By Deborah Hayden and David Stifter
There is a popular belief that the names of the ogam letters derive from words for trees: indeed, a quick Google search will typically unearth numerous links to websites where the script system is referred to as the ‘Irish Tree Alphabet’, the ‘Celtic Tree Alphabet’ or similar. This association has a long history, one that is closely linked to discussion of ogam in the Middle Irish (ca. A.D. 900–1200) commentary to the earliest grammar of the Irish language, Auraicept na nÉces (‘The Scholars’ Primer) – a text discussed in an earlier blog on this site as a key source for the initial development of manuscript ogam and for the transmission of ideas about the script into the early modern and modern periods.
In a passage from the Auraicept on the invention of the Latin, Greek, Hebrew and ogam alphabets, for example, one commentator suggested that, when ogam was discovered by the Irish grammarian Fénius Farsaid following the fall of the Tower of Babel and the confusion of languages, the individual letters of the script were named after twenty-five of the noblest scholars in his school (it should be noted that the number ‘twenty-five’ is an indication of the late, learned origin of this account, since the original core ogam alphabet had only twenty letters; the other five, referred to as the forfedae or ‘additional letters’, were invented later and seem to have been intended to represent sounds that were felt to be missing from the alphabet). The Auraicept-commentator’s claim that the ogam letters were named after people is, however, immediately refuted as follows:
Asberat immorro araile co nach o dhainibh itir ainmnighter fedha inn n-ogaim isin Gædhelg acht o chrandaibh gen gu haichinter anniu araile crand dibh. Air atat ceithri hernaile for crandaib .i. airigh fedha ⁊ athaig fedha ⁊ lossa fedha ⁊ fodhla fedha; ⁊ is uaithibh sin a ceathrur ainmnighter fedha in oghaim. Airigh fedha quidem .i. dur, coll, cuileand, abhull, uindsiu, ibur, gius. Athaig fedha .i. fern, sail, bethi, lemh, sce, crithach, cærthand. Fodla fedha andso .i. draighen, trom, feorus, crand fir, fedlend, fidhat, finncholl. Lossa fedha .i. aitean, fraech, gilcach, raid, lecla .i. luachair ⁊rl.
Others, however, say that it is not from men at all that the Ogham letters are named in Gaelic but from trees, though some of these trees are not known today. For there are four classes of trees, that is, chieftain trees, peasant trees, herb trees, and shrub trees; and it is from these four that the Ogham vowels are named. Chieftain trees, that is, oak, hazel, holly, apple, ash, yew, fir. Peasant trees, that is, alder, willow, birch, elm, whitethorn, aspen, mountain-ash. The shrub trees here, that is, blackthorn, elder, spindle-tree, test-tree, honeysuckle, bird-cherry, white-hazel. Herb trees, that is, furze, heather, broom, bog-myrtle, lecla, that is, rushes, etc.1
It is probable that these arboreal classifications of the ogam letter-names ultimately stem from the metaphorical extension of the Irish word fid ‘tree’ to mean ‘letter’, possibly because trees were one of the physical media on which the ogam letters were commonly inscribed (though little evidence for inscriptions on such organic objects survives today) and the vertical orientation of the script in this context was itself seen to resemble a tree. It is noteworthy, in this regard, that the individual strokes of ogam letters were referred to as flesca, lit. ‘rods, sticks’ (cf. Modern Irish fleiscín ‘hyphen’, and also German Buchstabe, ‘letter’, but lit. ‘stave of beechwood’). In other words, the vertical stemline along which the ogam letters were inscribed on physical objects was likened to the trunk of a tree, while the strokes on either side of the stemline were the branches or twigs that stemmed from that trunk – an idea that is reflected in the common designation of the script as ogam craobh (‘branching ogam’) in many early modern and modern manuscript sources. An understanding that this arboreal imagery was figurative in nature is indicated by another passage in the Auraicept, where it is explained that there are two kinds of fedae (pl. of fid): the ‘natural’ tree that one finds in a forest, and an ‘artificial’ tree, which is explained as ‘the tree of ogam’ (fid inn oghaim).2 The preface to the tract known as In Lebor Ogaim (‘The Book of Ogam’, on which see further here) likewise offers the more explicit comment that ogam letters were named after the trees of the forest ‘metaphorically’ (trin troip).3
The analogy was subsequently extended to other linguistic concepts. For example, the word fedae also came to have the more specific linguistic meaning of ‘vowels’ in the commentary to the Auraicept, as distinguished from so-called táebomnai ‘consonants’ (from táeb ‘side’ and omnae ‘bole of a tree’) – a usage that may have been based on the orientation of the ogam consonant symbols relative to the stemline.4 The terms fid ‘letter/vowel’ and táebomna ‘consonant’ are, however, cited in the grammar alongside straightforward borrowings and loan translations from Latin for these same linguistic concepts, i.e. Irish gutta for ‘vowel’ (a calque on Latin vōcālis) and conson for ‘consonant’. Both fid and táebomna may therefore have been coined with the express purpose of creating a native terminology to describe the linguistic system represented by ogam, as opposed to that of Roman script. Such a distinction was superfluous, since both scripts were used to represent the same language; rather, its significance lies in it being a manifestation of a wider attempt on the part of some of the Auraicept’s scholiasts to demonstrate the superiority and comprehensiveness of the Irish vernacular in comparison to Latin (as discussed in the aforementioned blog post here).
The scholars who composed the grammatical commentary to Auraicept na nÉces attempted to reinforce the arboreal associations of individual ogam letter names by giving etymological explanations for their origin. Thus, for example, the word luis (ᚂ, the second letter of the ogam alphabet), identified with the rowan-tree, is analysed as lí sula ‘delight of eye’, on account of the beauty of the berries found on that tree.5 Such definitions employ the widespread medieval method of etymological analysis popularised by Isidore of Seville in his Etymologies, however, whereby the origin and meaning of words might be illuminated by analysing a given term as a compound word or lexicalised phrase consisting of two or more existing words from Irish (or occasionally another language). In most cases, such explanations do not correspond to word-origins established using the modern methods of historical linguistics. And indeed, the meanings of the ogam letter-names that are uncovered by more current etymological approaches are considerably more varied and not exclusively arboreal in nature! Let’s have a closer look (we also include the forfedae into this survey):
1. There is indeed a group of letter names that are unambiguously identical with words that are otherwise found as names for trees in medieval Irish literature. They are especially common in the first two aicmi. Where known, the Proto-Celtic (PC) and Proto-Indo-European (PIE) reconstructions are given. The meanings of the reconstructions are the same as the younger stages, unless indicated otherwise.
ᚁ Beithe ‘birch’ < PC *betu̯ii̯o- ← PIE *gʷetu- ‘resin, pitch’; perhaps so called because of the similarity with the Greek letter name beta.
ᚃ Fern ‘alder-tree’ < PC *u̯ernā- < PIE *u̯er(H)neh₂-.
ᚄ Sail ‘willow-tree’ < PC *salik- < Western Indo-European *salik-, perhaps a loan from a pre-Indo-European language.
ᚇ Dair, older daur ‘oak’ < PC *daru- < PIE *dₔru- ‘tree, wood’.
ᚉ Coll ‘hazel’ < PC *kollo- < Western Indo-European *koslo-.
ᚊ Ceirt needs more discussion. Its value is Q, which means that its initial sound was originally PC *kʷ. Its meaning is traditionally given as ‘apple-tree’, but in view of its manifest British cognates Welsh and Cornish perth ‘hedge, thicket, bush’, ‘hedge, bush’ may have been the original meaning in Irish, too. It is never attested in the sense ‘apple-tree’. The word is usually reconstructed as PC *kʷerχtā- < *perkʷ-teh₂-, as if related with Lat. quercus ‘oak’ < PIE *perkʷu-. It is, however, semantically better set up as PC *kʷers-tā-, derived from the possibly pre-Indo-European root *kʷers- ‘wood, forest’ (cf. OIr. crann, W pren ‘tree’ < PC *kʷresno-; W prys, Sc. Gael. preas ‘copse, grove’ < PC *ku̯res-ti̯o-).
ᚑ Onn ‘ash-tree’ < PC *osno- < PIE *Hh₃os-no- (uel sim.). The more common word for the ‘ash’ in Old Irish is the derivative uinnius < PC *osnistu-.
2. A larger group of names is not independently attested for trees. However, in glossaries medieval scholars want to make us believe that they referred to trees, but the real, non-arboreal meanings of those words are usually well attested.
ᚂ Luis, allegedly ‘rowan-tree’, but cáerthann is the usual word for that tree. The original meaning of the letter name is unknown; it could either be ‘light, shine’ < PC *luχsi- < PIE *luk-s-i- or ‘plant, vegetation’ < PC *lutˢi- < PIE *h₁ludʰ-ti- (cf. OIr. lus ‘plant’ < PIE *h₁ludʰ-tu- ‘growing’).
ᚅ Nin, said to mean ‘ash-tree’, but we already saw the regular word for that tree above. The conventional meaning of nin is ‘part of a weaver’s loom, fork’.
ᚆ hÚath is explained as ‘whitethorn’, but the common word for that tree is scé. Both the sound value and the meaning of this letter name are disputed. Peter Schrijver suggested that the name is identical with OIr. úath ‘fear, terror’ < PC *φou̯to- < PIE *pou̯(H)to-, and that the value H represents the reflex of bilabial *φ < PIE *p. A more plausible alternative in my view is that the original value of the letter was J ‘yod’, i.e. *i̯. This Primitive Irish sound was lost on the way to Old Irish. Úath can be the regular continuation of the Greco-Latin letter name iōta; the traditional value H would only be a late rationalisation within the orthographic system.
ᚈ Tinne, said to mean ‘holly’ or ‘elder’ (already the reference to two unrelated plants should raise suspicion), but used as the regular words for ‘bar of metal, ingot’ and ‘salted pig, flitch of bacon’. There seems to have been a Proto-Celtic word *tanno- ‘green oak’, continued in Gaul. tanno- and Bret. glastannen, but it is unlikely that tinne is connected with it. ‘Holly’ is cuilenn in Old Irish, ‘elder-tree’ is tromm.
ᚋ Muin is said to refer to ‘vine’, but, again, there is no actual evidence for this. ‘Vine’ is usually called fínemain, a loan from Latin uindēmia. There are a number of real words of the form muin in Old Irish, meaning ‘love, affection’ or ‘neck’ or ‘treachery’; the letter name could be any of these. There is a word muine ‘a brake, thicket, grove’, but this has a final -e, which the letter name never has.
ᚌ Gort is said in Auraicept na nÉces to stand for ‘ivy’ or ‘fir’, but it is really a word for ‘field, territory’ < PC *gorto- < PIE *gʰorto- ‘hedged-in area’.
ᚍ nGétal is assigned the value NG by medieval sources, but this contravenes the internal logic of the ogam set to represent Primitive Irish archiphonemes, not mutated sounds. The etymology of gétal ‘piercing, wounding’ < PC *gʷantlo- < PIE *gʷʰn̥-tlo- confirms that it must originally have stood for Gʷ, i.e. *gʷ, before that sound lost its labiality and fell together with *g. Medieval scholars explain it as gilcach ‘reed’ (in Modern Irish, also a word for ‘tweet’).
ᚏ Ruis is another word claimed to mean ‘elder-tree’, but in fact the verbal abstract ‘reddening’ of rondaid ‘to dye red’ < PC *rutˢi- < PIE *h₁rudʰ-ti-, perhaps influenced by the past participle ruisse ‘reddened, red’. (For more on all this see the article Study in Red).
ᚎ Straif has a somewhat uncertain value. Medieval scholars transcribe it as Z and as SD/ST. Allegedly it means ‘sloe’, but the regular word for this is draigen. Alternatively, even medieval sources explain it as sraib ‘sulphur’ < PC *strabi- << Western Indo-European *strap- ‘lightning’, a loan from a pre-Indo-European substrate language in Europe. This is undoubtedly the correct explanation.
ᚑ Úr is once said to mean ‘heath’, which is normally fróech or fráech. Instead, it really means either ‘fresh, new; green’ < PC *ugro- < PIE *h₂ugro- ‘having strength’ or ‘earth, clay, soil’.
3. Finally, there are names that don’t seem to have a real-life reference outside of the alphabet at all, even though medieval scholars again invent tree references:
ᚐ Ailm for ‘pine-tree’ could be a play on words. The sound ‘A’, represented by ailm, can be said to be an uchtach ‘scream, cry’. By way of associative etymology this may have been equated with ochtach ‘pine’. Its original signification is unknown; perhaps it is a deformation of alpha. Unlike other dubious letter names , there is at least one arboreal attestation of ailm in the poem A Marbáin, a díthrubaig a.k.a. King and Hermit (cf. https://twitter.com/ChronHib/status/1283655990999748608), but this could be an educated use of the letter name and does not constitute full proof for the authenticity of the word.
ᚓ Edad is said to stand for ‘aspen’ (usually: crithach). This seems to be a fantasy name with no meaning outside of the alphabet, perhaps modelled on the next item as an assonating partner.
ᚔ Idad is said to mean ‘yew’. It is possible that the original name for this letter had a yew reference, namely Primitive Irish *iu̯o– << PIE *h₁ei̯u̯-. However, that word regularly became éo in Old Irish and was therefore no longer suitable as an acronymic name for I. On an off chance, idad may reflect fidot, a semi-certain word for ‘wild cherry’, through the grammarians’ practice of díchned ‘cutting off a letter’ (namely the first). Alternatively, it is just as artificial as its partner edad, whichever came first.
4. The forfedae are later additions, and their names illustrate the persistence of the need to find connections with the floral world.
ᚕ Ébad, allegedly ‘aspen’ (we have seen that one before) or ‘elecampane’ (real name: elenn). Ébad is probably just a meaningless variation on edad.
ᚖ Ór means ‘gold’, a loan from Latin aurum. Not even the stoutest arborologists tried to find a floral meaning for this.
ᚗ Uillenn’s explanation ‘honeysuckle’ in Auraicept na nÉces looks suspiciously contrived, especially since even its gloss edlenn is restricted to grammatical contexts. In fact, the word means ‘elbow’ < PC *olīnā- << PIE *h₃elēn-.
ᚘ Iphin, Pín is probably called from Lat. pīnus ‘pine, fir’, after the model of the other tree names in the core ogam set. But note that this evident connection to pīnus did not stop scholars from adding a meaningless i- at the beginning and from turning p into ph.
ᚙ Emoncholl ‘twin coll’ is self-referential within the ogam framework: it alludes to its shape ᚙ that resembles that of two colls ᚉ overlaid at a 90° angle.
ᚚ Peith is derived from the younger form beith of the letter ᚁ beithe by devoicing the initial.
 Ed. and trans. Calder, Auraicept, pp. 88–93 (with some minor modifications to the translation).
 Calder, Auraicept, pp. 30–1.
 See Calder, Auraicept, pp. 274–5.
 McManus, Guide, p. 3.
 Calder, Auraicept, pp. 90–1.
Calder, George, Auraicept na nÉces. The Scholars’ Primer (Edinburgh, 1917).
McManus, Damian, A Guide to Ogam (Maynooth, 1997).
McManus, Damian, ‘Irish letter-names and their kennings’, Ériu 39 (1988), 127–168.
Moran, Pádraic, ‘Comparative linguistics in seventh-century Ireland: De origine scoticae linguae’, Language & History 63/1 (2020), 3–23.
Schrijver, Peter, ‘On the nature and origin of word-initial h- in the Würzburg glosses’, Ériu 48 (1997), 205–227.