This is hardly the blog post I have been planning to sit down and sketch out over the past three weeks (on the coronavirus, naturally) now but needs must. You see, lector carissime, I have done it; I have defeated the final boss of Indo-European philology. Look at this glorious meme:
This is it. The height of #cheekychariotbois philology. I want to commemorate this moment in a blog post. They say nobody has ever saved a joke by explaining it, this is true, but – jocularity aside – I do want to expand on it a little. You see, there was no Indo-European bear taboo. There was a bear taboo in certain Indo-European languages (see the meme, above), but that is not the same thing as one being present in the parent language.
I: Proto-Indo-European taboos
What do we mean by taboo? Please forgive the “the dictionary defines this as…” vibe – taboo (antiquated spelling tabu) is a Polynesian word first brought to Western attention by Captain Cook in the 18th century; it is a form of avoidance speech. If something is marked as taboo it is not to be said or mentioned. It has, at least in Maori, as an antonym noa – we might say a euphemism or perlocution. Think of the logic behind this as something similar to Plautus’ nomen est omen. It’s not so much that naming defines/predicts, but naming something can either summon it (oh no, a bear!) or profane it (Jewish cultic avoidance of the name of god). Noa words therefore arise in order to avoid catching something’s attention, profaning it, or perhaps as a way of appeasing it. Such as the Greek habit of referring to the furies as eumenides (“kindly ones”) or the Black Sea as the euxine (< Εὔξεινος Πόντος, friendly sea).
Actually, I am glad I took the time to define it. Look at the Google NGram results, from Cook’s time onward. Clearly in a society where you can pay people to pee on you in a nightclub in Berlin (seriously Germans? θεός νύ τι καὶ τὰ νεμεσσᾷ), we need to be reminded of the concept of a taboo. Now that is a taboo Hesiod (Works and Days, 758-60) should have listed.
μηδέ ποτ’ ἐν προχοῇς ποταμῶν ἅλαδε προρεόντων
μηδ’ ἐπὶ κρηνάων οὐρεῖν, μάλα δ’ ἐξαλέασθαι:
ἐναποψύχειν: τὸ γὰρ οὔ τοι λώιόν ἐστιν.
Do not ever into the streams of rivers into the sea pouring
or into springs, urinate; much better to avoid it
it is not seemly to relieve oneself therein.
Did the Indo-Europeans have taboos? In the linguistic sense I mean, all cultures have taboos in the physical/cultural sense. Some of these are so deeply embedded they clearly go back to a deep evolutionary kernel long before speech (think of the story of Oedipus), others can not be much younger (the typical Eurasian reaction to the consumption of dogs and cats), whilst others still are evidently much younger and more culturally conditioned (touching wood, Friday 13th). The initial tweet and Hesiod quote show that these were broadly present across the daughter languages. Being a philological blog, we naturally want to discuss words.
One of the strongest pieces of evidence is the development of divine and theophoric names. Whilst some correspondences are apparent across a broad range of languages (PIE: *dyḗws-ph₂tḗr; IIr: *dyā́wš-pHtā́; Greek: Ζεῦ πάτερ Italic: *djous patēr, if you were wondering how the Romans got Iuppiter), in other situations we are left with functionally and cultically cognate deities without tenable reconstructions. It is a fundamentally untestable, but eminently reasonable, that in some cases divine taboos have rendered us unable to find the proper roots and correspondences.
This is best illustrated by Greek, but as I said above it is a requisite of good methodology when working with compranda to illustrate points across language families. We need to be able to distinguish being Indo-European (as in pertinent to the original parent – Proto-Indo-European) and daughter cultures, which are Indo-European in their phylogeny. I am guilty of being loose with my terms here, but I like to think my readers can prise the mens from the madness.
Let’s take Indra, as I have said before my favourite of the Indo-European gods (I swear this is relevant to bears, eventually), as an example of PIE taboos confounding. He is occasionally referred to as Parjánya. Scholars have postulated correspondences with the Slavic Perun and the Baltic Perkūnas we could add to that the Norse Fiorgynn (which requires a glide). The entirety of this data set goes back to associated words for striking (*per) and the oak (*perkʷ-). We might even try to hypothetically recreate a PIE god, *Perkʷuni(y)os.
From a philological standpoint, this is nonsensical. We require special pleading for the Germanic (god, don’t we always?) and such a root in Sanskrit would give us not Parjánya but *Parkunya. The situation is made even more untenable when you look at Baltic variations for Perkons and Perkūnas (in dainos, in Old Prussian, Lettish etc etc). Do we do violence to the our older, better, evidence in order to support our younger and weaker ones? Ordinarily, no. But there is a strong semantic framework involved (especially between Baltic and Indic), and if we allow possible PIE taboos to have existed, we solve some difficulties. We can even account for *κεραυνός as a cultic name with mutation from π/κ as being part of the same divine semantic field. 
There may, and I know I am stretching this now, have been some form of taboo avoidance/noa usage in the PIE habit of rendering inanimate objects as deities (fire, friendship/bonds, water etc), but I would not be willing to put money it. This is, briefly, covered in my review of Il Primo Re here.
I think that the Indo-Europeans had linguistic taboos, just like their descendent ethnolinguistic groups, I also think these may have operated enough force to confound philologists. They could therefore conceivably have applied it to the bear, but it is my contention that they did not.
II: The chonkiest of bois: Proto-Indo-European Bears
Famously, we may reconstruct a PIE word for bear. The eventual decipherment of Hittite and other Anatolian languages allows us to render, *h₂ŕ̥tḱos. A perfectly functional o-stem noun. The descendants of this word are particularly widespread: Hittite: ḫartákka; Greek: ἄρκτος; Latin: ursus; Sanskrit: ṛ́kṣa; Brythonic Celtic: arth etc.
The presence of such a productive reflex in Anatolian is significant due to the relative chronology of various Indo-European subfamilies. Anatolian (Hittite, Luwian, Palaic) isn’t just the oldest preserved branch, it represents a very early form of PIE: The laryngeals still have consonantal reflexes (sadly, Saussure did not live to see this), the noun is divided into animate/inanimate rather than the later m/f/n, and the PIE perfect is used to form present tense verbs (which will make sense if you have ever wondered at οἶδα or how the sequence of tenses in Latin works).
You may recall a few paragraphs earlier where I mentioned the importance of moving across language groupings when reconstructing things. Something only present in Greek is not necessarily inherited. Something in a relatively well attested isogloss, like Greek and Sanskrit, may not got back much further than the posited isogloss (sometimes called Greco-Aryan). Likewise, something in historically convergent areas – like Latin or Greek, or Germanic and Celtic – may only represent a much later, shared innovation. The latter ought not ever to be underestimated. The famous centum-satem split is perhaps the most famous example of an innovation taking over an incredibly large area.
Tying this back to bears, the fact that the *h₂ŕ̥tḱos may be found across so many different languages make it 100% certain that the parent language had this word. Meanwhile, let us look at the languages which practice linguistic taboos. We have Germanic words, like our bear, allegedly descending from a root *bʰer- which means something like brown. I say allegedly, because *bʰer- scarcely looks Indo-European. Ringe has argued for a link instead to *ǵʰwer-, wild animal (cf Latin fer; Greek θήρ). As usual, Germanic requires a host of special fucking pleading. As I have said before, I blame Matt Scarborough. The taboo here is obviously strong within each culture but cannot be said to hearken back to the parent language or culture.
Next, I’m taking Slavic. Not Balto-Slavic, you wisely ask? Indeed not, as will become clear. The Slavs have arguably the most charming perlocution in all the PIE languages, *medvědь. The actual compound etymology is something like *medu-ēdis from PIE *médʰu and *h₁édti, rendering honeyeater. The charm comes from a folk etymology I wish was the real thing; *médʰu and *weyd- giving us honeyknower. Serious images of Winnie the Pooh.
Baltic is, to be fair, also interesting. In Latvian we have lacis, in Lithuanian lokys which folk etymology sometimes renders from the verb to lick (PIE *leyǵʰ-, Lithuanian laižyti). I am reminded of the classical/medieval myth of bear cubs being born formless and having to be licked into shape by their mothers. Actually, I recently read a poem by someone on this. If you know whose it was, please comment, it was charming. Anyway, an etymology from licking or lapping (Balto-Slavic *lakti) is formally impossible. The presence of an Old Prussian variation with an anlaut in c or t – clokis/tlokis allows us to render an etymology of “hairy one” or “bristly one”. Less poetic, but more descriptive.
So, whilst the Indo-European languages broadly confirm to a word for bear, *h₂ŕ̥tḱos, German has a word of uncertain providence and actual meaning, but is conventionally taken to mean “brown one”; Slavic has “honeyeater” and Baltic has “bristly one”.
Incidentally, yes, we can follow the sound changes in each language to work out what the words would be. The Germanic form would give us something like *urhtaz (if we follow Ringe in allowing the metathesis of /tḱ/ > /ḱt/, modern English would have *ourt) and the Balto-Slavic *irśtvā́. I suppose you’re welcome to go around Germany, Scandinavia and the Baltic yelling out these words. No guarantee you won’t get eaten by an *ourt though.
Before I go on to wildly hypothesise how bears were conceived in PIE culture, let us reiterate why we cannot speak of an Indo-European taboo. First, there is a perfectly reconstructable root present in the majority of the languages right from the earliest stages of the language. We know this to be the case because Anatolian is behaving. Secondly, the languages with the taboos all have wildly different noa words. This does not make sense in the case of an inherited taboo. But the strongest piece of evidence is that Slavic and Baltic both, somehow, have different words. Balto-Slavic clearly form an isogloss, and therefore if the taboo had present at any early stage these two at least should share it.
Let me state clearly: in each and every case, the taboo is an unconnected innovation.
The converse just beggars all logic and requires special pleading several orders of magnitude beyond what Karl Popper would allow on his fanciest cocaine binge. It requires that PIE had a bear taboo, that not a single of our genuinely ancient languages inherited, but it may be found amongst our weirdest and of our youngest testified , moreover not only are all these words different, but an isogloss does the unprecedented thing of creating (at least) two versions of this taboo. A taboo allegedly so strong as to go back to the mother tongue. Hmm.
I’m not going to entertain the idea that the PIE word itself has undergone taboo transformation; that puts far, far, too much weight on the Indo-Aryan evidence. The thinking seems to be that there is a connection with Sanskrit rákṣa (destroy). Except there is no taboo whatsoever about mentioning the various demons given this name, the usage is restricted to Indo-Iranian, and the parent word spreads just as easily – amongst those who don’t have bear taboos. Just as sensible to posit a taboo of hyper protective bears with a root in rakṣ (protect).
So, we can conclude that whilst certain Indo-European peoples had a bear taboo, it was not an inherited one. We can even look at the map and hypothesise that is because they were more likely to run into bears (though that brings the Welsh and the Albanians into questions, what? Where they just…not scared of them?). Which seems sensible. I’d be far more scared of coming across a bear den (berloga in Russian!) on foot in the German forest than I would on horseback out on the steppes…
But can we say anything about bears in Proto-Indo-European culture? After all, the lack of avoidance language does not at all preclude cultic engagement, mythology, and all sorts of ursine goodness.
III: *h₂ŕ̥tḱoes?? néh₂u h₁moí?? kʷod!?! It’s more likely than you think!
Little can be said, by me at least, about the Hittite religion. Hittite culture arose in a confluence of “indigenous” Hattic and Indo-European speakers, locally, whilst at the same time partaking in the wider Mesopotamian (Sumerian and Akkadian, later, the Egyptians too; hence Qadesh and Amarna) koine. Sources speak of “thousand gods of Hatti” and they are not wrong! Some gods are transparently PIE, others are simply names of rivers or tutelary deities, others – like the goddess Belat or the god Enlil – are imported, others still must be “indigenous” Hattic deities. So, yes, complex.
Hittite culture seems to be surprisingly legalistic, with firm categories in place between the animal world (divisible into domestic and wild), gods, and men. But such strictures only serve to make liminalities saucier. That said, there seems to be little crossing in a religious sense with animals (and bizarrely well thought legal strictures contra bestiality; was it such a big problem!?). One text speaks of a “bear man”, hartagga, being shot at by a female archer. The text is obviously ritualistic/religious/magical in nature, but there’s nothing about bear veneration here. My gut tells me the main aspect here is to do with the subverting of norms more than anything. Moreover, the text itself is full of Hattic words and may not represent anything at all very Indo-European.
There are examples of theriomorphic deities (or similarities to them) in the Hittite tradition, which certainly have PIE parallels. As in Indic and Baltic, the bull represents the storm god; the god of the hunt/wild, (K)Runtiya seems to be cerviform. But the bear seems not the play a role.
Hittite has played such a prominent role in this section because, as I said, of the relative chronology of its divergence within Indo-European cladistics. Do not misunderstand: the complicated context (above) would not have exonerated any parallel from the usual philological and structural rules, if anything it would have exacerbated exactitude – but it would have made postulating something in the proto culture a little easier.
Sanskrit (rkṣāḥ; RV 1.24.10) and Greek share an ursine root in their name for the Ursa Major constellation, though the former later replaces it a name meaning “seven seers” with a new myth to match. The Greek name has a myth to go with it (the story of Arkas), I am unaware of any Indic parallel. Doubtless one must have existed at some point. The story of Arkas should be, I think, well known to readers of this blog. Folklorists have long noted that several other cultures have a similar myth of a bear hunt interrupted or pictured in stasis. The broad distribution in time, space, and language rules out even the laxest of areal spreads, we are clearly dealing with several cases of independent invention. Whilst the story (in all its variations – though the Finns win this one) is interesting, it doesn’t tell us anything particular about the Indo-Europeans.
It is a shame that we don’t have an Indic version to compare the Greek to, but as mentioned earlier that would at best give us a Greco-Aryan mytheme and not necessarily one shared by other PIE descendants. #isoglosses matter. A final note on Greek, before I close this section, and the goddess Artemis.
A number of non-specialists unaware of the dialectical variety of Ancient Greek (sadly not just neopagans, esotericists etc, but nowadays even linguists working from data sets rather than learning languages) try to tie Artemis to the root for bear. Alteration of the vowels from i/e (a in Boetian! Unless an error) renders this untenable. Mycenaean iirc even has the variation with i. When Greek does simplify the cluster further, it is the dental that gets eaten. I don’t have it to hand, but I 100% bet Beekes will say it’s pre-Greek… No idea, but the name can’t come from bear.
But but but what about the sanctuary at Vravrona? Whilst Artemis Brauronia has a heavy bear element (maidens played the part of little bears) it hardly holds that one epichoric shrine, in defiance of all evidence and method, holds the real meaning. I do think there is an element of folk etymology involved, though I dare say if we could question an Athenian priest they would rightly remind us that the major elements of this cult are all to do with its role as a centre of initiation. I do think that divergences amongst the aetiological myths and imagery are really interesting. They just have nothing to do with PIE bears.
Well. This has been a monster post. We could continue picking individual PIE cultures, but it seems that in addition to there being no inherited taboo, there are scarcely any wide-ranging parallels. Our best candidate – the myth of Arkas (sometimes called “the cosmic hunt” by folklorists) – seems to be so widely prevalent as to tell us little. I hope this has at least been interesting!
Perhaps like me you find bears charming. Well, there’s a way you can help! Since 1992 the Greek charity Arcturus has been rewilding bears in Greece. You can head over and see what they have done to date – and what they’re doing (despite fire, economic depression, and now the Corona Virus) at https://www.arcturos.gr/en/. If you like what they’re doing, you can donate. For just £20 you can cover the daily needs of your near own bear. How cool is that? Artemis would be proud.
I am thinking of changing the look of the blog. Long-time reader(s) can probably tell that the cheekily bad photoshopped header images have gone and the posts are slowly becoming more multimedia and linked as I learn. I liked the initial design (especially the main page banner), but think the pages are appearing increasingly cramped with this much text and footnotes. Let me know.
Lastly, if you found this interesting, have questions, absolutely must castigate it…feel free to do so in the comments below. If you *really* like it, like and retweet. At the very least you’ll be able to recruit a bigger mob against me. Pitch forks are cheaper in bulk order.
Endlings and suchlike
 As per Cook’s diaries (free online and worth your time), he took the word from the Tongans who pronounced it tapu. There are obvious variations across the Polynesian and Oceanic languages with regards to the (de?)voicing of the consonant and the quality of the vowels.
 This last may be slightly more complicated. It may be a reaction against the local Iranian name axšaina, blue/turquoise/dark, which sounds as if it has a privative alpha in Greek, Ἄξεινος, unfriendly. This Iranian hydronym does seem fairly widespread see e.g the old name of the Vardar, Ἄξιός, which must come from the same root.
 That said through comparison of Latin and Sanskrit sources, we can uncover a staggering amount of ritual taboos, especially as they apply to the Roman flamen and Indic brāhmaṇa.
 Should interest exist, I would like to return to the thundergod. Easy enough to write pages and the vast, vast, majority of material online has been written by morons who would benefit greatly from a basic course in Latin.
 Hence the master Jakobson on his study of the Slavic god Veles: “a rigorous, pedantic application of…grammatical rules to… hieratic onomastics would be sheer fallacy” in Jakobson, R., & Rudy, S. (1985). Contributions to comparative mythology; Studies in linguistics and philology, 1972-1982. Walter De Gruyter. pp 44-5. Neither free nor online, certainly worth your time.
 The importance of Anatolian to this reconstruction can not be overstated, but it absolutely can be boring. No less a luminary than Brugmann argued that the word required a thorn cluster in position final. Anatolian put paid to that and eased our reconstruction for other such important words like the one for earth/ground. Burrows had a fantastic article on this whose name I can’t recall.
 I am not as current with my Celtic philology as I should be, I suspect the aspirate is a Brythonic thing and that the o of the o stem should be kept, giving us *art(i)os or *art(i)us. We have an inscription from a Romanised Celt to the deae artioni in the suspiciously convenient city of Bern in Switzerland. Hmm.
 This is not the place, but I am nailing my colours to the mast that whilst I think Italo-Celtic is a bs grouping, the widespread genitive singular –i is due to areal convergence. I’d also like to note that all Messapians are cowards. Fuck you.
 Ringe, D. (2006). From Proto-Indo-European to Proto-germanic: a linguistic history of English: Volume I: A linguistic history of English: OUP. p106
 Sanskrit, of course, always be flexing with an actual epithet of honeylicker madhulih, but this is more of a joke. If you’ve watched The Jungle Book, you’ll be familiar with the character baloo (Hindi bhalu) which goes back to Sanskrit bhallūka. This has the sense of ‘lad’. Who the fuck, in ancient India, saw a bear and went “lad”? IIRC it can sometimes be used of cats or dogs too, but that makes it weirder. How can anyone confuse those animals? I say this knowing full well I am destined to die at the hands of a bear, a gypsy cursed me in 2011 with this.
 I think Ringe’s comment that these changes causing “baroque alternations” within paradigms is just perfect, by the by.
 As usual, Odin himself could not tell us wtf is going on with Germanic.
 Even Albanian keeps the root word! art/h
 As so often, we risk a false dichotomy here, or at least one that obscures the complexity of reality. PIE *deywós survived in Hittite cult not as a sky or thunder god, but as a god of the sun (Siu-summin or “our sungod”), whereas the god of the storms, Teshub, has his name from the local language despite his PIE trappings: he is a bull, he slays the serpent Illuyanka etc.
 Ever after going through Elements of Hittite I *still* have no idea how specialists catalogue their materials. Just search for KUB 58.14. It’s part of CTH 500 (fragments of festival and summoning rituals from Kizzuwatna). You will find it.
 Bros, can you imagine what Frazer or Graves would make of it? “A Neolithic survival; the ritual is meant to symbolise some sort of sympathetic magic to bring back the bear – a prime source of early sustenance – the archer is female to the underlying worries about fertility and gestation” etc. Pass the port, chap.
 I mention in part (largely) because he lived on well into the Roman period and was often associated with Hermes/Pan. Luwian versions seem to make him capriform, hence Pan. Yes, it’s interesting. You’re welcome.
 If you’re interested generally in the cult, Kahil, L. “L’Artemis de Brauron: rites et mystere” AntK 20 (1977) 86-98 if your best start.