The Altaic family controversy
So now that we’ve been introduced to the three language groupings that are hypothesized to constitute the core of the proposed Altaic language family — Turkic, Mongolic and Tungusic languages — we can consider the evidence for and against the Altaic family proposal.
The evidence for the common descent of Altaic languages, as with other language families, involves cognates from the basic vocabulary and recurring sound correspondences. Among the proposed Altaic cognates are first and second person pronouns, case and plural suffixes and some color terms. Compare, for example, the early Turkic män- ‘me’ with the written Mongolian min- and the Tungusic min-. Similarly, the accusative case marker in both Turkish and Mongolian is -i, the genitive case marker is -in, and the plural marker is -lar in Turkish and -nar in Mongolian. The possible cognate color terms include the Turkish kara and the Mongolian xar for ‘black’, the Turkish gök and the Mongolian xöx for ‘blue’ and the Turkish boz and the Mongolian bor for ‘grey’.
Based on this sort of similarities, scholars like Nicholas Poppe and Roy Andrew Miller have reconstructed Proto-Altaic sound system, lexical and even morphological material. This proto-language would have been spoken sometime during the Neolithic period somewhere in northern Eurasia, but at this juncture it is difficult to be more precise about its homeland, given how little is known about this possible proto-language at all.
Furthermore, Turkic, Mongolic and Tungusic languages share certain grammatical features: vowel harmony, agglutinative morphology, Subject-Object-Verb (SOV) word order, postpositions rather than prepositions, auxiliary verbs following rather than preceding the lexical verb. However, there is good independent evidence that the last three features (SOV, N-P and V-Aux orders) may well be part of a typological pattern controlled by the same linguistic parameter — the headedness parameter. Hence, this selection of features does not provide good evidence for common descent of Turkic, Mongolic and Tungusic languages.
An additional reason to doubt the mutual relationship between these three groupings of languages comes from a comparison of similarities across modern languages and across earlier stages of these languages. Here is the logic of this argument: the further back into the past we go, the closer to the proto-language from which all of these languages have presumably sprung, the more similarities we should find. This is certainly true for Indo-European languages: there is more similarity between Latin and Sanskrit than between Italian and Hindi. But the same does not seem to work for Altaic languages: an examination of the earliest written records of Turkic and Mongolic languages, for example, reveals less rather than more similarities between them. In other words, we can observe convergence rather than divergence between Turkic and Mongolic languages – a pattern that is easily explainable by borrowing and diffusion rather than common descent.