Altaic theory to be questioned:

The Altaic languages are a group of languages and language families, widespread in and dominating parts of Central and Northern Asia. The name Altaic alludes to the Altai mountain range in southern Siberia, where nineteenth-century scholars located the original habitat of the speakers of these languages.

Linguists usually consider the Turkic, Mongolian, and Tungusic languages, often referred to as micro-Altaic, as the principal members of this group; some scholars also include Korean and Japanese, referred to as macro-Altaic. While linguists agree that all these languages show a high degree of structural uniformity and many shared lexical (word-related) and, to a lesser degree, morphological (form-related) materials, the "Altaic theory" continues to be one of the most controversial issues in contemporary comparative linguistics.

The gist of this debate is the question of whether these language families�or a subset of them�are to be viewed as members of a higher-level family of languages (often called a macrofamily), for which a common linguistic ancestor (a protolanguage) can be reconstructed and from which the languages diverged, or whether the common traits and elements found in these languages arose during millennia-long processes of "areal convergence," or large-scale language contact, of originally unrelated languages.

Scholars noticed the high degree of structural similarity shared by the Altaic languages as early as the late eighteenth century and thus assumed that the languages were genetically related. One salient typological feature is vowel harmony: words may contain only vowels belonging to one of two or more mutually exclusive classes, such as front versus back or rounded versus unrounded vowels. The principal morphological technique in Altaic languages is that of agglutination: largely monofunctional affixes (almost exclusively suffixes) added to the roots they modify, in a fixed order, and without the high degree of fusion common, for instance, in Indo-European languages (as in the Latin word librorum ["of the books"], where the ending -rum simultaneously carries the functions of plural and genitive). The basic word order in ordinary sentences in Altaic languages is generally subject�object�verb; modifying constituents, like adjectives, always precede the constituent they modify; postpositions rather than prepositions are found.

Modern linguists, however, no longer view such typological similarities as indicative of genetic relations, mostly because the similarities can change (especially in situations of intensive language contact), and because these and other phenomena once viewed as "typically Altaic" traits occur on a global scale unknown when the Altaic theory was first developed.

Furthermore, critics of the Altaic hypothesis often point to the fact that some typological hallmarks of Altaic languages are historical innovations. For instance, the earliest Mongolian texts (thirteenth century CE) have elements atypical for Altaic (when compared with Old Turkic, for instance), including grammatical gender, no rigid verb�final word order, and adjectives following rather than preceding nouns. Mongolian later developed into a "typical" Altaic language. Somewhat similarly, in Tungusic, the Evenki and Even languages, spoken on the northern and western fringes of the language family's territory in central Siberia and Mongolia, are typologically much more divergent from the Altaic archetype than those spoken in the center, where they have been in contact with Mongolian for centuries.

Proponents of the genetic relations theory usually maintain that the Altaic languages share many lexical items from all semantic spheres, including basic vocabulary. These lexical commonalities are characterized by highly regular phonological correspondences, the most important criterion for genetic relationship. A considerable number of shared morphological elements (affixes) have been identified, and all these observations imply that these languages share a common origin.

Critics of the genetic approach to Altaic linguistics often argue that while the great number of lexical commonalities cannot be denied, most if not all can be explained as early borrowings. In micro-Altaic, the most readily identifiable layers of borrowings are Turkic loans in proto-Mongolian and Mongolian loans in proto-Tungusic. From the time of Ghengis Khan (c. 1162�1227) on, Mongolian elements begin to abound in Turkic languages, and a thin layer of Tungusic loans in early Mongolian is sometimes acknowledged. In particular, the fact that many words common to Turkic and Mongolian show, in Mongolian, traits of one of the subgroups of Turkic, namely the Bolgharic branch (of which modern Chuvash is the sole survivor), leads to the hypothesis that a strong proto-Bolgharic influence on proto-Mongolian is responsible for most commonalities in the two language families. The borrowing hypothesis is further strengthened by the fact that the Altaic languages share few lexical items that belong to those semantic areas that are generally thought to be stable over time and least amenable to linguistic borrowing. Altaicists claim that phonological correspondences between the languages are highly regular, which presupposes the acceptance of many etymologies; these are problematical for numerous reasons, including philological problems of determining the earliest Turkic, Mongolian, and other forms that alone should enter any external comparison; problems of inexact or vague semantic mapping in forms and words compared; and so on. The systems of correspondences proposed by Altaicists contain gaps, as well. Morphological elements compared by Altaicists are mostly confined to derivational morphology and usually involve very short morphemes, often consisting of merely one phoneme, the functions of which may be vague.

Some critics of the Altaic theory maintain that, once all comparisons for which these and similar objections may be brought forward are removed, the remaining potentially comparable items are so few that they could all be explained as chance similarities. Others do not completely reject the Altaic theory, believing that the evidence, when sifted through, will justify a leaner version of the theory.

All these points have been and continue to be addressed by proponents of the theory; both the methodological principles (such as insisting on allowing internal reconstruction of proto-Mongolian, proto-Turkic, and so on always to precede external Altaic-level comparisons) and the factual claims (for example, the acceptability of individual etymologies and sound-correspondences based on them) of their critics have been challenged. Thus, the Altaic debate still continues and has developed into an ideal testing ground for the methodology of assessing (or rejecting) the genetic relations of languages in general.

Stefan Georg

Doerfer, Gerhard. (1963) T�rkische und mongolische Elemente im Neupersischen, Band I: Mongolische Elemente im Neupersischen. Wiesbaden, Germany: Harrassowitz, esp. 51�106.

Georg, Stefan, Peter Michalove, Alexis Manaster Ramer, and Paul J. Sidwell. (1998) "Telling General Linguists about Altaic." Journal of Linguistics 35: 65�98.

Janhunen, Juha. (1996) Manchuria: An Ethnic History. Helsinki, Finland: Finno-Ugrian Society.

Martin, Samuel Elmo. (1996) Consonant Lenition in Korean and the Macro-Altaic Question. Honolulu, HI: Center for Korean Studies.

Miller, Roy Andrew. (1996) Languages and History: Japanese, Korean, and Altaic. Bangkok, Thailand: White Orchid Press.

Poppe, Nicholas. (1965) Introduction to Altaic Linguistics. Wiesbaden, Germany: Harrassowitz.

Because a later development, some Turkic dialects have exceptions on this issue. Like Uighur 'gherdash' instead of old 'qardash'.