1 Celtic Religion

Our sources for Celtic religion are sharply divided between ancient Greek and Roman authors, whose testimony is contemporaneous (starting from the first century BCE) but indirect and concerning only Gaul and Britain, and medieval Irish writings, whose material is the original mythology but heavily euhemerized by its Christian monkish authors. So in the Irish sources what were originally gods and goddesses must be detected behind stories of heroes and demons, and the priests are presented as sorcerers, whereas in the Latin sources the gods are given Roman names, assimilating them to the nearest Roman counterpart—an early instance of interpretatio romana, which was later extended to the Germanic pantheon by Tacitus and others. Hence a challenge that has faced modern scholars is to match these Roman names to the characters of the Irish myths. Help has been sought in the iconographic and inscriptional evidence unearthed by archaeologists, which is completely absent in Ireland (which was never conquered by the Romans) but plentiful in Gaul; here, however, it is heavily Romanized and so of uncertain reliability, compounding the problem. In any case, the principal reason of interest for us is that the Celts possessed a highly developed priestly class, the druids, even though we know precious little about their operations. In what follows, we focus on the continental Gauls and the Irish, because the British Celts are only occasionally mentioned by the Roman sources, left no archaeological traces, and after the Romans’ withdrawal were swamped by the Germanic Anglo-Saxon tribes which only left a record.

1.1 Theology

The clearest concise description of the classic continental Celts’ pantheon was written around 50 BCE by Julius Caesar in his Gallic War (VI, 17). It is worth reproducing in full:

Among the gods, they most worship Mercury. There are numerous images of him; they declare him the inventor of all arts, the guide for every road and journey, and they deem him to have the greatest influence on money-making and commerce. After him they set Apollo, Mars, Jupiter, and Minerva. Of these deities they have almost the same idea as all other nations: Apollo drives away diseases, Minerva teaches the first principles of arts and crafts, Jupiter holds the empire of heaven, Mars directs wars. To Mars, when they have determined to engage battle, they vow as a rule whatever spoils they may take: if then they win, they sacrifice such living things as they have taken and all the other effects they gather into one place.

The main problem of interpretation here concerns not so much the deciphering of the Gauls’ deities underlying these Roman names as the correspondence between Gaulish and Irish deities in the structure of a pantheon. In what follows we will essentially follow Le Roux (1970) as our guide, adding citations to other scholars when necessary to add to or depart from it; at the end of the section we will try to assess the results of our survey against alternative interpretations.

The plethora of votive inscriptions and figures found in Gaul, which is much more abundant than for any other native deity, confirm Caesar’s leading role of Mercury in the pantheon. The native Gaulish god that lies beneath the Roman Mercury can be identified as Lugus, which provided the names of many towns throughout Celtic Europe (including today’s Lyon). In turn, Lugus is the counterpart of Irish Lug, the king of the Irish pantheon, a god of the bright sky (the etymology is “shining, luminous”), at the same time a priest, a universal sovereign, a warrior who fights with his magic, and the inspirer and protector of arts, thus encompassing in himself all the three functions. This is a range of jurisdiction that goes well beyond the competencies of the Roman Mercury or of his Greek counterpart, Hermes, and is more comparable to the Germanic Odin/Wodan, who was also identified with Mercury in the Roman period—see Sect. 3.2.1 below (De Vries 1963, 62; Davidson 1988, 89–92). To him was consecrated one of the four major Celtic festivals, Lughnasad or “Lug’s assembly” (August 1), marking the beginning of harvest (see Davidson 1988, 38–39, for the Celtic festivals).

The king of the Roman gods, Jupiter, comes only in fourth place in Caesar’s list. His main Gaulish name is Taranis, or “the thundering one” (cf. the Romans’ Jupiter Tonans). A god of the heavens, of all the natural elements, as Caesar suggests, he is often portrayed with a wheel or with the thunderbolt in Gallo-Roman iconography—both cosmic symbols. His Irish counterpart is the Dagda, the “Good God”, “which means potent and gifted rather than benevolent” (Davidson 1988, 204). He wields a mighty club, corresponding to Jupiter’s thunderbolt, which can equally give death or life. He holds a huge cauldron to which crowds are always welcome to come and eat and whose contents never give out; he is himself a big eater and drinker, a great lover, a druid, and a musician. He is the patron of boundaries, oaths, contracts, law, and friendship, and overall is comparable to Thor (Davidson 1988, 204–207). One of his cognomina corresponds to the Dis Pater (a Roman god of the underworld, later identified with the Greek Hades) which, according to Caesar (1917, VI, 18), all Gauls claim as their common ancestor. Another aspect of Jupiter appears in Ireland as Manannan, whose relation to the Dagda is the same as that of Neptune to Jupiter: in the mythology, the lord of the sid, the Other World, and hence euhemerized as a lord of the sea because only from the sea can the sid be accessed.

Mars comes third in Caesar’s list. The god of war, obviously important to a people that was engaged in internecine warfare all the time, as Caesar points out elsewhere (VI, 15), he receives the spoils of battle as offerings, as Caesar’s passage above explains in detail. His counterpart in Irish epic and legend is twofold: on the one hand the “king” Nuadu and on the other hand the “champion” Ogma, lord of sacred writing (ogam) and also of brute force and warlike furor. The latter’s continental correspondent, Ogmios, leads the souls to the Other World. So Ogma-Ogmios represents the dark side of sovereignty of which the Dagda represents the luminous side. The Gallo-Roman Mars is described in the inscriptions by epithets that are both warlike and regal. So the Nuadu-Ogma pair is outside the classic Roman framework and points to a more archaic Indo-European structure, a dual conception of sovereignty: as Caesar says, Mars directs the war but does not actually fight it, while the warrior/hero does the fighting, since among the Celts (like the Germans) war was a collection of individual contests. This also explains the confusion of epithets and attributes between Mars and Mercury in Gaul.

Apollo, second in Caesar’s list, is presented only as a healer, to which the Irish correspondent is Dian Cecht. But other Irish legends and Gallo-Roman inscriptions indicate that the Celtic Apollo was seen more broadly as a solar, youthful, elevated god, in keeping with his classic eponym. One of his most widespread epithets, or hypostases, was Belenus, “shining”, whence the name of one of the four great Irish festivals, Beltene (May 1), marking the beginning of summer. Another epithet was Borvo or Bormo, “boiling”, which gave many thermal centers their name to many thermal centers. So this god does not belong to the third function; medicine is practiced by the druids and so falls within the first (priestly) function.

Patroness of arts and crafts, Minerva, the last in Caesar’s list, is an exalted female deity called by name-epithets like Belisama (“the shining one”) and Brigantia (“the most high”). Her Irish cognate is Brigit, daughter of the Dagda (like Minerva is daughter of Jupiter), worshiped by poets, smiths, and physicians. In the legend, under her other name of Dana, she is the mother of the gods, whose company is called Tuatha Dé Danann or “the tribe of Dana”. Seen sometimes as a single deity, sometimes as a triple one, she is reflected or personified by countless specific female deities, which together comprise a Great Goddess, encompassing the three functions. Unfortunately she was effaced from the written texts because Christianization turned her into an enormously powerful saint, St Brigit, patroness of Ireland together with St Patrick. She presided over one of the four great Irish festivals, Imbolg or “purification” (February 1). A technical homologue of Brigit is the smith god Goibniu, close to the Roman Vulcan, of whose worship there is ample epigraphic record in Gaul. De Vries (1963, 122–149) lays great stress on the Celtic goddesses, from the Matres or Matronae and Epona worshiped in Gaul to the Irish personifications of Mother Earth.

Summarizing her survey, Le Roux (1970) thinks that the “rules of polytheism” that preside over the classic pantheon do not fit the Celtic system. Caesar did his best but could not understand a flexible, fluid system in place of the rigorous classification of Roman religion. Here the main god, Lug, presides over the dyad Dagda-Ogma, joining spiritual and legal authority with warlike force and magic; then over the triad Dagda-Ogma-Nuadu, druid-warrior-king; then there are some artisans (a healer, a smith); and finally the primeval goddess. The greatest difference from other Indo-European religions is the total absence among the Celts of a great third-function deity of fertility like Venus or Freyja. On the other hand Davidson (1988, 214–215) sees parallel pairs of ruling gods, Lug and the Dagda, Odin and Thor, comprising the first and second functions and with identities shifting in the course of time, and opposed to a third function group comprised mainly of goddesses. De Vries (1963, 62, 154–164) also stresses the similarity between Lug and Odin but then, when he tries to extract a well-defined tri-functional pantheon of a few great gods, corresponding to Caesar’s Romanized structure, from the mythical story of two great battles between tribes of gods and demons preserved in Irish texts (the Two Battles of Mag Tuired), his findings are hard to square with Le Roux’s account given above.

In De Vries’ reconstruction (ibid., 162–163), the Dagda is the supreme god, then there is a triad Nuadu, Lug, Ogma (notice the difference with Le Roux’s), whereby the first, sovereign function bifurcates into “two aspects, one priestly and juridical, the other magical and kingly”, represented here by Nuadu and Lug, corresponding to the Germanic Tyr and Odin respectively—which leaves one to wonder in what sense the Dagda is supreme. The second, warlike function is represented by Ogma, corresponding to the Germanic Thor (whom, however, we introduced above as corresponding to the Dagda). Finally there are seemingly technical, third-function gods, like the smith Goibniu and the physician Dian Cecht, already mentioned; these, however, really belong in the priestly aspect of the first function because the smith was believed to possess magical powers while medicine was in the hands of the druids. So it is not clear who truly represents the third function in this pantheon. The purpose of walking the reader through this slog is only to show that, perhaps due to the cryptic, confusing state of the euhemerized Irish material, even between sworn Dumézilians like De Vries and Le Roux there are contradictions and shifts of interpretation, which—Littleton’s (1973, 167–174) ecumenical review notwithstanding—make a clean summary of Celtic theology anything but straightforward.

1.2 The Cult and the Priests

All the ancient authors were impressed by the learning and authority of the Celtic priestly class, which included the druids, the bards or poets, and the seers/diviners. In the Irish texts, probably because the monks who wrote down the ancient legends were converted bards, the last two groups are conflated into the single figure of the file (poet-seer). The druids were properly the priests: they conducted the sacrifices, held the courts of justice, advised the kings, were teachers and seekers of high learning. They were men; the so-called Celtic “priestesses”, important as they were, belonged in the third group, as seeresses, prophetesses, and fortune-tellers, since apparently they never practiced sacrifice or teaching.

Caesar, partly relying on earlier Greek writers, devotes two full chapters to the druids (VI, 13–14).

13. (The druids) take part in the cult, take care of public and private sacrifices, and regulate the religious observances. (….) they decide almost all disputes, public and private, and if any crime has occurred, or murder has been committed, or there is any dispute about succession or boundaries, they also decide it, determining rewards and penalties; if any person or group does not abide by their decision, they ban them from the sacrifices, which is their heaviest penalty. (….) But the one who has the highest authority among them rules over all the druids. At his death, either the one who is outstanding succeeds, or, if there are several of equal standing, they contend for the primacy by the vote of the druids, or sometimes even by the force of arms. At a certain time of the year, the druids meet within the borders of the Carnutes, whose territory is reckoned as the center of all Gaul, and sit in council in a consecrated spot. All those who have disputes convene here from everywhere, and they obey the decisions and judgments of the druids. (….)

14. The druids usually hold aloof from war and do not pay taxes like everyone else; they are excused from military service and exempt from all liabilities. Tempted by these great rewards, many young men assemble of their own motion to receive their instruction (disciplina), or are sent by parents and relatives. It is reported that there they learn by heart a great number of verses, and therefore some remain twenty years under instruction. And they do not think it proper (fas) to commit these utterances to writing, although in almost all other matters (….) they make use of Greek letters. (….) The first principle which they seek to teach is that souls do not die, but after death pass from one to another; and this belief, as the fear of death is thereby cast aside, they hold to be the greatest incentive to valor. Besides this, they ponder many questions regarding the stars and their movement, the size of the universe and of the earth, the order of nature, the authority and powers of the immortal gods, and hand down their lore to the young men.

This passage is largely self-explanatory, and the Irish legends confirm its main points, including the druids’ key role in educating the warrior nobility (those whom Caesar calls the knights—equites—i.e. warriors who could supply and mount a horse), in pursuing and transmitting higher learning, in practicing the magic arts, and in adjudicating disputes among tribes and kings at their annual gatherings. The kings issued from the warrior class but the druids, while keeping separate, controlled the choice of kings and their performance. One of their fundamental tasks was the blood sacrifice, about which we know very little because the Irish Christian records obliterated all traces of it. Like with the Germans, these rituals probably took place in a forest clearing or grove (Gaulish nemeton) or other consecrated open space, since no temples have been found in Ireland and those in Gaul are so obviously a Roman influence.

Elsewhere (VI, 16) Caesar says that the Gauls practice human sacrifices in the event of serious illness or great danger, and vow to sacrifice their own life in a pitched battle (parallel to the Roman devotio), and that the druids minister such practices. Human sacrifice among the Celts was widely known in antiquity and divided the ancient writers, some of whom, near the end of the first century CE, including Pliny the Elder, Tacitus and Suetonius, stigmatized it as savage and cruel and on this basis endorsed the decrees of Emperors Tiberius and Claudius which proscribed druidic practices and specifically suppressed their schools.Footnote 1 It is not clear how frequent or widespread such sacrifices were, as archeology has found hardly any unambiguous evidence of it (as opposed to the execution of criminals or of prisoners of war, both of whom the ancient sources claim were offered as sacrifice to the gods)—an ambiguity that extends to the analogous problem with Germans and Vikings. In particular, one is not sure what to make of the horrific description by Caesar (VI, 16; confirmed by Strabo IV, 4, 5) of huge, upright wickerwork figures that were filled with living men and set on fire—something practically not easy to do, as remarked by Davidson (1988, 60). In any case, the druids are clearly stated to be essential participants in these sacrifices, as could hardly be otherwise.

Based on both the Irish and the classical sources, Le Roux (1970) and Le Roux and Guyonvarc’h (1986, Chap. 5, III) claim that the immortality of the soul was believed for all men, who are revived in a distant Other World (sid) of bliss, known by various names, located in some islands beyond the sea to the far north and west of the known world. But one may doubt this. It seems more likely that such life beyond death was the preserve of the aristocracy, as in the Germanic, Vedic, and Iranian religions, since Caesar (VI, 13) states that the commoners (plebs) were held in the same regard as slaves, and in the passage quoted above (VI, 14) he reports that the druids believed that the doctrine of another life after death worked as “the greatest incentive to valor”, so it was something for warriors. On the other hand, in contrast with the Norse, in either Celtic mythology as we have it or in ancient reports on their beliefs there are only faint traces of eschatology or a conception of the end of the world, possibly hidden in the Irish account of the Second Battle of Mag Tuired, mentioned above (Davidson 1988, 188–189, 193–94).

Caesar’s description of the druids sets the Celts apart from their contemporary German neighbors, whose priestly class, as we will see in the next section, was in an infant stage of development and did not grow to maturity with their Norse descendants. It is the best evidence against the claim, often made by scholars on the basis of late Gallo-Roman archeological evidence, that the Celts’ religion was all made of local deities and tribal cults: such an organized and unified priesthood must have insured a unified theology (De Vries 1963, 154–157, 212–222; Le Roux and Guyonvarc’h 1986). The only apt comparison is the Vedic Brahmans, with their long-lasting instruction within hearing distance from the teachers and rote memorization of verses, their providing a high priest to support and advise the king (purohita), and their doubling as scholars and masters of religious ceremonies. Even the Brahmans, however, did not have any regular meetings nor a chief priest overseeing the whole class. Like the Brahmans, the druids may have been specialized by cultic function (although this is not attested) but there is no mention of specialization to the service of particular gods.

2 Germanic Religion

Like with Celtic religion, our access to Germanic religion too is characterized by a great chasm in time, space, and nature of the sources. Our earliest sources are Greek and Roman authors of the classical era, prominent among them the Roman historian Tacitus who wrote Germania, a detailed account of the continental Germans at the end of the first century CE. To the Romans, the Germans were the tribes living east of the Rhine to the Vistula and including Denmark, Frisia and the southern tip of the Scandinavian peninsula. These continental Germans moved west in the so-called Age of Migration (fourth-sixth centuries CE) and overrun the Western Roman Empire, where they finally settled and were soon Christianized. They left no written records of their own, so our evidence for the postclassical period is limited to some archaeological and inscriptional evidence.

By contrast, the Scandinavian branch expanded vigorously in the northern seas and lands in the Viking Age (eighth-eleventh centuries) and converted to Christianity much later than their continental cousins, between the tenth and the eleventh century. The Vikings bequeathed to us a rich literature that has a bearing on mythology and religion, including a collection of poems, known as the Elder Edda or Poetic Edda, some of which composed in pre-Christian times; a rich body of heroic sagas and family sagas composed in Iceland in the thirteenth or fourteenth centuries but containing some earlier material; a history of Denmark written in Latin in the early thirteenth century by the Danish scholar Saxo Grammaticus; and above all the Prose Edda, an invaluable account of Norse mythology derived from early poetry, sagas and oral tradition and written in Old Norse by the gifted Icelandic scholar Snorri Sturluson in the early thirteenth century (see Turville-Petre 1964, Chap. 1, for an extended description of these sources). For the period prior to the Viking Age, the only evidence is some inscriptions in the runic characters, archaeological finds, and place names. Thus the Norse religion is documented from inside sources dated more than 1,000 years after the outside sources that provide us with an account of the continental Germans’ religion. Any suggestion that the two sets of accounts capture different stages of evolution of the same religion must, however, take into account the vast differences in environmental and historical conditions between the two societies.

2.1 Theology

For the early continental Germans, the classic place to start is Tacitus’ (undated, 9) account:

Among the gods, they most worship Mercury, to whom on certain days they deem lawful to sacrifice even human victims. Hercules and Mars they appease with the animals usually allotted for sacrifice. Some of the Suebi also sacrifice to Isis; what was the cause and origin of this foreign rite I could not discover, except that her representation in the image of a galley suggests an imported worship. Otherwise, the Germans consider it unworthy of the greatness of celestial beings to confine the gods within walls or to portray them in any form of human likeness: they consecrate woods and groves, and apply names of gods to the hidden powers that they only see with the eye of reverence.

In another chapter (40) Tacitus describes a ritual practiced by a group of Suebi tribes, near Denmark, who worship a goddess called Nerthus, whom he identifies with Mother Earth. She dwells in a sacred grove on an island in the Ocean, where there is a consecrated chariot, draped with cloth, which none but a priest may handle. He escorts the goddess in her chariot, drawn by cows, on a tour around the country, and the people rejoice and welcome her. They lay down their weapons, and then and only then do they enjoy peace, until she is satisfied of human company and the priest brings her back to her sanctuary. Then the chariot, the cloth, and the deity herself are washed in a hidden lake by slaves, who are then immediately swallowed up by the lake itself.

So these Germans worshiped their deities in sacred groves, had no built temples, scorned anthropomorphic images, had priests at least for some rituals, knew public festivals, and occasionally practiced human sacrifice. Tacitus’ description of the goddess Nerthus’ cult is paralleled by stories of deities parading in wagons in Icelandic sagas and by archaeological finds of ceremonial wagons found in peat bogs in Denmark around Tacitus’ time and in ship burials of the Viking period (Davidson 1964, 92–96).

The names of the days of the week provide a clue to the identity of the Germanic gods underlying the Roman gods chosen by Tacitus as their closest approximation. Mars seems to have been Germanic Tîwaz, Scandinavian Tyr, hence Tuesday (dies Martis) (Davidson 1988, 208). Mercury was identified by the Romans with Odin’s Germanic predecessor Wodan, hence Wednesday, (dies Mercurii) (ibid, 90). The Germans in the Roman armies equated their god Donar, the Scandinavian Thor, hence Thursday, with Hercules (Davidson 1964, 82, based on Tacitus’ chap. 3), but this came to replace the Roman dies Jovis, Jupiter’s day, which points to a shift of divine identities that we will find in Viking times (below). The equation of Wodan/Odin to Mercury may at first seems surprising, until we recall that the latter was the god of trade, the patron of wisdom and learning, the god who was carried by his winged sandals over land and sea, and the guide of souls to the underworld—which points to the original character of Odin as a god of the dead (Davidson 1964, 140–141; De Vries 1970).

Dumézil (1959, 40) interprets Tacitus’ list as foreshadowing, in the expected hierarchical order, the later Scandinavian, tri-functional structure: Mercury as Odin, Hercules and Mars as Thor and Tyr respectively, and “Isis” as a goddess of fertility—replaced or joined, in some other place in northern Germany, by Nerthus, described above as a goddess of peace and prosperity. This structure shows permanence. The so-called Old Saxon Baptismal Vow—a formula that the Saxons converted by Charlemagne were asked to recite, dated to the ninth century—asks the convert to specifically forsake a triad of gods: Thunar (i.e. Donar), Woden (i.e. Wodan), and Saxnot. The latter (also attested in Saxon England as Seaxneat) probably simply means “companion of the Saxons”, and is thus equivalent to the Roman Quirinus and to the Scandinavian Freyr (see below) as a representative of the common people in their totality (ibid., 41).

For the terminal stage of the Scandinavian religion, we have a description of the proceedings at a great temple at Uppsala in the eleventh century, when apparently the traditional religion was still holding sway in Sweden. The description is due to Adam of Bremen, a German cleric contemporary with the events who wrote a history in Latin of the diocese of Hamburg and of the Christian missions to the Nordic lands. It reads (Adam of Bremen 1876, IV, 26–27):

In that temple, all coated with gold, the people worship the statues of three gods; Thor, the most powerful, has a seat in the middle with Wodan and Fricco to his sides. These gods have the following meaning: Thor, they say, rules the atmosphere which controls thunder and lightning, winds and rains, the fair weather and the crops. Then Wodan, which means fury (furor), conducts war and grants men valor against their enemies. The third is Fricco, who bestows peace and pleasure on mortals, and whose idol is provided with a big phallus. They actually portray Wodan in arms, as [our pagans] do with Mars, while Thor with a scepter seems to represent Jupiter. (….) They have priests appointed to all their gods, who offer sacrifices on behalf of the people. If disease or famine is looming, they make offerings to Thor’s idol; if war threatens, to Wodan; and if a wedding is to be celebrated, to Fricco.

So unlike the West Germans of Tacitus’ time, these Swedes do represent their gods in human image and have a temple served by priests. Fricco, that is, Freyr here takes the place of Isis, Nerthus, or Saxnot of the previous examples as patron of the third function. The triad (or tetrad) does indeed appear remarkably stable over time, but note that the functional attributes shift: in Adam’s account Donar/Thor is a sky and thunder god who benefits agriculture and health and who at the same time is likened to Jupiter (no longer to Hercules), while Wodan/Odin is the god of war and is assimilated to Mars (no longer to Mercury).

To account for this shift, which makes the Germanic triad Odin, Thor, Freyr at variance with the parallel triads embodying the tripartite theology—such as the Romans’ pre-Capitoline triad Jupiter, Mars, Quirinus and the Vedic triad Mitra/Varuna, Indra, Ashvins – Dumézil (1958, 57–58; 1959, Chap. 2) argues that in the ideology and practice of the Germans warfare “colored” everything, including their theology, as already noted by Caesar (1917, VI, 21) in an early sketch. The consequence of this ubiquitous, hypertrophic preoccupation with war is twofold. On the one hand Tyr—the counterpart of Mitra/Mithra, the juridical side of sovereignty—has almost lost his original place in the first function and, identified with Mars, is already in Roman times (by Tacitus, Chap. 9 above) associated with Hercules/Thor al the level of the second function. On the other hand Odin/Wodan becomes the god of war in the Viking age (for Adam of Bremen, above), even though in the Norse mythology he does not fight with arms but with spells and binding, as a magician would. And, as a consequence of this last shift, Thor in the Viking age is, as it were, displaced from the second function and becomes a fertility god, which tends to move him to the level of the third function. On this level we find a divine company called the Vanir (Davidson 1964, Chap. 4; De Vries 1970) who, in Norse mythology, were once separated and hostile to the great gods, called Aesir, but then happily joined them in one great pantheon. The principal Vanir deities include Njord and his son Freyr and daughter Freyja (whose names—meaning “Lord” and “Lady” respectively—are really titles, not proper names); scholars have recognized that Njord is the same deity as Tacitus’ goddess Nerthus, discussed above, except that for some reason he is now a male god. This group together, in various ways, takes care of economic production, health, reproduction, love, and pleasures.

Dumézil (1959, Chap. 4) is at pain to stress that it is only through the happy outcome of the atmospheric battle, the rain, achieved by wielding his great hammer—the counterpart of Jupiter’s and Indra’s thunderbolt—that Thor benefits agriculture, and not through any inherent power over the fertility of land and livestock, in the same way as by wielding his hammer he protects the gods from the giants and other cosmic enemies; so he does not really belong in the third function and cannot be confused with the Vanir. This, however, is more of a fine point of comparative theology, which changes nothing in the facts of actual cult. In the Viking age, there is no doubt that Thor (Davidson 1964, Chap. 3; De Vries 1970) was widely invoked and worshiped as the protector and “friend” of the common people, despite—or rather, owing to—his defining mythological feature: an overwhelming, colossal display of physical force, taken to almost comical effect in the myths. He was seen as a beneficent power, close at hand, who could be counted on to right wrongs; this makes it difficult to see Thor linked with the warrior class, while his upholding of rights moves him closer to the sovereign function (Davidson 1988, 200–201).

In the Viking age, the god of the warriors was Odin (Davidson 1964, Chaps. 2, 6; De Vries 1970) who, helped by the battle-goddesses, the Valkiries, chose the cream of the fallen warriors and heroes and welcomed them into his heavenly abode, the Valhalla, where they would forever enjoy a life of dueling and feasting while awaiting the end of days, the Ragnarok (discussed below); the other people were consigned to a murky realm of the dead. This warlike function expands the Varuna-type sovereign function of Odin, father and king of the gods and god of chiefs and kings, inspirer and patron of poetry and letters, great magician (with shamanistic features) and god of (some of) the dead, and knower of hidden things past, present and future—in myth, he sacrificed one of his eyes to gain access to the source of all knowledge. Being a god of the royalty and the elite, and a disquieting one—he sometimes claims human sacrifices, in the Icelandic sagas already in Tacitus (above)—it is not surprising that patronymics containing Odin’s name are virtually nonexistent and that place names containing his name are rare in continental Scandinavia and completely absent in Iceland (an egalitarian republic of free emigrant farmers). The common people, farmers and sailors, would more naturally name their children and their sites for deities closer to their concerns: the beneficent storm god (Thor), the god of winds and sailing (Njord), the god and the goddess of earth, animal, and human fertility and procreation (Freyr and Freyja) (see Turville-Petre 1964, Chaps. 2, 3, 7, for the evidence of personal names and place names). In the same way, Indo-Iranian peoples have personal names containing the names of Mitra and Indra but not that of Varuna (Dumézil 1959, 64–66).

Tacitus’ tetrad had shrunk to a triad by the time of Adam of Bremen: Tîwaz/Tyr, assimilated to Mars in Roman times, seems to have lost importance in Viking times. His identification with Mars seems strange as he represented the juridical, contractual, peaceful side of sovereignty, as attested by inscriptions dedicated to Mars Thingsus—that is, a Mars presiding over the thing, the political assembly of free men. The rationale seems to be that for the Germans warfare was not an exhibit of brutal violence and unrestrained slaughter, but was ruled by legal norms and proceedings, to the point that a battle between two armies could be replaced by a judicial duel—just like with the Romans (De Vries 1970). Tyr had certainly been an ancient Indo-European god, as attested by the etymology of his name, which is cognate to Vedic Dyaus, Greek Zeus, Roman Jupiter: a god of the bright sky. But he then receded into the background, and in Snorri’s Edda he is only a shadowy figure, although one which men prayed to for victory; the only surviving myth about him was the binding of the monster wolf Fenrir, the deadly enemy of the gods, a feat for which he sacrificed his hand. He seems to have ceded his role as battle god to Odin and his role as patron of the law to Thor (Davidson 1988, 208; 1964, 56–61).

These seem to have been the deities who received a lively cult. There are some other members of the Aesir group or associated with them: one is Heimdall, the guardian and watcher of Asgard, the heavenly abode of the gods, residing at the border between the worlds, who will sound his horn to arouse the gods to the final battle—perhaps a god of beginnings like Janus in Dumézil’s view (Davidson 1988, 211); another is Balder, the dying god, son of Odin. But there is no evidence of cult associated with either, and it is not even clear if Balder was a god or a deified hero (Davidson 1964, 176, 183). Still less could there be a cult of Loki the trickster, the deceiver, a malign being that will take the lead of the forces of evil at the end of days. All three will have a role to play at Ragnarok. Some other gods and goddesses are little more than names. So in the final period of the traditional religion we have a small company of gods, parallel to the Celtic one—Odin, Thor, the Vanir—which shows no sign of having swollen since our earliest testimonies; if anything, it had somewhat shrunk, with ancient gods like Tyr and perhaps Heimdall having receded into the background. Furthermore, the shifting of functions and consequent allegiances between gods, discussed above, seems to have implied the idea that people could “choose” their personal patron god or goddess (typically Thor or some of the Vanir) as an almost all-purpose protector—an important point to which we will return (Davidson 1988, 214–215, 220–222).

Ragnarok (“Twilight of the gods”) is the Old Norse name for doomsday, the time of the cataclysmic destruction of the gods and the world, described in the great poem Voluspá and recounted by Snorri with some variation (Davidson 1964, 35–38, 202–210; 1988, 188–195). As the story goes, hatred, disloyalty, and warfare will spread among men, followed by a terrible winter of three years of bitter cold with no more warming up as the wolf Fenrir swallows the sun and the stars fall from the sky. Then all the bound monsters break loose and, together with the giants of frost, mount an assault on Asgard. Heimdall, the watcher, arouses the gods to fighting, and Odin, his company of heroic warriors (who had been chosen to live a glorious afterlife in the Valhalla for this purpose), and the other gods come out to face the evil host in a final battle. There all the gods go down fighting in single combat, and the monsters and giants are destroyed with them. In the process, the sea rises to engulf the land and fire spreads over it to destroy the human race. Yet this is not the end. Earth will rise again from the waves, green as never before, cleansed of all suffering and evil and lighted by a new sun; it will be peopled by a new race of men begotten by a surviving couple. The sons of the old gods survive the great battle to reign peacefully on this new earth.

Despite some possible echoes of Christian apocalyptic teaching in some of the Norse authors, this eschatology seems authentic: it has especially close parallels with the Zoroastrian prophecy of the world’s end (see Sect. 6.2.1 below) and some similarity with what survives of an Irish Celtic myth of a great battle between gods and demons (see Sect. 3.1.2 above). Indeed, scholars of comparative mythology have brought out from the eschatological myths of various Indo-European peoples the contours of a proto-myth of cosmic catastrophe and renewal, centered on an epic final battle, even though its remnants have lost their place in practiced religion and found a niche in legendary history or epic literature in the Indian, Greek and Roman cases (the battle of Kurukshetra in the Mahabharata, the Titanomachy in Hesiod’s Theogony, and the battle of Lake Regillus in Livy’s history of Rome) (see the summary in Bray 2000 and the discussion of cosmic death and resurrection in Lincoln 1986, Chap. 6).

2.2 The Cult and the Priests

The Germans “have no druids to preside over things divine”—such was Caesar’s (1917, VI, 21) terse remark, emphasizing the great difference with the Gauls. As we have seen above, though, in the following century Tacitus knew of a priest presiding over a festival of the goddess Nerthus. In another chapter, Tacitus (undated, 43) briefly mentions an ancient cult paid by a remote German tribe to a pair of young divine brothers, called Alcis, whom he equates with the Roman twins Castor and Pollux (the counterparts of the Greek Dioskouroi, “Zeus’ boys”); the cult was performed in a consecrated grove and was presided over by a priest, “dressed in woman’s attire”. Thus the evidence is mixed for Roman times.

Both Caesar and Tacitus, as well as the Greek geographer Strabo, emphasize the importance of divination among the Germans, as among the Celts of Gaul and of Britain, and the role of highly respected professional seeresses, especially as regards decisions about going to war or engaging battle. Foreknowledge and omens could be obtained in various ways, from sacrificed animals and humans, to observation of natural phenomena and wild animals or birds, to interpretation of dreams, and this continued to be the case later among the Scandinavians and the Irish (Davidson 1988, Chap. 5). In the Viking age the seeress (volva) practiced a special kind of witchcraft called seidh, where she fell into a state of trance and then answered questions about hidden things; these women often traveled about the land in groups and were especially associated with the cult of Freyja (Davidson 1964, 117–120).

To get a sense of the cult practiced in the Viking age, it is best to start with a report. In his Saga of Hakon the Good, a king of Norway in the tenth century, Snorri (2016, Chap. 14) gives the following description of a sacrificial feast:

Jarl Sigurd maintained all the ritual banquets on behalf of the king there in Trondheim country. It was an ancient custom, when a ritual feast was to take place, that all the farmers should attend where the temple was and bring there their own supplies for them to use while the banquet lasted. At this banquet everyone had to take part in the ale-drinking. All kinds of domestic animals were slaughtered there, including horses, and all the blood that came from them was then called hlaut (‘lot’) (….); with [it] the altars were to be reddened all over, and also the walls of the temple outside and inside and the people also were sprinkled, while the meat was to be cooked for a feast. There would be fires down the middle of the floor in the temple with cauldrons over them. The toasts were handed across the fire, and the one who was holding the banquet and who was the chief person there, he had then to dedicate the toast and all the ritual food; first would be Odin’s toast—that was drunk to victory and to the power of the king—and then Njord’s toast and Freyr’s toast for prosperity and peace. Then after that it was common for many people to drink the bragafull (‘chieftain’s toast’). People also drank toasts to their kinsmen, those who had been buried in mounds, and these were called minni (‘memorial toasts’). Jarl Sigurd was the most liberal of men. He did something that was very celebrated: he held a great ritual feast at Lade and stood all the expenses.

So here we have a temple, a feast requiring universal attendance, a local nobleman sponsoring it and acting as a priest, a group of gods honored which is at variance with the tri-functional triad (Odin the god of war and royalty, two Vanir gods as fertility powers, but not Thor unlike in the Uppsala triad), the slaughter of cattle, horses, and more, a full sharing of the sacrificial meat among the participants, and an emphasis on drinking.

Unlike the Germans of Roman times, the Scandinavians of Viking times did have built temples, even though worship in open spaces or sacred woods or groves continued to be important (Turville-Petre 1964, Chap. 12; Davidson 1988, 31–35). The temples (usually called hof) were typically wooden structures, hence subject to decay, but nevertheless archeologists have found remains spread all over continental Scandinavia as well as Iceland. Often—in Iceland particularly—these structures played the double role of sites of religious rituals and of secular gatherings or community festivals. Often, too, the smallest structures were probably family shrines devoted to private cult—which was also very common and did not require the presence of a priest.

Sacrifice (Davidson 1988, 41–45, 49–58; Turville-Petre 1964, Chap. 13) does not seem to have involved a burnt offering to the gods, in contrast to Greek and Roman practice. The spilling and sprinkling of the blood was seen as well-wishing and participants ate all the meat, accompanied by generous helpings of beer or mead hallowed to the gods—Odin was said to have an unlimited supply of mead to welcome his heroes in the Valhalla. The chief sacrificial animals were the boar, the bull or ox, and the horse—the first especially dedicated to Freyr and the Vanir, the second to both the Vanir and Thor, the last to Odin. The horse sacrifice also figured prominently in the ship burials of chiefs and kings or queens. A sacrificial feast (blot) could be held communally in the hall of chieftain’s farm or in a temple, like Jarl Sigurd’s above, or in private homes.

There were three seasonal festivals, lasting several days each, communally enjoyed: one at the beginning of summer (April) when men sacrificed to Odin for good luck and victory in the upcoming season of raids and warfare; another at the beginning of winter (October) when they sacrificed to Freyr for plenty; and a mid-winter festival for the growth of crops, also given to Freyr, which later was made to coincide with Christmas (Davidson 1988, 39–40). Other feasts were held at longer intervals or on special occasions such as times of danger or as thanksgiving for victory or at the funeral of a king, when human victims as well as animals were sacrificed (Davidson 1988, 58–68). This was already widely reported by Greek and Roman writers for the ancient Germans, and Adam of Bremen (1876, IV, 27) gives a gruesome description of a festival held at the Uppsala temple every nine years, in which for nine days men, horses, dogs, and other animals were slaughtered daily and left hanging and putrefying on trees nearby, while the gods were satisfied with their blood—so no communal meal as in the usual blot.

In the Viking age there is hardly any evidence of professional priests (Turville-Petre 1964, Chap. 13; Davidson 1988, 157–158). In continental Scandinavia, priestly functions, including conducting public rituals and sacrifices and presiding over religious festivals, were part of the responsibilities and privileges of chieftains, aristocrats, and kings, alongside political leadership and judicial functions. In Snorri’s account above, a jarl (“earl”) maintained all the ritual banquets “on behalf of the king”, dedicated the toasts and the sacrificial meat to the gods, and paid for one great feast, and this seems to be the general pattern. In newly colonized Iceland—a republic of free farmers—each district was ruled by a godhi, a title that originally must have meant priest (it derives from godh = god) but which then came to designate a chieftain with a retinue of followers; alongside his political and judicial powers, he was in charge of maintaining the building (hof) which was used for both religious and profane meetings and of presiding over the ritual feasts there.

It is possible that with the strengthening of royal power, near the end of the pagan period, priestly institutions began developing around the court, after the pattern of other religions. The Uppsala temple described by Adam of Bremen (above) had priests appointed to the gods and conducting the sacrifices, and this was not just for an occasional or periodic festival but apparently a permanent institution. In the seventh century, when King Edwin of Northumbria called an assembly to decide on whether the kingdom should convert to Christianity, a “high priest”, named Coifi, was in attendance and supported the assembly’s decision, proceeding thereupon to destroy the temple. He is likely to have been a priest of Odin—which is in character with his being a royal chaplain, as Odin was the god of royalty—since he destroyed the temple by first hurling a spear to it and then setting it on fire (both practices in accordance with the sacrificial rites associated with Odin) (Davidson 1964, 50–51; 1988, 31, 157–158). But these signs of institutionalization of priestly office were soon undercut by Christianization. So overall, despite the huge differences in social structure, the prevailing Scandinavian pattern, where priestly functions were not the preserve of an independent profession but were joined to secular office, leadership status, and/or property ownership, reveals an unexpected analogy with the Greek and Roman patterns.

3 Main Takeaways

It is hard to provide a clear picture of Celtic theology because for the Gauls we have an early Romanized description of the pantheon while for the Irish we have late Christianized accounts, hidden under legends, by medieval monks. Caesar gives a neat picture of five deities under Roman names, each with a specialized function, but from the inscriptions in Gaul and the myths in Ireland these deities appear to correspond to multiple Celtic counterparts, so that the roster grows and the specialization becomes blurred. Understandably, scholars are divided about the structure of the pantheon. One attractive, though not unanimous, interpretation sees a broad correspondence between the two major Irish gods, Lug and the Dagda, and the Scandinavian Odin and Thor, though with attributes shifting and expanding in the course of time, producing overlap; these are followed by a group of goddesses of the third function. All the scholars agree on the central role of an exalted Great Goddess (called Brigit in Ireland), encompassing all the functions and personified by many female deities. The Celts had an aristocratic conception of the afterlife, in which the souls of the warriors were revived in another world of bliss.

According to Caesar’s description, confirmed by the Irish legends, the druids were a priestly class that oversaw all sacrifices, regulated the religious observances, held courts of justice, and adjudicated disputes between tribes; they were seekers of high religious and secular learning and educated the children of the warrior nobility through rote memorization of verses for many years, without any reliance on writing. Kings or chiefs issued from the warrior class but the druids, while keeping separate, controlled their choice and advised them. This picture is strikingly similar to that of the Vedic Brahmans but goes beyond it in that the druids elected a chief druid to oversee the whole class and met in council every year—a centralized organization that the Brahmans never had.

The continental Germans of Roman times, as described by Tacitus, had a small pantheon of three gods and one goddess, reasonably fitting a tri-functional structure; they had no built temples but worshiped their gods in sacred groves, only occasionally assisted by priests, and scorned anthropomorphic images. The Scandinavians of the Viking age did have wooden temples housing gods’ images. By then, Tacitus’ tetrad of deities had not swollen but shrunk to a triad, in which, however, attributes and functions had shifted away from the original structure. Odin was the god of warriors and kings, patron of poetry and knowledge, and great magician; Thor was the beneficent storm god, the upholder of law and rights, and the friend of the common people; and a group of deities called Vanir (principally Freyr and Freyja) were in charge of production, fertility, and love. In the terminal period, the idea seems to have developed that people could “choose” their personal patron god or goddess as an almost all-purpose protector.

In the Viking age there is scant evidence of professional priests, who were just beginning to develop around the strengthening royal courts. Priestly functions, including conducting public rituals and sacrifices and presiding over religious festivals, were usually discharged by chiefs and noblemen alongside their political and judicial functions. Divination, entrusted to seeresses, was held in high regard. Sacrifice of cattle and horses was central to worship; the animal was not burnt but fully shared among participants in a meal, accompanied by plentiful drinking. Upon death, a choice of warriors fallen on the battlefield was welcomed by Odin in the Valhalla to enjoy a life of delights while awaiting the end of days, while the rest of the people went to a dark underworld. The Norse had an eschatology predicting the end of the world in a cataclysmic battle between the gods and the forces of evil, where all will die on both sides, after which a new, perfected earth and a new human race will rise again and the sons of the old god will reign peacefully on it.