#subgrouping

@a-little-bit-of-maxi-in-my-life asks (via the old blog):

Do you also belive that Yurats was the indirect middle of Nenets and Enets? I'd be glad to hear some other opinions. And I know, it is really hard to gain information on Enets and especially Yurats, but still. Thank you for reading this.

Depends on what “indirect middle” is supposed to mean…

From what I’ve heard recently (via Olesya Khanina & Valentin Gusev), there’s firstly a related issue where even Enets itself appears to be partially due to secondary convergence. Aside from ethnographic indications, such as the Forest Enets and Tundra Enets not considering each other a part of a single people, there’s a surprizing amount of exclusive Tundra Enets + Nganasan isoglosses. Even some of the basic vocabulary differences between the two Enets varieties fall in this category. Gusev raises the possibility if this means that either Nganasan and TE have formed an areal group from very early on, or even that the primary dialect division in Northern Samoyedic might fall between Forest Enets and Tundra Enets, not between Enets and Nganasan.

I am skeptical of at least this latter suggestion though. There are still also numerous features distinguishing Nganasan from the Nenets–Enets group, or even from the entire rest of Samoyedic. One of the former is *a > *ä in N–E, early enough to trigger the Common Samoyedic palatalization of *k (e.g. *kajwå ‘shovel’ > N–E *säiwa > TN сива = śīwā, Enets seo, versus Ng kajbu; *kaliŋ ‘armpit’ > N–E *säliŋ > TN сялʼ = śāl°, FE śeri, TE śeðiʔ, versus Ng kaľiŋ). Two of the latter: *j > ∅ in *jïntə ‘bow’ (TNe ӈын = ŋin°, Yurats ngidde, Enets iddo, Selkup ynty, Kamass īnə, Mator mindi, versus Ng ďintə; *j- is original as shown by other Uralic cognates like Finnish jousi, Northern Sami juoksa), and Nganasan having *-ntənA (> -ntənU) and not *-kənA as the ending for the locative case.

But the Enets-as-partly-areal analysis sounds still plausible, and is actually visible also elsewhere. For one example, there is a very deep phonetic isogloss where Proto-Samoyedic word-initial *ä- results in Tundra /e-/ but Forest /na-/, suggesting that the FE/TE split goes pretty far back. This seems to predate numerous common Enets innovations, such as loss of initial *ŋ, depalatalization of initial *ń, and general raising of *ä to /e/ — that is, in FE there is an intervening development *ŋ́ä > *ŋ́a > *ńa before any of these other shifts happen.

(You can now find Khanina, Shluinsky and Koryakov’s presentation covering differences between the two Enets varieties from the program page for the recent 7th International Conference of Samoyed Studies.)

Back at Yurats, then: the most visible feature it shares with Enets is the denasalization of nasal + stop clusters (*mp, *nt, *ŋk > bb~b, dd~d, g), in contrast with *nt > *n and retention of /mp ŋk/ in Nenets proper. IMO this seems fairly trivial though. Already within Uralic, the same happens also in e.g. Hungarian, Permic and most Sami languages, and so this could be easily areal rather than common inheritance. Yurats also lacks some other related consonant cluster assimilations; e.g. in Enets *ms *ns *rs all > *źź (> TE ď, FE z > s), but in Yurats *ms > bs (əmså > *ŋămsa > ngabsa ‘food’) and *rs > rs (*märkä > *märsä > merse ‘wind’) (no examples of *ns are attested). Innovations that Yurats clearly shares with Nenets are not too much more common either, but they include e.g. *ə > a, *ŋ́ > ń and apocope (versus in Enets: *ə > o, *ŋ́ > ∅, no apocope). IIRC the lexicon also points slightly towards Nenets, though I don’t have at hand Helimski’s article where he discusses this.

So if we want to put these languages on a tree and not just call them a dialect continuum / linkage, what seems to me like the best approach is Yurats being a sister group to Nenets; this “wider Nenets” plus Enets (be it in one branch or two) combining to form “Northwestern Samoyedic”; and this being likely independent from Nganasan entirely, even if they share at least *ŋ-epenthesis.

For general Enets sources, I mentioned a few reference materials in my recent main blog post on Kamassian (but these are maybe not news for you). There indeed are not a lot of materials for the language(s) though. Not published at least. Helimski’s collections on Tundra Enets remain in edition, and at the Samoyedology conference I also asked a few people if Natalia Tereshchenko’s archives might have something; but if there is, it is apparently not public knowledge. Also, Florian Siegl is still preparing some additional reference materials on Enets, even if he is otherwise no longer working in Samoyedology.

The forthcoming linguistic volumes of the Manuscripta Castréniana series will have their share of Enets coverage too, but I don’t think there is an ETA yet.