Is Danu a real goddess?

One of the things everyone knows about Irish mythology is that the deities we all know (Brigit, the Dagda, Boand, Ogma, etc.) are all members of the Tuatha de Danann, the people or tribe of the goddess Danu.

However, none of the Irish sources mention this goddess, who was an important enough ancestor that her name identified all of her descendants. (Imagine if the Olympian gods of Greece were known as the Gaians, for example.) Some Celtic scholars have gone so far as to doubt the existence of any such deity. Why?

In fact, the name Danu comes only from linguists’ reconstructions, based on the name Tuatha de Danann. (Like *Rigantona and *Vindos, they represent a “best guess” about what a name would look like.)

What they’re saying is that based on the word Danann, if we take it to mean “of Dana”, the name of the goddess would be *Danu in Irish. (To be grammatical for a second, Danann is in the genitive, or possessive case, while *Danu is in the nominative, or subjective case.)

Anu and Danand

While the sources do not mention Danu, they do mention a goddess Anu, who was based in Kerry, and a goddess Danand.


The Paps of Anu, on the Cork/Kerry border. Wikimedia.

Anu is mentioned in the Sanas Cormaic, a glossary of over 1400 words in Irish. The king-bishop Cormac mac Cuilennáin who is credited with compiling it died in 908 CE. He explains Anu as the mater deorum hibernesium:

Ana, viz. the mother of the gods of the Irish people. Well did she nourish the gods. From her name is said ana, that is “abudance,” and from her name are named the two breasts of Ana in west Luachair, as hte legend is told, viz. as the storytellers say. Or Ana [is, corresponds to] anyon in Greek, which is interpreted as [Latin] dopes, i.e. “food”. (trans. Clark)

Danand is mentioned in the genealogical section at the end of the Lebor Gabala Eirenn, or the Book of the Taking of Ireland. There it says that she was the mother of three gods, Brian, Iucharba and Iuchar:

The six sons of Delbaeth s. Ogma s. Elada s. Delbaeth s. Net, were Fiachra, Ollam, Indui, Brian, Iucharba, Iuchar. Donann the daughter of the same Delbaeth was mother of the three last, Brian, Iucharba and Iuchar. These were the three gods of Danu, from whom is named the Mountain of the Three gods. And that Delbaeth had the name Tuirell Bicreo.

The sons of Tuireann, have the same names, but obviously have different parents. Otherwise these three are obscure.

Some scholars see Anu and Dana as the same goddess, like Proinsias Mac Cana:

Dana: Sometimes given as Danu and cognate with Anu. A mother goddess from whom the Tuatha de Danann take their name. There is a school that believes that Dana is not the same deity as Anu, though most agree that she is. If her counterpart in Brythonic Celtic tradition is anything to go by, then her husband, never mentioned in the Irish tradition, is Bile, the god of death. The Dagda is her son. In some texts it is she, not Brigit, who is said to be the mother of Tuireann’s children, Brian, Iuchar and Iucharba. (79)

The Welsh Goddess Dôn

Another argument in favour of an Irish mother-goddess is that the Welsh pantheon has one: Dôn, mother of Arianrhod, Gwydion and many more. You may be dismayed to learn that Don’s gender has also been the subject of some controversy, although the tale of Math in the Mabinogion says that Arianrhod and Gwydion are the children of Math’s sister, and one of the Welsh Triads calls Arianrhod Beli’s daughter.

At any rate Don is usually assumed to be a goddess, and so the Tuatha de Danann, like the Plant Dôn (Children of Dôn) must be descended from a similar figure, the elusive Dana or Danu. (Of course, the Welsh Classical Dictionary points out that the existence of a Irish mother-goddess was used as proof of Dôn’s female gender, so it all gets a bit circular.)

Rivers and the Goddess

Another possible support for Danu comes from Indo-European linguistics and myth. As the Wikipedia entry on Danu puts it:

The etymology of the name has been a matter of much debate since the 19th century, with some earlier scholars favoring a link with the Vedic water goddess Danu, whose name is derived from the Proto-Indo-European root *dhenh2- “to run, to flow”, which may also lie behind the ancient name for the river Danube, Danuuius (perhaps of Celtic origin, though it is also possible that it is an early Scythian loanword in Celtic).

The Indian Danu was a goddess of the primeval waters, and the mother of the Danavas, who rebelled against the gods. She is mentioned once in the Rig Vedas (RV 1.32.9):

9 Then humbled was the strength of Vṛtra’s mother: Indra hath cast his deadly bolt against her.
The mother was above, the son was under and like a cow beside her calf lay Danu.
10 Rolled in the midst of never-ceasing currents flowing without a rest for ever onward.
The waters bear off Vṛtra’s nameless body: the foe of Indra sank to during darkness.

In later tradition, she became the daughter of Daksha, and the wife of Kasyapa.

There is also a river Danu in Nepal, and across Europe the Danube, Dnieper, Dniester and of course the Don rivers continue the watery theme. All of these names go back to the Indo-European root *dhenh2-.

James MacKillop follows this line of argument in his entry in the Dictionary of Celtic Myth:

Dana: Speculative name for the mother goddess of the Continental Celts based on the evidence of place-names, e.g. Danube (L. Danuvius; Hungarian Duna; German Donau); also a variant for the Irish *Ana (a prosthetic d– = Ana) and linked to the Welsh *Dôn. Other goddesses named Danu appear as far afield as Russia and India; in India’s Rig-Veda the name of the goddess Danu signifies ‘stream’ and ‘the waters of heaven’. (128)

Another Celtic goddess, Matrona, was connected with several rivers in France, and had a shrine at the source of the Marne. Damona, Divona, Belisama and many more were also river-goddesses, so Danu would fit.

But does Danann imply a Danu?

There are several worthy arguments against the idea that Danann necessarily means “of Danu”:

  1. the name Tuatha de Danann is not an ancient one, as you would expect from the people of a primordial mother goddess, but post-Christian, and introduced with a specific aim in mind (Williams).
  2. Danann has been mistranslated, and refers to the aristocratic gods of various skills. (Hamp)
  3. her Welsh counterpart Dôn is not a goddess, but the actual earth, and linguistically Dôn = Danu doesn’t work anyway. (Koch)

The People of God, or God-People?

Williams points out that Old Irish texts referred to “the kin of the Dagda” as the “god-people” (tuath/a dé) or else the god-men (fir dé). Around 900 CE, the name Tuatha de Donand came into use, and by 1200 it had become the familiar Tuatha de Danann. He sees this as a learned usage rather than a popular one, which solved an embarrassing problem.

There were two groups of god-people in Irish thought, and the other one was the Israelites, known from the Bible as the “People of God,” and not to be confused with pagan gods. (People of God and People of Gods looks the same in Old Irish, so both groups were the tuath dé.)

Williams thinks they solved this problem by substituting a particular god/dess, Donand, for the more ambiguous Dé. It would also help to make the old Tuatha Dé seem further distant in time, an important consideration to people who were still dealing with the remnants of pagan tradition.

Of course, this still leaves the question of who or what was Donand. Williams suggests that the phrase tri dé dána, the three gods of skill may have been hybridized with known tribal names including the element Domnann. (188) He acknowledges that all the steps in this process cannot be traced, and that who or what Donand was or what inspired it will remain unknown.

Alexei Kondratiev proposed a work-around for this problem: the Irish Anu was conflated with the name Danand which may have come from other Celtic peoples who influenced Ireland. He assumes it comes from the Belgic language group, since it is close to the Welsh Dôn, and would come from a root word dan- meaning “low ground” or “moist earth”.

Of course, there’s still a lot left unexplained in Kondratiev’s theory, mainly the actual origin of the name Donand. But, as Williams says, that will probably never be known.

The People of Skill

Hamp does not address the Donand issue, but says the name Tuatha de Danann does not refer to a family at all. According to him, they were a collection of skilled, aristocratic gods. He starts with the epithet often applied to *Danu’s brother, the Dagda, “the Good God”:

Instead of the ‘good god’ I suggest that the dago-dieuos was the ‘god of the good ones’ the maithe (Welsh deon), the highest social classes (druids, warriors and kings, and persons of skills and crafts). (Hamp 166)

(Sharon Paice MacLeod turns this around in her paper on Dana, suggesting that Dana herself was a goddess of skills and a patron of the skilled. The Old Irish dán means “gift, skill”, and MacLeod suggests her sons were actually the three craftsmen deities Goibniu, Luchta and Crédne, who are called the trí dé dána or “three gods of skill” in the Cath Magh Tuiread.)

Hamp’s suggestion is not so far-fetched, since Irish sources themselves make a distinction between dee and andee, aristocratic gods and the gods of herders and farmers. (For example, in the Tain, during the episode when the Morrigan is healed, “Now these were their gods, the mighty folk: and these were their non-gods, the folk of husbandry.”)

People of Earth

I mentioned earlier that scholars see the Welsh ancestor-goddess Dôn as a parallel to Danu. By now you won’t be entirely surprised to hear that her existence has been questioned as well. John Koch, editor of the five-volume Celtic Culture: a historic encyclopedia, is not convinced by her:

If what we have in the tale of Math indeed reflects a pagan creation myth or theogony another explanation should be put forward. The name Dôn is found only as semantic genitive, mostly preceeded immediately by merch, mab, or plant. Therefore, we may entertain the possibility that it is a petrified inflected genitive, the exact cognate of Old Irish don ‘place, ground, earth’ (occurring only as genitive, dative, and accusative singular), apparently originally functioning as the oblique of nom. ‘place, location’. The same ablaut is seen in the Greek cognate nom. khthōn, gen. khthonós, ‘the earth’. The Plant Dôn are therefore the Children of the Earth, like the Titans of Hesiod, the primaeval beings of the mythical age. (Koch: 5)

I’m not qualified to comment on this argument, although unlike the Plant Dôn the Tuatha de Danann were not chthonos, in the sense that they were “come-from-aways” who invaded Ireland. Still, the family of Dôn may well have been a family of the earth, in the sense of “this place”.

Gods of Skill or Children of Dana?

I think the verdict has to be the frustrating Scottish one: not proven. Nor is it clear how you would prove it to everyone’s satisfaction. Both Kondratiev and Williams point out that it is impossible, linguistically, to get from the original Irish form Donand to Danu, although Kondratiev does offer a (speculative) way around that.

It is clear that the Tuatha de Danann could equally well be the “gods of skill” without changing them at all. They are gods and goddesses who possess the skills of magicians, druids, warriors, healers, bards or skilled craftsmen. The goddess Brigit, for example, combines three high-status skills: poet, healer, smith.

I think that while these arguments may convince many Celtic scholars, it is unlikely to sway the large number of people who follow Celtic Reconstructionist or other pagan/Wiccan paths, and who would argue that they know Danu exists, because they’ve experienced her.

To me, she is a modern goddess, and I think she meets several modern needs well. Both she and the Welsh goddess Dôn are a welcome departure from the Uranus-Saturn-Jupiter pattern of male rule that characterizes Graeco-Roman and Middle Eastern myth.

They show that a pantheon with a goddess at its head is possible, and help to set the insular Celts apart from other European pagans, including their continental cousins, who traced themselves back to Dis Pater.

Still, the whole discussion is a welcome reminder of how little evidence we have in dealing with the pagan past, and how our modern assumptions can influence the way we interpret the bits and pieces we do have.


If you liked the image at the top, click here

Hamp, Eric P. 2002: “The Dag(h)d(h)ae and his relatives” in Donum Grammaticum: Studies in Latin and Celtic Linguistics in Honour of Hannah Rosén, eds. Lea Sawicki and Donna Shalev, Peeters Publishers. (Google Books)
Koch, John T. 1989: “Some Suggestions and Etymologies Reflecting upon the Mythology of the Four Branches”, Proceedings of the Harvard Celtic Colloquium 9 (1989): 1-10.
Kondratiev, Alexei: “Danu and Bile: the Primordial Parents”, retrieved from the Wayback Machine, originally published in An Tribhis Mhor: the IMBAS Journal of Celtic Reconstructionism 1/4 (Bealtaine 1988).
Mac Cana, Proinsias 1997: Celtic Mythology, Chancellor.
MacLeod, Sharon Paice, 1998-9: “Mater deorum Hibernensium: identity and cross-correlation in early Irish mythology”, Proceedings of the Harvard Celtic Colloquium 18–19 (1998–1999): 340–384. (JSTOR)
MacKillop, James 2004: An Oxford Dictionary of Celtic Mythology, OUP. (Oxford Reference)
Williams, Mark 2016: Ireland’s Immortals: A History of the Gods of Irish Myth, Princeton University Press. (Google Books)






20 thoughts on “Is Danu a real goddess?

  1. greg

    I think you are right to privilege current living experience over verified historical continuity. In the case of historical sources from after the establishment of christianity, it seems to me quite possible to recognise divine figures in them (gods can inhabit stories) and also to argue that, at least in some cases, the author is aware of the divine nature of characters, however these may actually be presented. This certainly, in my experience, is the case with Rhiannon whether or not the process of linguistic development can take her name back to a reconstructed *Rigantona (though I believe it can). Epona, of course, is attested in earlier sources and we can trace other characters by the same process from attested deities such as Mabon who similarly develops from the fully attested Maponos.

    In the case of Dana, the arguments you summarise do show that her construction as the progenitor of the ‘god people’ is a later development of the tradition. But the gods, as they appear (namelessly?) in Nature, may appear in our stories about them and in our worship of them, with different names and different times. If the ancient ‘Tuatha Dé’ only adopted Dana in later sources, and only in some of these, she may nevertheless speak to us in a divine voice. It is, after all, in the nature of the gods to be elusive.


  2. Faye

    Irish gods appear to be particularly illusive. Even when Patrick came to Ireland to ‘bring the Christian faith, he is believed to have commented………Their gods are everywhere often hidden always ‘out of sight).


    1. solsdottir Post author

      I like that. The Celts apparently didn’t make images, and instead of temples they had groves. So you can see why Patrick was frustrated. He would have been used to the Romanized Britons, who had both.


  3. SimpleHuman

    Goddess Danu is the mother of Danavas, a clan of gods. Goddess Diti is the mother of Daityas, another clan of gods. Goddess Aditi is the mother of Adityas (Olympians). There was a war in Swarga Loka, where sons of Diti and Danu fought the sons of Aditi and other allies. Indra/Thor/Appolo slayed the dragon son of Danu and there was a battle between Indra and Danu. The defeated Danavas went to Patala, netherworld where they still plot ways to regain the lost kingdom of heaven. 🙂


    1. solsdottir Post author

      Interesting. You can see how people made connections between her and the hypothetical Danu of Ireland, especially since the Irish Danu was supposed to be the mother of the Tuatha de Danann.


  4. SimpleHuman

    Goddess Danu is the mother of Danavas, a clan of gods. Goddess Diti is the mother of Daityas, another clan of gods. Goddess Aditi is the mother of Adityas (Olympians). Danavas and Daityas can be compared with Titans. There was a war in Swarga Loka, where sons of Diti and Danu fought the sons of Aditi and other allies. Indra/Thor/Appolo slayed the dragon son of Danu and there was a battle between Indra and Danu. The defeated Danavas went to Patala, netherworld where they still plot ways to regain the lost kingdom of heaven. Since Danavas and Daityas resides in the underworld, they are often referred to as demons or anti-gods. Their teacher is Shukra, god of venus. They are actually equal to gods. All the class of beings in indian stories are just like humans, there are good bad and ugly characters among all of them. There are no pure angel and demon class. 🙂

    Liked by 1 person

  5. SimpleHuman

    In indian stories there are several secret portals located at various places on earth through which one can reach the nether regions. They are more like dimensions. Many characters gets teleported to one of the 7 lower lokas when they enter certain caverns, caves, doors in mountains or some underwater places. The underworld is not some hellish place in indian stories. They are described to be more beautiful than the upper worlds and more materialistic and luxurious, full of pleasures of senses. The upper worlds are for the more evolved souls who have gone away from material pleasures. The swarka loka among the upper lokas is the most pleasurable one, equivalent to abrahmic heaven, ruled by god-king Indra.


  6. SimpleHuman

    These are the major Danavas 🙂

    Liked by 1 person

  7. Ali Isaac

    Very interesting post. I’m not qualified either, but it seems to me that it is possible that the Goddess Danu may have been a convenient medieval invention. As you say the name is genetive so perhaps when they were called ‘the people of Danu’ it actually referred to a leader, or even their place of origin. Early antiquarians who were discovering all this in the 17th and 18th centuries could only interpret it in terms of their classical education, and tried to make it fit with the Roman and Greek pantheon. Of course there is no similarity there. If anything the Vedic connection and even Native American shamanism shares more similarities. Just my opinion. 😊 I enjoyed reading your post.


    1. solsdottir Post author

      When I first read someone saying that the goddess Danu might be an invention, it blew my head off. But when you look into it the idea becomes more and more plausible.
      It’s ironic because if she was a Christian invention, she has taken on a life of her own since, with many followers.


  8. Pingback: Andee: the non-gods of Ireland | We Are Star Stuff

  9. Pingback: Ogma: champion of the Tuatha | We Are Star Stuff

  10. Pingback: The Ambiguous Status of the Tuatha De Danann | We Are Star Stuff

Comments are closed.