of Celtic gods would have increasingly diverged from one another. The cult of Teutates may been equated to Mars in one place and to Jupiter in another, as there was no longer any pan-Celtic religious institution to connect and standardize such practices. Not all of the diversity of Celtic theonyms can be explained as the distortion of Iron Age traditions by Romanization: there is also evidence that new cults and theonyms were being actively created and promoted in the Roman era. Although we should expect some degree of continuity from earlier periods, the hybrid forms that emerged during the era of Roman rule may have been very different from those of the Iron Age.
Another way of thinking about the evidence is that rather than having been a coherent group of beings with specific and fixed names and personalities throughout the Celtic world, there may instead have been a set of common roles and functions which was interpreted according to the local traditions and lore of each community. As Proinsias Mac Cana has argued,
this incoherence simply reflects the decentralised structure of Celtic society, in which each tribe functioned as an independent political unit, the inference being that political autonomy was coupled with religious autonomy and that each tribe had its own special gods, which might, or might not, be common to neighbouring tribes. (Celtic Mythology, 18)
Some inscriptions and early magical texts give the name of a Celtic nation as an epithet along with the name of the god, as if to underscore the divinity’s allegiance to the group as their patron. The Chamalières curse tablet, for example, written in Gaulish not long after Caesar’s conquest of Gaul, invokes Maponus, known elsewhere in the Celtic world, but calls him “the Arvernian.” Similarly, the god Mars was Celticized and “localized” by a Celtic nation called the “Caturiges” by giving him the epithet “Caturix.”
The concept of a locally-rooted tribal god, such as suggested by the epithet Teutates, accords well with the idea of the local hero-ancestor cult discussed above. Some Celtic gods of Iberia were identified not by the place at which they were worshipped but by the kin-group who worshipped them. The idea of the tribal god seems to be echoed by a formulaic oath which occurs in many early Irish tales: tongu do dia toinges mo thúath “I swear by the god my people swear by.” There are further variations of this in the Irish tales, such as “I swear by the god the Ulster-people swear by” and just “I swear by the gods,” and it has been suggested that this may be an Indo-European oral formula in origin. Could this be the reason that Strabo (Geography § 3.4.16) was under the impression that the Celtiberians venerated a nameless god?
Lug: a Pan-Celtic Deity?
The challenges of reconstructing pagan Celtic religion can be illustrated by the attempts to locate the cult of the god Lug in Celtic Europe. From the literature of the medieval Insular Celts, one would expect Lug to be a major pan-Celtic figure: Irish literature calls him a god and represents him as a model of kingship, the consort of the divine female personification of the landscape, and a master of all arts and crafts.
On the basis of the last characteristic, he has often been assumed to be the god equated by Caesar (§ 6.17) with Mercury, whom he says was the most important deity of the Gauls and who was the inventor of all of the arts. The Gaelic god Lug (one of whose epithets is “of the Long Hand”) corresponds directly to the Welsh hero Lleu of the Skillful Hand. His name forms the basis of the Gaelic harvest festival of Lugnasad, held around the beginning of August, some celebrations being held on hilltops.
There are several interpretations of the meaning of the god’s name. The conventional analysis has been that his name was derived from a root meaning “bright, light; to shine,” and indeed he is described as bright and associated with bright objects in some Irish tales. The root of his name is also related to a term used in the swearing of oaths, leading to the suggestion that he was the “god of the oath,” invoked implicitly when bonds were given. The evidence about him, and contemporary perceptions of him in post-Roman Europe, may have been complicated by other words that resembled his name, such as the raven (Gaulish lugos). The root Lug appears in Celtiberian, Gaulish, Old Gaelic and Old Welsh personal names.
Given his importance in Insular Celtic tradition it is strange, then, that there is so little direct evidence of him in continental Europe. His name does not seem to appear among any of the surviving early magical texts, although the words luge and luxe, close homonyms interpreted by most linguists as referring to oaths, do appear. Only the Celtiberian rock inscription on the hilltop of Peñalba de Villastar seems to invoke him as a singular god (although there are alternative interpretations of the inscription). There are about ten inscriptions in Celtic Iberia and one at Avenches, Switzerland, the ancient capital of the Helvetii (shown on the left), to a set of divinities – the Lugoves (e.g., taking a multiple rather than singular manifestation) – whose gender has been argued to be feminine rather than masculine. Is this the same god as the Insular Lug, a later development of a regional cult based on the same root concept but in multiple, female form, or a completely different deity?
There are many place names around Europe, often associated with oppida and seats of power, which originally took the form Lugudūnon “Fortress of Lug.” While a dedication to a god of kingship or craftsmanship seems like a logical origin for such names, we don’t know for certain exactly which meaning of Lug/lug was intended by the people who originally named these places.
The difficulties in interpreting such evidence are well illustrated by in the case of Lugudūnum (modern Lyon), Gaul, founded in 12 BCE. In a section about the names of rivers and mountains in his treatise The Morals, the Greek author Plutarch (†120 CE) offers us a legend (of uncertain origin) about the origins of the city:
Near to the river Arar stands a mountain called Lugdunum, which was given its name because of the following event. When Momorus and Atepomarus were dethroned by Seseroneus, in fulfillment of the oracle’s command they decided to build a city upon the top of the hill. But when they had laid the foundations, great numbers of crows with their outstretched wings covered all the neighboring trees. When this happened Momorus, being a person well skilled in augury, called the city Lugdunum, as lugdon in their language [i.e., Gaulish] signifies a crow, and dunum signifies any spacious hill.
The Roman province of Gallia Lugdunensis was named for Lugudūnum, which became its capital. The union of numerous Gaulish nations under Roman rule was symbolized by an altar at the capital which, Strabo (Geography § 4.3) informs us, bore “an inscription of the names of the sixty Gaulish nations, and also images from these nations, one from each.” In time, Lugudūnum became one of the most important cities of the Roman west, being the birthplace of two Roman Emperors. A festival to the Emperor Augustus was held there on the first of August (a Roman month named for him).
What are the arguments for and against the hypothesis that the capital of Lugudūnum and the cult of Augustus was modeled on the cult of Lug? What purpose might have using such a pre-existing set of symbols and beliefs had in the Romanization of Gaul?
Consider the distribution of the following ethnonyms based on the element lug. Do these confirm a pan-Celtic cult to the god Lug, or are there other ways to interpret the evidence?
- Lugii in east-central Europe
- Lugi in north-eastern Scotland
- Lougei and Luggoni in north-west Spain
- Luigni in Meath and Sligo, Ireland
Consider the distribution of the following place names, all originally containing the elements equivalent to Lugudūnum (names for which there is no ancient record are italicized). Do these confirm a pan-Celtic cult of the god Lug or are there shortcomings to this hypothesis? Are there other ways to interpret the evidence?
- Lyon, France
- Dinlleu, Gwynedd, Wales
- Dinlle, Shropshire, England
- Carlisle, England
- Lothian, Scotland
- St-Bertrand de Comminges, Garonne, France
- St-Lizier, Ariège, France
- Montlahue, Drôme, France
- Leiden, The Netherlands
- Laon, Aisne, France
- Loudon, Sarthe, France
- Lion-en-Sullias, Loiret, France
- Laons, Eure-et-Loire, France
- Laudun, Gard, France
- Lauzun, Lot-et-Garonne, France
- Monlezun, Gers, France
- Montlauzun, Lot, France
- Loudun, Vienne, France
The discussion thus far has concentrated on male deities: what about goddesses, which feature so prominently in the literature of the medieval Insular Celts? These female figures, often appearing in triplicate, personify the landscape, confer legitimacy upon kings, prophecy the future, and bestow life, death, victory, defeat, and fertility. Is there evidence to demonstrate that this is a continuity of older conceptions of sacred feminine principles?
The names of goddesses are closely tied to territory and features of the landscape in Celtic regions and the names of many rivers in Celtic regions imply female deities. Many of the major rivers of Gaul, for example, had associated goddesses with similar names, such as Sequana, the deity of the Seine; the Marne, a tributary of the Seine, is named for the Mātr-on-ā “Great Matron.” Similarly, the river Boyne in Ireland is personified in medieval Irish literature as the female consort of the male deity the Dagda (although her name itself, Boand, means “White Cow”). The matriarch of the pagan Irish pantheon in medieval literature, Danu/Donu, shares her name with an Indic aquatic goddess, the root of the name of the river Danube (and many others), and the Welsh mother-figure Dôn. Rivers bring moisture and fertility and actually shape the landscape over time; they carry a power and force that must be respected and is often feared. According to Sextus Propertius (The Elegies § 4.10.39-44) the Gaulish chief Virdomarus claimed to be born from the river Rhine, a way of boosting his status by claiming a divine mother.
It is striking that while Caesar lists several Gaulish gods (under Roman names), he mentions only one goddess, whom he equates with Minerva. There are some noteworthy correspondences between Minerva and some of the Insular Celtic goddesses, such as the Irish goddess Brigit. The name “Brigit” means “the high, elevated one” and may originally have been more of a generic title for a goddess than a specific individual character. Unlike inscriptions to most of the Romano-Celtic gods, Minerva’s name is seldom given an epithet in Latin dedications, as though her name was itself a title or epithet. Like Brigit, Minerva was also associated with poetry, medicine and artisans (among other things).
Names or titles derived from the same root as Brigit appear in several other parts of the Celtic world: the goddess Brigantia appears in seven inscriptions in Britain, mostly in the north, one pairing her with Jupiter; the female deity name Bregissa appears on a magical text (paired with Branderix “Raven-king”) from Le Mas-Marcou, France, written in first-century Gaulish; a dedication to the Matres Brigiacae was carved at the Peñalbo de Castro of Spain; and so on.
There is another specific female deity who appears in many Romano-Celtic inscriptions: Juno Regina, often depicted on top of Jupiter columns in Gaul with her husband Jupiter. Juno Regina was the divine protector of the nation and had military overtones, being the mother of Mars. Despite the Roman tradition of using many epithets in invoking her, she is often just referred to as Regina “Queen” or dea Regina (“Queen goddess”) in Romano-Celtic tradition. She is even paired with Minerva on some carvings.
There are many dedications to divine females represented as triplets under such names as Matres (“Mothers”) and Matronae (“Matrons”), sometimes depicted with symbols of fertility (especially the cornucopia), such as the altar to the Matres at the Gaulish capital of Vertillum (see left). Although these dedications to female goddesses do not begin to appear until the 1st C BCE and then only in Gallia Narbonensis, this dissimilarity to the Roman tendency to make multiple female deities into a single one suggests that these triplets are a continuity of Iron Age Celtic beliefs. Similar triple goddesses can be found in the inscriptions and iconography of Celtic Iberia and are very common in medieval Irish literature.
Recent work on an ancient sanctuary in Noricum on the banks of the river Sava has shown that a local cult to two Celtic female water deities, Savus and Adsalluta, were displaced by the incoming cult of the Magna Mater “Great Mother” during the Roman era. The Celtic subjects of Rome, like many others in the empire, seem to have found the transition from female deities personifying local aspects of nature to the universal Great Mother a fairly easy one.
The Great Mother, worshipped under different names, was known to all peoples and was known also to the Celts, although among them the cult of various Mother Goddesses is usually documented in the plural. The evidence shows that the cult of the Great Mother soon also became popular in the Celtic provinces, and that its popularity could have in some cases eventually outshone local Celtic goddesses.
We may also learn something about the nature of female Gallo-Roman deities from iconography depicting divine couples, especially in eastern Gaul: it is significant that in many cases a Roman god is paired with an indigenous goddess, a pattern not derived from Roman practice. This coupling seems to symbolize Roman rule (the male leader) over Celtic territory (the female personification of landscape), a theme fully developed in medieval Insular literature.
Borsje, “Omens, Ordeals and Oracles.”
Campanile, “Meaning and prehistory.”
Enright, “Fires of knowledge.”
Ginoux, “Pendragon’s Ancestors.”
Haeussler, “From Tomb to Temple.”
— “How to identify.”
Koch, Celtic Culture.
Kruta, The Celts.
Mac Cana, Celtic Mythology.
MacLeod, “Mater Deorum.”
Mees, Celtic Curses.
Ó hÓgáin, The Sacred Isle.
Šašel Kos, “Adsalluta and Magna Mater.”
Simón, “The Cult of Lugoves.”
— “Religious Practices.”
Sopeña, “Celtiberian Ideologies.”
Watkins, “Some Celtic Phrasal Echoes.”