• Belatucadrus’ name has long been interpreted as dubious variations of “Shining,” and “Handsome,” or “Strong.”
  • More recent attempts by linguists to revisit this name have been fruitless.
  • A strong argument can be made to link the 1st component to “fragrant, blooming,” and the 2nd components to “fortification, wall.”
  • Belatucadrus’ cult corresponds to the western side of Hadrian’s Wall, which was defended by an earthwork fortification or turf wall.
  • Belatucadrus is thus best understood as the tutelary deity of this boundary: the guardian Who defends the “hedge,” or “flowering wall.”

Belatucadrus is a theonym many authors have struggled to contrive an explanation for over the years. We have oft been regaled with the old-hat theory that it comes from *gʷelH-, an Indo-European root meaning ‘fair, shining,’ or ‘bright, burning.’ That root could certainly yield something like **bel- in Celtic, phonologically speaking.

But J.P. Mallory has pointed out that antiquated Proto-Indo-European source books often have a low bar for inclusion, with lax criteria governing how they adduce such roots. Mallory opines that half the PIE lexicon in such reference works seemingly (and implausibly) just describes bright, shiny balloons!

The *gʷelH- etymology seems like a model example of this flimsy, “Bright Balloon” lexicon. More modern linguistic sources, by contrast, reflect the lack of evidence that that full-grade root ever descended into Celtic (other grades of the same root, or a similar one, may have produced words like glo, gual ‘charcoal’ in Welsh and Irish).

Tossing out old ideas: Belatu- as ‘throwing, dying’?

More recently, Delamarre has gone back to the drawing board, relating belatu- instead to Beltaine and Old Irish epeltu (< *eks-beltu) ‘death, dying, perishing.’ This in turn, he says, is supposed to descend from the PIE verbal root gʷelə- ‘suffering, death.’

While suffering or dying could be semantically fitting for a Martial deity, Delamarre rightly casts doubt on it in light of the fact that it does not seem very appropriate for the many personal names derived from it. I would add, especially not for such aggrandizing or diminutive ones like Belatumara or Belatulla.

Delamarre mentions a bigger nail in the coffin from Stüber, who reconstructs epeltu as *eχs-blātiyū, connecting it to *gʷelh₁ ‘throw.’ More recently, Gordon reconstructed it from the ø-grade of that root *-gʷlh₁- which would likewise yield *exs-blation- in Proto-Celtic. From this line of thinking, it does not very well explain Belatu-, even phonologically.

Delamarre ultimately opts to leave belatu- undefined...

A fresh look: Fragrance and flowers

Since scholarship has done little else that I could find to analyze this name, we’ll need to do our own analysis:

Something I find useful in Celtic comparative studies is looking at the *-maro- compounds. Belatumara is attested in Gaulish, so any related adjectives in the Insular languages would make a good starting point. As it happens, Old Irish does have a somewhat similar term: boladmar, baladmar ‘fragrant, sweet-smelling.’ While the root vowel doesn’t match, this would semantically be a much better fit for personal names like Belatumara.

When root vowels alternate like this, we should consider the possibility of ablaut. The primary term, bolad, balad meaning ‘smell, scent’ (usually in a positive sense, or related to hunting), has never had a proposed etymology, but seems possibly related to bláth < *blātu- ‘flower.’ Bláth is normally reconstructed back to PIE *bʰleh₃-, but Matasović notes: “is also compatible with *bʰlh₃-tu-, and the zero-grade of the root is expected” (although he removed this remark later, in his addenda and corrigenda).

Earlier, Schrijver also noted the possibility of bláth being from a zero-grade. He preferred the full-grade only on the grounds that it is what’s more commonly used when constructing terms with the *-tu ending.

Given that, we could be tempted to see Irish as having an o-grade in bolad, as opposed to its ø-grade bláth, with the proto-root being something like *bʰVlh₃- instead of the traditional *bʰleh₃-. This would make Gallo-Britonnic belatu- an e-grade, which may also have parallels in MIr. bile and Gaulish bilio- (*bel-yo-) ‘tree, trunk,’ which Matasović suggests could possibly be from *bʰelh₃-.

If that’s the case, then we could propose a CeRH reconstruction vs. the traditional CReH and see a correspondance in boladmarBelatumara.

Whether this reconstruction is workable outside of Celtic is beyond my ability to determine. We might see Latin as supporting it with flos, folium, although these are probably false friends (folium being better reconstructed from *dʰolyom). Perhaps Greek βλαστός and φύλλον are relevant, but I know less about Greek to say.

In any case, whether there is a “true” etymological link or not, it’s clear these different Irish roots are near-homonyms with related semantics and thus have a connection in a practical sense, if not a theoretical linguistic one. More importantly, there are plenty of examples where Belatucadros is spelled either Balatu-, or Blatu-, showing that the engravers were comfortable with spellings that are identical to what we would expect for both balad and bláth.

It is also worth noting that belatu- reflects the correct stem vowel to match bolad, even in spite of Gaulish’s penchant for replacing said vowels with an O in composition. It seems improbable for there to be an unrelated homophone root shaped like bVlatu- that is so frequent in Gallo-Brittonic onomastics, yet somehow lacking any parallel in any Insular Celtic language’s corpus.

So in the end, while acknowledging its many uncertainties, defining belatu- akin to ‘fragrance’ or in some other way like ‘flower, bloom,’ seems fair enough, as it is both linguistically arguable and lacks any real competitors.

Having explained the first term as satisfactorily as possible, we should now turn to the second one: cadrus.

Arriving at the wall

Delmarre notes that older sources like Jackson would rely on a reconstructed PIE root defined as ‘shine, sparkle,’ meaning we have a “Shiny Sparkler” on our hands! But after discarding yet another old, Bright Balloon-ism, Delamarre finds the closest comparison to be in Brittonic forms like cadr, cazr, kaer, ‘beau, puissant’ (handsome, strong). He notes, however, that there is a phonological difficulty in this derivation since we would expect those forms to descend from *kat- (which in turn probably relates to catu- ‘combat’).

Given the regular development of *tr > *dr in Proto-British, I would assume *T > D is perfectly expected in this context rather than a difficulty to surmount.

The *kat- root seems to be used in a variety of constructions ranging from ‘battle,’ to ‘hate,’ to ‘helmet,’ to ‘fortification.’ It’s hard to suss out the exact relationship and origin of this family of terms, but *katr- in particular pertains to fortifications (or maybe ‘protective strength’ in general, which could be extended by cadr as ‘beau, puissant’). This observation should be highlighted given the context of Belatucadrus’ cult in the vicinity of Hadrian’s Wall.

But that’s not all—some variations of His name include a different term than -cadrus: -gagrus, -cairus, -caurus.

Rather than believe there is a full D > ø elision taking place, I find it much more likely that -gagrus represents a different word from -cadrus, and that the absence of a consonant in this position (ø) represents the lenition of that middle G. Since G and C were often interchanged by epigraphers (or perhaps misread by lecturers), -cairus and -caurus may be variations of the same (although the latter could potentially be a third term, *kawaros > W. cawr ‘giant, hero’).

In other words, *kagros > gagrus, cairus, etc., could explain these numerous variants. The significance of this is that Proto-Celtic *kagros means ‘enclosure, fort,’ and its close relative *kagyos means a ‘hedge-row, fence.’

Some Greek connections?

Mann and Vanderspoel, alternatively, suggested the Greek phrase ό τοῦ βλαστοῦ καιρός “the time of shoots” could have influenced some of the variations here, and that Blatucairus was thus understood as a springtime deity. While there was considerable Hellenistic presence in Roman Britain, this theory does little to accord semantically with the Celtic endings of His name, nor does it explain -caurus in any way. But besides the Hellenistic look of -cairus and a smattering of dedications to Syrian Gods around Carvoran, I will also note that one of the dedicants attested in a Belatucairus inscription, Peisilus (RIB 2056), could have been a Hellene (his name may be built off the root πεῖσις). Other votives, on the other hand, were made by Roman and Celtic dedicants (e.g., Censorinus, RIB 2045, and Audagus, RIB 774).

While more research is warranted in this area, I don’t believe the Greek hypothesis is explanatory for Belatucadros’ cult as a whole, even if some Hellenistic folk etymologies might have occasionally exercised an influence on the spelling of His name.

Hadrian’s Western Turf Wall: The ‘blooming hedge’

Which brings us to our ultimate conclusion: Belatucadros is no ‘Shiny-Sparkler,’ nor ‘Death-’ or ‘Throw-Beau.’ He is not even a ‘Fragrant Beau.’ No, Belatucadros is no less than a flowering wall: a hedge. A verdant barrier. Meaning Mars Belatucadros is the tutelary deity Who confers and maintains the wall’s protection. Now you may wonder, why would anyone refer to Hadrian’s Wall as a blooming hedge?

This is made rather obvious by His cult’s exclusive affinity to the Western half of Hadrian’s Wall (and the settlements downstream from there). Let’s turn our attention to the Western wall’s development over time:

From Milecastle 49 to the western terminus of the wall at Bowness-on-Solway, the curtain wall was originally constructed from turf, possibly due to the absence of limestone for the manufacture of mortar. (Wiki)

Since Belatucadros’ cult correlates with the Western half of Hadrian’s Wall, it follows that this earthwork defense, which must have grown over with groundcover that included flowering plants, would be the eponymous “Belatucadros.”

And besides the temporary earthen curtain wall, there was the vallum: “a ten-foot (3 metres) deep, ditch-like construction” that lay behind the stone wall, “with two parallel mounds running north and south of it,” which could just as well take such a name.

So in sum, after considerable philological journeying, and investigation of the site context wherein we find Belatucadros’ cult, we can finally arrive at a rather specific and apt explanation for not just His canonical name, but even the various alterations of it. Belatucadros is the guardian of the frontier—our defender who maintains the hedge, the flowering wall.


  • Delamarre, X. (2003). Dictionnaire de la langue gauloise: Une approche linguistique du vieux-celtique continental (2nd ed.). Errance.
  • Gordon, R. C. (2012). Derivational morphology of the Early Irish verbal noun. University of California.
  • Mallory, J. (2019). Proto-Indo-European, Proto-Uralic and Nostratic: A brief excursus into the comparative study of proto-languages. In B. Olson, T. Olander, & K. Kristiansen (Eds.), Tracing the Indo-Europeans: New evidence from archaeology and historical linguistics (pp. 35–58). Oxbow Books.
  • Mann, M. L., & Vanderspoel, J. (2003). The End of Winter: Belatucadrus as a Spring Deity. Mouseion: Journal of the Classical Association of Canada, 3(3), 399–414.
  • Matasović, R. (2009). Etymological dictionary of Proto-Celtic. Brill.
  • Schrijver, P. (1995). Studies in British Celtic historical phonology. Rodopi.

Previous Post Next Post