I recently ran across a reference to Imbolc, an Irish festival marking the beginning of spring. It will not surprise anyone who has read much of this blog to learn that I care little about the festival but a great deal about the name, to wit: why is it spelled that way? The Wikipedia article begins “Imbolc or Imbolg (/ɪˈmɒlɡ/ i-MOLG),” but it is much more commonly referred to under the former spelling, which uses the Old Irish convention of c to represent /g/. But it is very unusual to use Old Irish spellings except in an Old Irish context; we talk about Maeve, not Medb, and Cooley, not Cúalnge. The Etymology section says:

The etymology of Imbolc/Imbolg is unclear. The most common explanation is that is comes from the Old Irish i mbolc (Modern Irish i mbolg), meaning “in the belly”, and refers to the pregnancy of ewes. Another possible origin is the Old Irish imb-fholc, “to wash/cleanse oneself”, referring to a ritual cleansing. Eric P. Hamp derives it from a Proto-Indo-European root meaning both “milk” and “cleansing”. Professor Alan Ward derives it from the Proto-Celtic *embibolgon, “budding”. The 10th century Cormac’s Glossary derives it from oimelc, “ewe milk”, but many scholars see this as a folk etymology. Nevertheless, some Neopagans have adopted Oimelc as a name for the festival.

The usual Wikipedia farrago, but it doesn’t provide any explanation. The eDIL entry says:

n o (? v.l. Stowe), TBC 2473. o lúan taite ṡamna co tate imbuilg (go taitte n-earraigh v.l., Stowe), 3186 . fromad cach bíd iar n-urd | issed dlegair i n-Imbulc, Hib. Min. 49.27 . iar n-imbulc, ba garb a ngeilt `after Candlemas (rough was their herding),’ Met. Dinds. iii 370.61 . Fanciful etymological explanation: óimelc .i. ōi-meilg .i. is[ī] aimser andsin tic as cāirach. melg .i. ass arinní mblegar, Corm. Y 1000.

Which is fun but also no help. This reminds me forcibly of the geas/geis mess we hashed out a few years ago; since I’ve studied both Old and Modern Irish, it doesn’t surprise me a bit that that fine language is such a good provider of conundrums, but I’d still like to get some clarity if anyone can provide it.


  1. J.W. Brewer says

    A quick pass through google books suggests that it’s just not a word (in either spelling) that is used all that frequently in post-1800 English texts outside of: a) descriptions (scholarly or otherwise) of Irish practice of a very-long-ago time; or b) revivalist use in self-consciously neo-pagan or wiccan or what-have-you circles. In other words, while the old-timey pre-Christian feast may have been assimilated to Candlemas or St. Bridget’s/Brigid’s/Bride’s Day or both, and influenced the folkloric celebrations of those occasions by Irish people up to the present day, the old-timey name is apparently not used at least in English for those more modern celebrations outside of self-consciously neo-pagan-etc circles, which may explain why Hiberno-English hasn’t generated an English-phonetic-spelling along the lines of “Maeve” or “Cooley” et al. The sidebar list of wikipedia articles in other languages lacks an Irish one, so I don’t know whether Irish-speakers (who aren’t self-consciously etc) use a word amongst themselves that resembles Imbolc/g or just stick to Lá Fhéile Bríde as suggested by the English wikipedia article.

  2. 2/2/1882 James Joyce born.
    2/2/1922 Ulysses published.
    2/2/1939 Finnegans Wake published.

  3. January First-of-May says

    we talk about Maeve, not Medb, and Cooley, not Cúalnge.

    Not in Russia, weirdly; here, the tradition is apparently to take the Old Irish and transliterate it literally, so we get Медб and Куальнге (yes, seriously, I’m not making it up).

    This creates the impression that Irish is full of horrible consonant clusters. I was very surprised to find out that this is not the case.

  4. The modern Irish spelling used to be Maedhbh, and the Ulster Folk and Transport Museum has a rather nice steam locomotive by that name

    However with reformed spelling it is now Maebh, as in Dr. Who character Maebh Arden, or sometimes Méabh.

    Cúalṅge is now Cuaille.

    Imbolc doesn’t conform to modern Irish spelling rules, but as it is rarely used other than to refer to an ancient festival I’ve haven’t seen it spelled any other way. As JWB says upstairs, in discussing folk customs the day is generally called Lá Fhéile Bríde.

  5. I thought modern Irish spelling was weird, but now I see it’s got nothing on Old Irish. Please tell me it’s at least more internally consistent than English.

  6. WP gives some examples of criticized spellings prior to the 1948 reform:

    beirbhiughadh (beiriú)
    imthighthe (imithe)
    faghbháil (fáil)
    urradhas (urrús)
    filidheacht (filíocht)

  7. David Marjanović says

    Please tell me it’s at least more internally consistent than English.

    That it is, but there’s still some controversy over how many consonant phonemes Old Irish really had…

  8. @J 1-of-M This creates the impression that Irish is full of horrible consonant clusters.

    Um wouldn’t that be the pot (Russian) calling the kettle black? Or is it that Russian consonant clusters are familiar therefore not horrible?

  9. Куальнге seems to be a nice rendering of the Irish into Russian, even transforming Irish -il- correctly into Russian -ль-.

  10. David Marjanović says

    The Russian consonant clusters, especially the weirdest ones from a global point of view, are mostly syllable-initial. The Irish ones are syllable-final.

  11. WP gives some examples of criticized spellings

    And so, some future expert on the 21st century, when annotating ancient electronic archives, will note that “WP is an abbreviation of ‘Wikipedia’, a 21st-century compendium of knowledge on the Internet (q.v.) which allowed participants to freely contribute and rearrange facts according to their preferences. Wikipedia was made obsolete by the later Zhishi Baoku (q.v.), itself a crude forerunner of the holistic ethereum-net (q.v.) experience constellations of the 22nd century.”

    And people will bow down at his great knowledge of ancient arcana.

  12. Tibetan spelling is similarly weird:

    bzhugs rogs gnang (pronounced “juro na”)

  13. In modern Irish the names of the [cross-]quarter days other than Imbolc are the names of the corresponding months [mí X] more often than the day that begins the month [Lá X]. These are Bealtaine May, Lúnasa August, Samhain November. The firsts of the months are Lá Bealtaine, Lá Lúnasa, Lá Samhna (though Halloween, “Oíche Shamhna”, is more common than “Lá Shamhna”; 1 Nov is more usually “Lá na Naomh Uile”, All Saints’ Day)

    Imbolc is not in the dictionary; February is Feabhra and 1st Feb is Lá [Fhéi]le Bríde. I’m guessing the Christianised Gaels found it easier to forget whatever pagan merriments they had attempted in the cold of winter than those that had occupied the long summer evenings.

    I see WP titles vary, with anglicised “Beltane” and pre-1948 “Lughnasadh”. I guess there were not enough neo-pagans in Ireland in 1948 to object to “Lughnasa” become “Lúnasa” and obscuring the connection to Lugh.

  14. Stephen C. Carlson says

    It occurs to me (inspired in part by Roger Wright’s work on Romance) that if Italian never had a spelling reform and kept the old Latin spelling rules, it just might be feasible, though frustrating to almost all.

    For example, here’s an example of an Italian sentence in modern spelling:

    “La percentuale di scritti dell’antichità che ci sono pervenuti più o meno per intero è minima.”

    And here it is with Latinate spelling:

    “Illa percentualis de scriptis de illa antiquitate qui hic sunt perventi plus aut minus per interum est minima.”

    Though it looks like Latin, the language is Italian. Scribes would have to be taught that there are lots of silent letters (final ‘m’, ‘s’, ‘t’, e.g., the first part of illa, etc.).

  15. It’s a strain, but it would be doable. Writing a topolect (usually a Chinese variety, but sometimes Korean, Japanese or Vietnamese) in Classical Chinese and then reading it back in the same or another variety was scribal practice in China for millennia.

  16. The best part, as in Chinese, Tibetan or Greek, is imagining how Classical Latin is read aloud:
    Gallia est omnis divisa in partes tres, quarum unam incolunt Belgae, aliam Aquitani, tertiam qui ipsorum lingua Celtae, nostra Galli appellantur.
    “Gaglia è ogne divisa in parte tre, quaro una incolono Belge, aglia Achitani, terzia chi issoro lingua Celte, nostra Galli appelantro.” or something like that.

    Must say it’s quite reasonable, much more reasonable than how it would work for French (“Jaille est onge divise en parts trois, quer une incueulent Belge, aille Aquitain…”)

  17. January First-of-May says

    Reminds me of one of my favorite parts of Lest Darkness Fall – Padway’s attempt at 6th century Vulgar Latin: Omnia Gallia e devisa en parte trei, quaro una encolont Belge, alia…

  18. Though it looks like Latin, the language is Italian. Scribes would have to be taught that there are lots of silent letters (final ‘m’, ‘s’, ‘t’, e.g., the first part of illa, etc.).
    I have used that idea for one of my conlangs…


  1. […] Hat examines the spelling of the Irish word “imbolc” or “imbolg”, used to describe a […]

Speak Your Mind