Below is a list of Gaulish Gods whose names can be plausibly connected to plant terminology. In addition to Gaulish theonyms, some Germanic and Latin ones from the Keltiké are included, especially with respect to the Matres, Whose cults exhibit a high degree of theonymic productivity and translatio (meaning that equivalent Celtic theonyms could have existed without surviving in the epigraphic record). The entries are presented in alphabetical order of the phytonym, even if it follows the main theonym.
It should be noted that there are, of course, Celtic Gods whose functions relate to the vegetal sphere without having a plant-based theonym, and vice versa: some deities below may be named in relation to a plant while otherwise exhibiting unrelated cultic functions (e.g., some Matres may have been worshipped as the Mothers of a tribe or place that was named after a plant).
This list should not be presumed complete at any point, as some theonyms may have been overlooked; new findings and analysis may result in additions or subtractions to the list; and the entries here are only meant to be brief discussions of pertinent information, rather than fully-fleshed theographical works (although complicated or problematic names may have longer entries to discuss the linguistic issues involved).
Some references are used all throughout, and are thus listed at the bottom of the page. References that are particular to a certain entry will be listed only at the bottom of said entry.
One thing that has grown evident from assembling this list is that Beech Cultus was a major phenomenon that has been overlooked. I will need to dedicate a separate project to that at a later date.
The ‘All-Tree Mothers’ have a Germanic theonym. Their name is structurally similar to another group of British Matres, the Ollototae ‘All-Peoples.’ Considering how frequently translatio takes place between linguistically different dedications to Matronic cults, a similar Celtic theonym could plausibly be constructed. De Bernardo Stempel suggests *Ollo-derw-yai as a cognate (which would differ slightly in meaning, as ‘oak’ rather than ‘tree’).
They are attested along with the Matres Campestres, a Latin epithet denoting the “parade ground.”
Explained by De Bernardo Stempel as the Germanic term *aluz-ô(n) ‘elderberry’ + the *-ikā suffix. This term in turn is supposed to be from the root *aliz- ‘alder,’ augmented with a nasal particle.
The Gaulish term for ‘elder’ would be odocos, as well as variations found in literary sources like ducone, ebucone, eubocone, etc. Perhaps these variants are the odoco- root augmented with a nasal particle like we see in the Germanic term.
This name is found on an impressive stela depicting the three Matronae.
De Bernardo Stempel explains this as composed of *ah(i)z- ‘ear of corn,’ + -ingj(a)- + *-n(a)- + *-ikā-. The compounding that resulted in this name may have gone through phases, with Stempel explaining the *-n(a)- as an intermediary stage that may have indicated a demonym for an otherwise lost toponym.
So perhaps this is best understood as the Matronae of the people of the grain fields.
Arinca, bracis, *sasyo-, etc., are known Celtic terms for grains such as barley, emmer, wheat, etc.
An altar to IOVI BAGINATI features a similar epithet to the Matres Baginatiae listed below, albeit slightly further North than Narbonensis. In CIL’s illustration, the name IOVI is flanked with “palm” leaves which could just as well be beech leaves to my eye, although the caption states that while the leaves appear in the source they referenced, the author did not see them on the altar itself.
While some scholars clearly favor reading this name in relation to ‘beech,’ it should be noted that a homonym meaning ‘combat’ also exists and is sometimes preferred by other authors. The -at- suffix is agentive, so this could lend credence to viewing the root as verbal. However, we do not see these bag- epithets related to Mars, Minerva, Victoria, etc., as would seem more likely if the theonym were war-related (although Jupiter and the Matres can of course well preside over martial matters too).
We generally believe the root means ‘beech’ instead, though, due to the preponderance of related toponyms, as well as the existence of theonyms elsewhere in Gaul that appear to be Latin translationes: e.g., Deus Fagus and Faginus. Cf. also the Roman Iuppiter Fagutalis.
De Bernardo Stempel suggests that this theonym is a Latinzed form derived from a Celtic base of *bāginā ‘beech forest,’ which is also known from toponymy. Whether the name is more of a localizing epithet versus one of vegetal significance (if not both) would need further exploration.
It should be noted that while De Bernardo Stempel translates bagina as ‘beech forest,’ it could be synonymous with the Latin cognate, faginus, which as an adjective means ‘beechen, beech green,’ or as a substantive means simply ‘beech.’
Whether this epithet is synonymous or merely shares the ‘beech’ root with Matres Baginatiae is unclear. I would favor the latter on account of the Baginus and Fagutalis theonyms.
The Latin -ensis suffix is clearly locative, seemingly making These the Matres of a place named Baginienses. Perhaps like the PAG BAG inscription (referring to a Gaulish pagus named in some way for the beech).
The altar is damaged, missing its base, and so far I have not found a more modern record or treatment of it than Allmer, 1889.
Baginus is thus the male accompaniment to these Matres, as we often see in Matres cults. De Bernardo Stempel rejects the Baginahae reading of this altar without explanation. Since I do not have access to the works she cited for that view, I will assume she sees it as a misreading of Baginatiae. Given the damaged state of the altar, it is not hard to believe that TI could be misread as H.
Simply explained as the *bago- root plus -Vno- suffix, making Him the Great Beech God. The caption from CIL suggests that the name is in fact spelled with a G rather than a C.
The reading of the name is complicated by the fact that the the O is nested within the preceding letter. While we would etymologically expect it to be G, alternations between G and C in epigraphy are common (either due to legitimate spelling or dialectal variants, or due to misreadings).
However, the fact that CIL proposed a G reading without even having an etymological motivation to do so should persuade us to see that as the canonical spelling.
On the other hand, His name possibly makes an appearance with the C spelling in the Passion of Saint Marcel. The passage seemingly states Marcel was sent to various Pagan ritual places for punishment: 1) before the statue of Saturn where he was racked, 2) before an image of the sun which was worshipped on the gate of the Sequana (Seine), and 3) to the court of the God Baconis* where an “olovitrea”† statue stood atop a lofty column (perhaps like a Jupiter column?)
* Apparently the text is garbled and this has been read various ways: Divi Hamonis, Decubaconis, Divionis, Divibavonis, etc., which makes this a weak connection but nonetheless worth noting.
† An obscure term that seems to be some sort of glassy, crystalline, or enamel material (“enamel or glass appliqué worked with crystal or enamel”).
His name has been strained to be explained a number of different ways, but see my article here for an in-depth argument of why it most probably means ‘fragrant’ or ‘flowering wall,’ in reference to the original turf wall that extended along the Western half of Hadrian’s Wall, which was only later rebuilt in stone.
While I have yet to find any linguistic treatment for this theonym, it seems nothing other than a precursor to Welsh blodeuog ‘flowery.’ It is normally read as BLOTVGI, but the cramped lettering is rough and imprecise, and it simply looks like a C to me (image can be found in citation below).
The spoons are dated to the 4th Century, so the late date could explain its departure from the more regular spelling of *blatucus, including if the letter is indeed a G.
Other spoons from the hoard that are likewise marked as belonging to various Fauni include those to: Andicrose, Ausecus, Cranus, Medigenus, and Narius. Symbolism such as the panther on other spoons attest to the spoons’ affiliation with the Bacchic cult, which seems confirmed by the epithet Medigenus ‘Mead-Born.’
I have chosen to read BRACIACÆ as a genitive, not only because of the rarity of masculine a-stems, but because Braciaca is probably the root of several French toponyms. If this may be referencing a similar (now-lost) placename in England, then it makes sense to read BRACIACÆ as a localization of Mars.
I would read this name as *kallī- ‘wood, grove’ + *rīg- ‘king’ + *-yos with lenition of the G in *rīg- (as has been found in other Celtic names). Silvanus is indeed the King of the Forest, so this epithet would seem in perfect agreement with the interpretatio.
De Bernardo Stempel gives a similar reading in *Kalni-r-yo-s, using the same first element as stated above, and as she sees in Vocallinae (cf. infra). However, she does not read *rīg- in the second element and did not explain the R.
Alternatively, Delamarre reads Callirius as ‘Hoof-Free’ in reference to wild horses (or, I would suggest, other wild and hooved creatures such as deer). This reading of the second element is phonetically plausible, but I would generally expect *riyo- ‘free’ to be the first element of the compound instead of the second. Delamarre believes *φriyos is found in the second position in a number of names which I would instead read as *rīg-yos like I did above, both of which are phonetically workable.
The term calli- meaning ‘hoof’ is not comparable to anything I’ve found in the Insular languages. However, Delamarre justifies it with reference to the attested epocallium, caliomarcos, glossed “ungula caballina.” I find that a solid enough basis for establishing the word, but I would point out that those examples are apparently yo-stems, which isn’t reflected in Callirius like it is in caliomarcos.
Delamarre points out other such names as Riocal(l)at(is) ‘Free-Hoof?’ and Riomandus ‘Free-Horse’ which would bolster his interpretation. Although calatis is not guaranteed to still mean ‘hoof,’ in my view, especially when it could instead be *kaleto- ‘hard, strong.’
In the end, I still choose to read this as the ‘Wood-King,’ primarily because of its Silvanian interpretatio, the appearance of calli- as an i-stem, and the order of the terms in the compound—although none of those fully rules out Delamarre’s theory either. The panoptic, “inter alia” approach would be called for here: The Wild-Hooved Forest-King.
The name is simply connected to *kannīnā ‘leek, garlic, daffodil.’ Cf. Irish cainneann, Welsh cennin.
Connected to vulgar Latin coccinus, coccinum ‘scarlet color, scarlet robe,’ or to the ‘plum’ (coccymēlum), plus the -ikā- suffix.
This is the *derwo- ‘oak’ root, augmented with a nasal stem suffix. It was later reshaped as an ā-stem.
These Goddesses are likely named ‘Foliage,’ or ‘Herbs’ (esp. medicinal), with the root being *dolā ‘leaf’ (e.g. πομπέδουλα, *pempedula ‘cinquefoil,’ from Pseudo-Dioscorides) and the collective ending *-yā. Cf. MIr. duille, W dail.
We should also keep in mind that this term is used metaphorically to name things like the “leaves” of a book, of metal, a brooch or a spear blade, etc.
Additionally, there is a similar root, *dolā (W dôl), that can mean ‘meadow, dale.’
Presumed to relate to the *perkʷu- ‘oak’ root by De Bernardo Stempel, which is certainly plausible in light of Ercura. But we should also consider Aquitanian as a possibility rather than Celtic due to its isolated attestation there.
If Ercura has been attested in that locality, perhaps Ercu(ianus) could have been regarded as parhedros to Ercura? The *-yo- particle could also have genitive meaning, although we cannot be sure whether both inscriptions reference the same name and thus if it is present in both.
Both inscriptions are unfortunately now lost, so further study is probably limited.
The etymology of Her name has been speculated and debated for quite a long time, but the consensus is emerging that it relates to the Proto-Indo-European *perkʷu- ‘oak’ root. If that is the case, then Her name is peculiarly archaic or dialectal for lacking the Q > P shift (cf. with Sequana, Whose name is similarly Q-Celtic in an otherwise P-Celtic zone).
The proposed etymology is strengthened by the existence of the placename Hercynia silva, as well as the well-known Lithuanian God Perkúnas, whose name similarly reflects that root. Additionally, the connection with the oak is semantically appropriate for a consort of Iuppiter Infernus (Jupiter, of course, being affiliated with the oak). In a Celtic context, He may be akin to the Caesarian “Dis Pater,” both of which have been sometimes identified with Taranus.
De Bernardo Stempel emphasizes that Ercura’s name is probably meant in the sense of a cosmologic World Tree rather than a literal oak. Indeed, Her find contexts and interpretatio do not point to Her cult being sylvian, arboreal, etc., as one would expect if Her name referenced an oak in the literal sense.
While De Bernardo Stempel sees the H in some variants of Her name as being a remnant of the etymological P that Celtic characteristically lost, I have my doubts and would instead assume the H is probably just a “Hellenized” pronunciation like is the case with Hesus, Helvetii, Hibernia, etc.
Suggested by De Bernardo Stempel to contain an archaic preservation of the *perkʷu ‘oak’ root. This is possible, although the retention of *P in that case would seem to make it definitionally non-Celtic. The coexistence of ERCU in the same region (cf. supra) would suggest a multilingual milieu in that case.
Barring further confirmation or study, this name should be considered potentially Aquitanian (whose meaning would thus elude us, unless it is a loan from an Indo-European language), since it is apparently isolated to that region and does not exhibit the characteristically Celtic loss of etymological *P.
However, the ‘oak’ interpretation here is strengthened by the Percerne epithet found in Narbonensis, which I can agree likely means ‘oak.’
Also spelled FAHINEIHAE, this theonym appears to be derived from Latin fagineus ‘beechen,’ augmented with the *-kā suffix, with Germanized phonetics.
De Bernardo Stempel suggests the name is an explicatio vel translatio Latina of the Baginatiae or Matres Baginienses.
Explained by De Bernardo Stempel as the Germanic term for ‘fern,’ plus *weni- ‘friend,’ and the -ikā suffix.
I would also add that the Celtic cognate *wenyā ‘family, kindred,’ could have had an influence here, and perhaps the -VINE- in this name is functioning similar to how the omnipresent Celtic *-gena- does in Matronic epithets. So perhaps ‘Fern-Born’ or simply ‘Related to’ or ‘of the Ferns’ is another way to read this.
Possibly derived from Celtic vidu- ‘wood’ + -genai ‘born,’ and similar to well-known names like Gwydion.
Nimidos is thought to relate to *nemeto- ‘sacred place, sanctuary,’ and could be compared with Nemetona, suggesting something like a God of the Sacred Grove or Sanctuary of the Wood-Born (Goddesses)?
Simply the Latin word cicera ‘pea’ appended with the *-yā suffix, also having undergone Germanic sound changes.
Note that Cicero was a Roman cognomen. Would there be any connection to the Tullia gens?
Ultimately descends from the root *púk'tokā ‘spruce’ with the -nā suffix appended. Cf. MIr. ochtach ‘Scots pine (?), ridge-pole.’
De Bernardo Stempel suggests a derivation from Pre-Celtic *PERKURA < *Perkʷu-nâ, to which I suppose *-nā was resuffixed. The *-nā suffix seems frequent for plant-related theonyms, lending some more weight to this reading.
It’s difficult to say how plausible this etymology is, given that the retention of etymological *P would define it as non-Celtic (perhaps Ligurian? Or something else entirely?)
However, the similarity between Perc- and Perkʷ- is great enough, combined with the semantic plausibility of connecting the Nymphae with oaks, that I find the ‘oak’ interpretation a fair reading.
Interpreted as relating to the ‘fern,’ in comparison with the Germanic Matronae cult to the Fernovineae. However, this has also been compared to Dea Ratis near Hadrian’s Wall, which I favor reading as a name that relates to the homonym ‘fort, rampart’ (cf. infra).
De Bernardo Stempel disagrees with the connection to the ‘fort, rampart,’ homonym on the morphological grounds that she does not believe it should have been an i-stem yet. However, she does not explain this position, and other authors do reconstruct it as such.
Pending further research, I would slightly favor reading this as related to the ‘fern’ due to its similarity to the other Matronae cult, but both options could be possible.
Robur is notably the term Pliny uses to signify the oaks that Druids considered sacred and harvested mistletoe from.
Similar to dervos, robur has secondary meanings of ‘hard, strong,’ and Breton nerth was even glossed as robur for this reason.
It is a cognate with Celtic *rowdo- ‘red,’ and the root of ‘robust.’
The phytonym *tennyo- is not precise, although it has a tendency to mean non-coniferous evergreens and is the source of terms like tanning, tannic, etc. It could refer to green oak, holly, elder, or another plant. It was appended with a *-wā suffix, then again with *ikā.
De Bernardo Stempel also notes that the alternative etymology proposed by Delamarre, *Te[p]ni-aveya ‘Souffle-de-Feu’ (Breath of Fire) is unparalleled by other theonyms.
De Bernardo Stempel traces this back to Indo-European *upo-kalni-nai, composed of what became in Celtic the preposition vo- ‘below’ + *kallī ‘hewn trunks’ or ‘forest’ and the derivational morpheme -nā, and sometimes also the -kā suffix.
One could also see calli- as ‘hoof’ the way Delamarre does (cf. Callirius, supra), in which case we could read this epithet as referencing something more like trampling ‘Under-Hoof.’
Other authors (cf. Simek 2020) have interpreted this as a dehydronymic byname, but De Bernardo Stempel argues that the -nā suffix most commonly applies to arborial epithets and argues that the Silvanae may be an explicatio vel translatio Latina of the Vocallinae.
The toponym Vacalliniacum (Wachendorf) is probably derived from these Goddesses.
Said by De Bernardo Stempel to derive from something like *pūro- ‘fresh’ > *φūro- > *uro- ‘verdant, sappy’ + *mrogi- ‘territory, region,’ so probably Goddesses of the lush, green fields.
There is no Matronic title attested with the epithet, but They are considered possibly to be Matronae due to being plural feminine divinities.
Stempel suggests the similar sounding theonyms, Ura Fons and Urnia contain a pre-Celtic lexeme for ‘water’ that is unrelated.
Cf. also Matrae Mageiae and Campestres ‘Fields’ (but usually taken in a military context).
Simply connected with the Celtic etymon *wroyko- ‘heather,’ and presumed to belong to Divine Mothers on account of the name being feminine plural (although other types of plural Goddesses could be possible).
The fact that this group of names so frequently receive pre-verb particles (e.g. ad-, ar(e)-, ande-, di-, uer-) casts some doubt on Delamarre’s tentative placement under his entry for the flower called bugios (> OIr. buga). Instead, I would propose that these names are variants of the Gaulish bogio- ‘breaker, attacker’ root, related to PIE *bʰewgʰ- > PCelt. *boug- ‘break, cut, reap,’ thus making names like Dibugius and Verbugía akin to terms like W. difo, diwung and OIr. forbach (‘destructive,’ ‘attacks,’ and ‘fine, tribute,’ respectively).
Matasović dislikes this connection due to inconsistencies with the root vowel. I would note that the nasal infix is evidently omitted in ancient orthography (e.g., Andecombogius seems to perfectly correspond to OIr. adcombongim, cf. Delamarre). We also find vowel variations with other nasal-infixed verbs (cf. *kenget- > cinget-), which is suggestive that the vowel may get reshaped in such cases. Additionally, there could be chronological or dialectal explanations for the differences attested in Gaulish that may need to be investigated further.
Such variations in the epigraphic record seem not uncommon in general, so I would see these as relatively minor hurdles that by no means should foreclose a connection from being proposed, especially when this verbal root is evidently so frequent in each respective language, while at the same time we are utterly bereft of alternative explanations.
Though I therefore interpret these names as not being floral, on the other hand it makes a certain amount of sense for a flower to be named for a root such as ‘cut, reap.’ The bugios flower may thus mean something like a ‘cutting’ or a ‘cutter,’ in reference to it being harvested in that manner, or for being particularly worthy of picking for its beauty, among other possibilities. So there may still be an indirect link.
While I could find no examples of the *are- prefix being paired with the *bougo- root in Irish or Welsh, we do see it used with semantically similar verbs and verbnouns: OIr. airlech ‘slaughter,’ ar·fich ‘does battle,’ airdbe ‘cutting off, slaying,’ etc.
So perhaps Arbugio’s name means something to do with being a Great Destroyer or Fighter.
I have not found any sources on this theonym aside from a brief mention in Delamarre’s Dictionnaire de la langue gauloise, under his section for uerna ‘alder, marsh’, where he simply says, “if not for *u(p)erno-” (an unrelated homonym, which he seems to be linking Aduernus to as opposed to ‘alder, marsh’).
While Delamarre does not explain his apparent rejection of the ‘alder, marsh’ interpretation, it is likely due to the ad- prefix in the name. Ad is a preposition meaning ‘to’ which appears to function as an intensifier when prefixed to adjectives. The homonym *u(p)erno- that Delamarre mentions is likely an augmented form of *uer- ‘on, over’, which in turn is an adjective that could well receive this *ad- prefix.
Not only could the -nus ending in the name be a variant of the well-known *-Vno- suffix (which signifies theonyms and occasionally tribal names, generally agreed to mean something akin to “great one”), but there is also the attestation of the adjective veronadas from the Larzac Tablet, which corresponds to its antonym substantive andernad- ‘those below,’ and thus confirms its meaning of something like those ‘on high’.
All this together means Aduernus is most likely ‘The Great and Highest One’, perhaps with a more particular cosmological sense of being the Lord of the Veronađ ‘Those Above’ (if one is inclined to read the Larzac term as having a corollary substantive to andernađ)—although we still cannot yet rule out the possibility that His name is instead related to ‘alder’.
Then again, the *ad- prefix could perhaps make sense with the ‘marsh’ definition, if there were a place whose name meant ‘towards the marsh’, thereby making this some sort of localized or detoponymic epithet. More research is obviously needed, especially if the location of the altar, or any imagery present on it, might help elucidate the issue.
In conclusion, though, of all phonetically plausible interpretations, ‘Great and Highest One’ would seem the most semantically appropriate for a God, especially since this proposed root is already applied to beings in a magico-religious context. It should thus remain the preferred interpretation until more evidence is brought into the discussion than has been found here so far.
The relief found at Bisley in the 18th Century has since weathered away. Delamarre has tried to propose some kind of metathesis to reshape the name as *Ollo-uidius, ‘All-Tree,’ which is tempting to compare with the Alaterviae Matres (cf. supra). However, I am not aware of such a metathesis occurring in any other context, and both inscriptions in these two separate regions are consistent in their lack of the letter I in that position. Furthermore, Delamarre himself says deuos had evolved into diuos and documents countless examples of it. So his explanation is unconvincing and seems like motivated reasoning.
I would instead analyze this as *Ollo-dius, with the composition vowel being raised through affection with the following I.
Dius could be none other than the Latin dīvus ‘divine, divinity,’ or its Celtic cognate, making this more akin to the ‘All-Holy’ or ‘All-God’ Mars, not too dissimilar to the Ollathair ‘All-Father,’ Dagda ‘Good-God’, and Devorix and Divorix ‘God-King’ epithets we see elsewhere in Celtic polytheisms.
Cf. also the names, Dagadius Marius, Diuogenia, Diuuogna, Sacrodiuus, etc., etc.
This Goddess, attested as Dea Ratis, may be named for the fern. This interpretation would accord well with the Matres cults mentioned above. However, the proximity to Hadrian’s Wall should probably be prioritized.
I believe the context or proximity to the wall favors an interpretation of the homonym meaning ‘fort, rampart,’ (or perhaps originally ‘dugout’) from *φrāti- > OIr. ráth, ráith and W bedd-rawd.
De Bernardo Stempel suggests that the ‘fort, rampart,’ homonym was not yet an i-stem in Common Celtic, but does not detail why this should be the case. I am following Matasović’s reconstruction of that root pending further elucidation.
I have found little treatment of this theonym in modern sources, aside from Delamarre’s mention of it in his dictionary.
CIL segments the name as VX · SACANO. It is preceded with LICO on the previous line, which is unclear to me if that is part of the dedicant’s name in 3rd declension nom., or a 2nd declension dat. theonym.
Since I can find little substantiation for LICO in Latin, I think interpreting it as a Celtic Licus (cf. the River Lech), perhaps related to *φlikkā ‘stone,’ is the most satisfactory reading.
I will tentatively accept the reading of Uxsacanus on the grounds that VX does not represent a valid word ending or numeral in this context, as far as I can tell. Additionally, we have numerous other attestations of an Vxa- prefix in Celtic epigraphy.
However, the segmentation might imply that sacanus is a discrete term, or perhaps it was understood to have a third segment (e.g. *uxs- + *su- + *kano-). Succamo is indeed documented as a personal name under Delamarre’s succos, ‘porc’ entry.
Speaking of husbandry, we should note that uxso- can mean ‘ox.’ We certainly have seen terms for barn animals compounded together before in Roman religion, such as the suovetaurilia. This speculation goes to show how semantically divergent the homonyms for this name can be.
However, reading uxso- as ‘high, above,’ is favored by the corresponding Andecanus.
Even if we read the second element as canos, though, our interpretation is complicated by homonyms. Cf. Welsh cân ‘song,’ or object of praise. A high or loud ‘song, voice’ would make perfect sense in correspondence with a ‘deep’ or ‘low’ one as represented by Andecanus.
Finally, Old Irish canach ‘wooly substance, down of plants, cotton grass’ could point to other phytological meanings (although this is perhaps still referring to the seeds of reeds/canes).
The significance of ‘High-’ and ‘Low-Reed’ would be unclear, with Delamarre noting it would have to allude to some mythological premise that escapes us. Which I think would lend itself to the argument that reading canus as ‘song’ may be preferable...
Finally, none of this even deals with LICO—whether it is a separate term, part of the compound, or how anything to do with a rock might factor in to the interpretation of the following name. If it is in reference to some river, perhaps that could relate to ‘reeds,’ but all of this is most unclear and more research is needed.