So here is Mari written in Arabic script:

پستملا يمو پيرکم پو
Pə̑s(t)mə̑la, Jumo, perkem pu
'Bi-'sm-illāh, Jumo [the Supreme Deity], grant [us] prosperity!'

Documented in Mustafād al-Akhbār fī Aḥwāl Qazān wa-Bulghār (1885) by Shihāb al-Dīn Marjānī.

We suggest that it was within the ethno-linguistic landscape of the Ar Land in the 14th and early 15th centuries that the Chuvash people emerged from Bulghar-speaking groups who survived the catastrophes of the period.

We link the emergence of the ethnonym to the 14th century, when archaeological and historical evidence shows a large migration of Bulghar groups from the left bank of the Kama to the right bank — the so-called Ar Land (Арская сторона), east of Kazan.

We argue that the term passed from a Bulghar-Chuvash dialect into Tatar as čuaš, and was later borrowed back into Chuvash as čăvaš.

In our view, only one etymology withstands criticism: a connection with Old Turkic joγoč ‘the opposite side of the river; those living beyond the river’. A regular Chuvash reflex of this word is preserved in Mari: Meadow Mari suas ‘Chuvash > Tatar’, Hill Mari suasla marə̑ ‘Chuvash.’

We rejected a number of popular explanations: derivation from a Turkic word meaning “humble, meek”, connections with the phantom tribe of the Suvars, and artificial constructions with invented meanings such as “pagan”.

Our task was simple in principle: to review every serious hypothesis and evaluate it against two criteria — the standards of historical-comparative linguistics and historical-cultural plausibility.

Ethnonyms rarely have clear lexical meanings. As a result, their etymology is notoriously difficult: many formally possible derivations can be proposed, and speculative historical scenarios can always be constructed.

This week, the leading Chuvash archaeologist Nikolai Myasnikov and I presented a major joint study. Combining comparative linguistics, archaeology, and written sources, we addressed one of the central problems of Chuvash ethnogenesis: the origin of the ethnonym Chuvash.

Yes, there is Bulgharic *r ~ Narrow Turkic *r. It has in fact been shown that the two phonemes in question are partially in complementary distribution (Helimski's Law): *ɻ > *r / _C.

In addition, reconstructing *z, *š implies that Proto-Bulgharic, rather than Proto-Turkic, was in contact with Samoyedic, Mongolic, etc., which is in my view a highly doubtful scenario.

In my view, Early Bulgharic is definitely more archaic than Early Narrow Turkic, so Chuvash data must be taken seriously in Proto-Turkic reconstruction. Also, *ɻ and *ɬ are not exotic at all for the macro-area where Proto-Turkic was spoken.

Not from the correspondences alone. Extrapolating Proto-Narrow Turkic reconstruction to the Proto-Turkic level is generally a questionable idea, likely shaped by Turkology being taught mainly via Turkish or Old Turkic while Chuvash remains understudied.

In my 2025 paper on Bulgharic-Permic relationship, I reconstruct a retroflex approximant *ɻ at the Proto-Turkic level. The l ~ š correspondence is best interpreted as a lateral fricative *ɬ.

His cautious advice: do not merge the roles of historian and nationalist. 36 years later — has much changed? Only partially, largely due to individual scholars. Systemically, the paradigm remains firmly in place — and this is to a great extent true for other peoples of the Volga–Kama region as well.

1990: Uli Schamiloglu notes that research on the ethnogenesis of the Kazan Tatars largely remains trapped in the paradigm of 19th-century romantic nationalism, while the “Volga Bulgars → Kazan Tatars” continuity functions as a dogma not only in politics, but in scholarship as well.

Our project on documenting Turkic runiform inscriptions has just published its first article in Turkish, devoted to our 2024 fieldwork. A good moment to recall that season: Southern Khakassia, searching for Yenisei Turkic inscriptions at local burial grounds, and life in a wooden yurt.

To sum up, none of the proposed Chinese loans stands up. They provide no evidence for long-term Bulghar residence near China. Early Bulghars must be sought elsewhere.

Finally, Vovin proposed Rus. слон ‘elephant’ ← Chuvash сăлан (~ сăлун) ← Chinese 象. This collapses too: Rus. → Chuvash borrowing is perfectly regular. Chuvash -a- ~ -у- as substitutions of Rus. -o- are well attested, cf. Rus. запон ‘apron’ → Chuvash саппан ~ саппун.

Chuvash çыр- ‘erode (riverbank)’, compared by Vovin to Chinese 溢 ‘overflow’, also has clear Turkic parallels — contrary to his claim.

Instead, çын(ă) has a perfect Turkic etymology (Mudrak 1993), corresponding regularly to Old Turkic jalŋuq ‘person’.

In any case, both alleged Late Middle Chinese loans collapse. Chuvash çын ‘person’ cannot derive from the LMC form of 人: its possessive çынн-и requires earlier çынă, incompatible with Vovin’s ad hoc *ǯin.

This is already absurd: Late Middle Chinese dates to around the turn of the 1st–2nd millennium CE. Were the Chuvash still near China then, arriving on the Volga with the Mongols?

In 2011, A. Vovin tried to prove the presence of early Bulghars in East Asia by proposing a few Chinese loanwords in Chuvash: one Old Chinese (≤1st c. BCE) and two Late Middle Chinese (!).

Thus, “East Asian Bulghars” are an extremely ephemeral construct, and attempts to give them substance usually fail.

As noted by E. Helimski, all peoples securely identified as Bulghar-speaking are localized in Europe, whereas all historically attested Turkic languages of Siberia and Mongolia are Common Turkic.

However, rejecting this fantasy does not imply that the East Asian localization of the early Bulghars is without issue. It relies not on specific Bulghar data, but rather on the assumption that Proto-Turkic was East Asian and the transparent continuity from Proto-Turkic to Proto-Bulghar.

Salmin’s works are scarcely distinguishable from pseudohistory. Yet in 2014 his monograph appeared in English, edited and introduced by Peter Golden — a curious episode in modern Turkology.

Claims exaggerating ancient West Eurasian roots of the Chuvash stem from a broader idea: their ancestors were never in East Asia, with a homeland supposedly in the Caucasus. Today this view is marginal, represented mainly by A. K. Salmin.

What seems clear, however, is that yupa is neither a recent internal formation nor an obvious borrowing. It is likely very old, reflecting Proto-Turkic—or perhaps even deeper—realities.