This modified Swadesh list, developed by Morris Swadesh in the beginning of the 1950-th and modified by S.A.Starostin in the middle of the 1980es, is used as a study tool for the history of the genetically related languages. The following table compares the 100 word modified Swadesh list for the languages of the Türkic group. According to the S.A.Starostin's modifications, the borrowings were excluded from the list.

Even with the inherent approximations and imprecision of the method, the results of the comparison tell about the closeness of the Türkic languages, which have the commonality of the dialects. For comparison, the Swadesh commonality for the English-French is between 30 to 40%, for the English-German is between 60 to 70%,  for Russian/Polish/Czech in between 70 to 80%, and, to get to the 90% rate, the commonality of Lisbon and Rio Portuguese stands at 89%.

The 91% per millennium conservation rate, accepted for the Indo-European languages, does not seem to be applicable to the Türkic group. That is being explained by different flexibility of flexitive (IE) vs. agglutinative (Türkic), Türkic being much more conservative because of a fixed system of unchangeable roots, when a change of a single phoneme would change a meaning of the word, or change a grammatical meaning of a sentence. The best example of the preservation is the Hunnic sentence uttered in 312 AD and formally included in the Chinese annals by Fang Xuanling in 648 AD. The phrase is unambiguous, because it had a Chinese translation, and translation of all its words (for more details, see V.S.Taskin Jie Huns Issue 2 1990 page 8):

Not only the lexicon was nearly completely preserved, but the grammatical suffixes survived intact to present. In the 11th c. the same suffixes were listed in the Mahmud Kashgari "Devon lugot it turk" dictionary, 700 years after the phrase was recorded. The 8 semantical components do not provide enough material to deduce a better consevation rate, but it points to a consevation rate approaching an order of magnitude above that established for the benchmark Indo-European examples.

On the other hand, the societal mobility and fluidity of the nomadic states brought them in long symbiotic contacts with a multitude of other languages, which can tentatively be corrected for by typological and statistical analisis of the constituent admixtures. A conservation rate change from 91% to 95% per millennium would effectively double the the estimated duration of geographical separation. There are other coefficients floating around, up to 83%. The table gives 91% and 83% results, both grossly contradicting the known history, even the 95% per millennium rate appears to be way too liberal. The Swadesh analysis also shows the weaknesses of the classifications, with the difference within a classification group larger than between different groups. The Swadesh analysis reflects the admixture effect, not accounted for in the linguistic tree model.

Turkish - Chuvash conservation rate

Chuvashes, called "Suasla Mari" by their neighbors, are a mix of Suases/Suvars with Türkified Mari, and so, the empirically established conservation rate is not applicable to a case which is not a natural divergence of a homogenous population into geographically separate areas. Thus, the number of separation years represents a hypothetical case as how long a homogenous population would have to develop independently to achieve this level of lexical divergence. There is no knowledge whatsoever as to the relationship between Bulgarian and Suvarian languages, though based on ethnological grounds, the Suvarian is believed to be a much-alike Bulgarian Türkic.

Identification of the Chuvashes with Bulgars, first advanced by not very literate Russian nationalists in the 19th c., was adopted first by the Czarist state, and then by the Soviet Russia, as an official doctrine, but it remains disputed by the Chuvash neighbors, who hold them to be Türkified Mari, i.e. genetically Finno-Ugrians. That position was corroborated by the genetical study published in Human Biology, Jun 2003 by Arnaiz-Villena et al., "HLA genes in the Chuvashian population from European Russia: Admixture of central European and Mediterranean populations", which contradicted the official adagio that "Chuvash  have originated from Türkic-Altaic Bulgar tribes who migrated in the 4th century AD from Central Asia together with the Huns to the western region of the Volga River". Nevertheless, an admixture with some Suvars cannot be ruled out, though the genetical trace of the Suvars remains to be found. In the official doctrine the Türkic Chuvash serves as a benchmark Bulgar language, and is widely used as such in linguistic studies, though the merits are at least dubious. With such arbitrary retention factors wrongly applied, and controversial background, no confidence can be afforded to the number of the separation years.