This brief and brilliant review of the state in the Alataistic philological classification gives a clear picture on the current status, and may exceed in objectivity the limits of understatement. Few accents may be added.

The first is the Türkic-Mongolic symbiosis. Though suspected, and may be reasonably surmised, there is no historical evidence about it in the 1st millennium BC, other then limited Sogdian, Tocharian and Hunnish enclaves in the presumably Mongolian territories. The first known mass integration happened around 202 BC, when Maodun dismembered and annexed Dunhu Mongols, separating Mongols into Türkic-affiliated and independent; then after 160 AD, when the Hunnish confederation split, and its many tribes that remained in situ interspersed with Mongolians, changing their ethnic appellation. The second happened in the Early Middle Age, when the Türkic confederation split, and again its many tribes that remained in situ interspersed with Mongolians. From what we know about demographic situations, in both cased the Türkic component outnumbered the Mongolic indigenous population. As a result, this cultural diffusion made the Mongolian language what we know it to be today. The impact of the 13th century Chingisid Empire, as far as we know it today, was minimal in respect to linguistic influences, and concentrated in fairly narrow semantic field. The linguistic diffusion happened in the 1st millennium AD. The fact that even after a millennium-long symbiosis Dr. Lars Johanson does not find evidence of the genetic connection between Mongolian and Türkic families certifies how distant were the languages before the beginning of the exchange.

Secondly, we know that Hunnic tribes, and later their splinters were in close and continuing contact with Dunhu-Tunguz tribes, enough to impose some linguistic impact directly and without Mongolian intermediary. That should not minimize the Mongolian influence, but a result would be a compound effect of direct and indirect diffusion.

A further review of the Mongolian, noted by N.Bichurin early in the 19th century, is a dialectal differentiation between the western and eastern Mongolian, the differentiation that clearly displays the language affected, and the language largely non-affected by the influence of the Türkic languages. Investigation segregated along the dividing line would bring results much different from an undifferentiated mixture of these dialects.

The genetics also gives us a hint. Agglutinating languages center in a huge radius around SE Asia, and so does the blood group B, these two maps are overlapping, and point to a land and a coast routes of the spread for both. While innovations over millennia separated the languages, the agglutinative philological backbone, and some other common philological markers were preserved. Notably, the linguistic definition so superbly stated in the preamble of the Lars Johanson's article does not find reflection in the following philological overview, leaving a major void in the deliberative content. In the 20th century philology, the typological structure of the languages took a back seat in respect to more detailed expressions of the  morphology, supposedly because agglutination as a typological trait that cannot be used as evidence of genetic relationship to other agglutinative languages. That is a nice hill to occupy if you want to present the pidgin analytical languages like English and Chinese as being the top of the human development. However, with the absence of the examples to substantiate this notion, and with the opposite example that Native Americans had been retaining their typological structure over the whole 10+ millennia period of their isolation, offhand dismissal of this fundamental philological  factor can easily lead to arimaspu, using the Scythian word for a half-eyed, and incorrect conclusions.

The bottom line of Dr. Lars Johanson discourse is: Türkic languages are a linguistic family, and neither Mongolian, not Tungus, nor Korean and Japanese belong to it. Any lexical admixture to these languages is from the cultural diffusion only. Why such a strong statement? Because, using L.Johanson own ringing words, there is a “lack of evidence for a common basic vocabulary in Altaic”. “Look-alikes that occur only in typical contact zones cannot easily be used as evidence for genealogic relatedness.” In Turkic-Mongolic, “The few pairs of corresponding words do not, however, relate to the most significant parts of the vocabulary”. And “there are hardly any plausible lexical correspondences between Turkic and Tungusic.” For the remainder, “problems with the Altaic affiliation of Japanese”: “Japanese (with Old Koguryo) and Ryukyuan to form a distinct family of its own”. Are we at a deathbed of the 19th c. easy hypotheses? Not so fast, first we need an uncomfortable sober-up period.

Page numbers are shown at the top of the page. The posting's notes and explanations, added to the text of the author, are shown in parentheses in (blue italics) or blue boxes.

A common designation for the typologically related languages of the Turkic, Mongolic, and Tungusic families is 'Altaic languages'; according to some scholars, this designation also includes Korean and Japanese. The common typological features of these languages include an agglutinative and exclusively suffixing word structure, sound harmony, verb-final word order, with dependents preceding their head, and use of numerous nonfinite verb constructions.

Altaic as “Ural-Altaic“

The term “Altaic“ was first used by M. A. Castren in the middle of the 19th century for a supposed family comprising Finno-Ugric, Samoyedic, Turkic, Mongolic, and Tungusic. This group of languages was later called “Ural-Altaic“. The Ural-Altaic hypothesis, which was largely based on general typological criteria such as agglutination and vowel harmony, was widely accepted in the 19th century. Later on, this hypothesis was seriously doubted. The works on “Altaic“ languages by W. Schott, M. A. Castren, J. Grunzel, H. Winkler, and others contain abundant incorrect data. Castren, however, rejected the purely typological approach and applied linguistic criteria of lexical and morphological comparison. There are not sufficient materials to establish a Ural-Altaic protolanguage (i.e. using an unmolested protolanguage as formulated in the nonsensical “Language Tree“ paradigm. In a diffusional paradigm the existence of a symbiotic, syncretic, and dynamic “protolanguage“ is quite viable).

Scholars of the following period, e.g., J. Nemeth and J. Deny, who took a more cautious attitude, published detailed works on phonology, word formation, etc. Syntactic typological arguments for the unity of Ural-Altaic were, however, discussed as late as 1962, by Fokos-Fuchs.

Altaic as “Micro-Altaic“

Scholars such as G. J. Ramstedt and N. Poppe argued for a “Micro-Altaic“ family (Comrie, 1981:39) that at least consisted of Turkic, Mongolic, and Tungusic, three well-established genealogical groups. Ramstedt is the founder of Altaic linguistics in a scientific sense, though his works contain many problematic details. His introduction to Altaic linguistics was published posthumously (1952-1957). Poppe's contributions to Altaic linguistics are not less important. His comparative phonology, planned as the first part of a comparative grammar, appeared in 1960. An example of phonological correspondences according to Ramstedt and Poppe is the supposed development of the initial Altaic stop *p- into Korean p- and pb-, into Tungusic p- (Nanai), f- (Manchu), and b- (Evenki), into Mongolic *p- (Proto-Mongolic), b- (Middle Mongolian), f- (Monguor), and Ø- (Buryat, Oirat, Kalmyk, etc.), and into Turkic h- (Proto-Turkic, some modern languages) and Ø- (most modern languages). Ramstedt's and Poppe's arguments were largely accepted until they were challenged by G. Clauson (1956, 1962). Opponents such as J. Benzing and G. Doerfer expressed doubts even against this Micro-Altaic unit as a valid genealogical family.

Whereas the Altaicists regarded certain similar features as a common heritage from a protolanguage, others claimed that the similarities were the result of contact processes. Thus certain common features in Mongolic and Chuvash could go back to Proto-Altaic or had been borrowed into Mongolic from a language of the Chuvash type. Clauson had criticized the lack of evidence for a common basic vocabulary in Altaic. In his huge work on Turkic and Mongolic loanwords in Iranian, Doerfer (1963-1975) refuted the Altaic etymologies presented by Ramstedt, Poppe, and others, arguing that similarities that can be attributed to general typological principles or to areal diffusion must be excluded from genealogical comparisons.

A possible Altaic unity must have been dissolved about 3000 BC. The crucial question in Altaic comparative studies is by which methods common elements due to early contacts can be distinguished from elements inherited from a protolanguage. One problem is the scarcity of early data. Whereas Indo-European is attested already in the second millennium BC, there are no real Turkic sources prior to the 8th century (East Old Turkic inscriptions in the Orkhon valley, Inner Asia). The first Mongolic materials are found in The secret history of the Mongols (believed to be written around 1240 AD, partly based on older materials). The first substantial materials documenting Tungusic emerge centuries later (Even an abundance of written materials does not ensure an undisputable resolution of the genealogical dilemma. The criticisms of the “Language Tree“ paradigm are precipitated exactly by the abundance of written materials, which provides ample tools for argumentation disputing the soundness that paradigm. The universal general typological and areal diffusion principles of the Doerfer arguments are equally applicable to the development of the presumptive monolithic Proto-IE and, say, “Micro-Altaic“ family).
31

The Turkic-Mongolic-Tungusic Relationship

As for the relationship between Turkic and Mongolic, it has been possible to establish a number of convincing sound laws on the basis of words with similar sound shape and content, and to find certain corresponding derivational and grammatical suffixes. The question is how to judge these similarities. The earliest Turkic and Mongolic sources hardly show any common features except for intercultural words such as qagan “supreme ruler“ and tegri “heaven“. Middle Mongolian displays a number of words with similar Turkic equivalents. The few pairs of corresponding words do not, however, relate to the most significant parts of the vocabulary, i.e., numerals, kinship terms, and basic verbs, nouns, and adjectives. A few common elements are found in morphology. On the other hand, it is obvious that later Mongolic languages have converged with Turkic by giving up some old features, e.g., an inclusive vs. exclusive distinction in pronouns and verbs, grammatical gender in verb forms, agreement between the adjectival attribute and its head, and the option of postposed adjectival attributes (Coming from such distinguished expert as Dr. Lars Johanson, this description of the language is overwhelming. The first Mongolian, spied from the earliest 13th century source, after a millennia of Turkic symbiosis, displays typical traits familiar to us from the IE languages: “shel“ vs. “shla“ - “he went“ vs “she went“ in Russian, “chico hermozo“ vs.“chica hermoza“ - “attractive boy/girl“ in Spanish, and postposed “chico hermozo“ vs. “hermozo chico“ - the postposed “boy attractive“ is as bad in English as it is in Türkic. Adding to this the incompatibility of “the most significant parts of the vocabulary, i.e., numerals, kinship terms, and basic verbs, nouns, and adjectives“, the idea of a Turko-Mongolic “Language Tree“ seems to be a clear absurd. Given that not an absence of materials or scholarship feeds the Turko-Mongolic linguistic family notion, the only reason for this scholarly absurdity is a preconceived arimaspu notion of “us“ vs. “them“. To put it square, this attitude is as absurd as the division of philologists onto “Alataists“ and “non-Alataists“. No real scientist can be marred by a belief, however ingrained it may be).

Many similarities may thus be due to contact processes. There were close ties between Turkic and Mongolic as early as the middle of the first millennium BC. Borrowings in both directions had taken place since early times. With the rise of the Chingisid Empire in the 13th century, many Turkic varieties came under strong Mongolic influence. The impact lasted longer in areas of intensive contact, such as South Siberia and the Kazakh steppes. The lexical influence is particularly strong in Tuvan, Khakas, Altai Turkic, Kirgiz. Kazakh, etc. Look-alikes that occur only in typical contact zones cannot easily be used as evidence for genealogical relatedness.

Mongolic displays early layers of loanwords from several Turkic languages and has developed many structural traits under Turkic influence. Words common to Turkic and Mongolic, e.g., Bulgar-Mongolic correspondences, are regarded by Altaicists as true cognates and by non-Altaicists as Turkic loans in Mongolic. Some scholars consider the possibility that correspondences between Turkic and Mongolic go back to a common adstrate, some “language X“ that might have delivered loans to both groups. Tungusic words considered by Altaicists as Altaic are rather regarded by non-Altaicists as loans from Mongolic in certain contact areas. Similar derivational and grammatical suffixes are very scarce. Mongolic and Tungusic had been in contact for a long time prior to the first documentation of Tungusic. Except for recent Yakut loans in North Tungusic, there are hardly any plausible lexical correspondences between Turkic and Tungusic. In a non-Altaicist perspective, the overall Turkic-Mongolic-Tungusic relationship thus appears to be due to diffusion rather than to genealogical relatedness. According to this view, words common to all groups may have wandered along the path Turkic — Mongolic — Tungusic (What bothers Dr. Lars Johanson is that the presence of the evidence against genealogical relatedness in the corners of philological science is viewed as an absence of  the evidence for the genealogical relatedness, and thus a necessity to assemble a body of scientific evidence proving it. Accumulating linguistic cognates, say, doubling or quadrupling their number, would create on the ground a situation where the mass of the facts overweighs the underlying fundamental philological differences, which would be lightly brushed away).

After decades of discussions, the nature of the relationship between the Altaic languages is still controversial. Many common features are the result of recent contact, often limited to certain languages within the groups. The question is what reliable correspondences remain to justify the recognition of Altaic as a family in the sense of Indo-European or Semitic. There is no consensus as to whether the relatedness is proven, still unproven, or impossible. Some scholars argue that too few features are common to all three groups, and only to these groups. There are clear lexical and morphological parallels between Turkic and Mongolic, and between Mongolic and Tungusic, but not between Turkic and Tungusic. All three groups exhibit a few similar features, e.g., in the forms of personal pronouns, but similarities of this kind are found in different unrelated languages, in the rest of northern Eurasia and elsewhere. Today, however, compared to the 1960s, the fronts between Altaicists and non-Altaicists are not always as rigid. For example, the pronounced non-Altaicist Doerfer, who had criticized the proposed Altaic sound laws as being construed less strictly or even ad hoc, has accepted the above-mentioned development of *p- into Turkic b- and Ø-: e.g., *pat “horse“, bat (Khalaj, etc.), at (most Turkic languages). Doerfer expresses his appreciation of the achievements of the Altaicist Ramstedt in the following way: “We must be grateful to the ingenious founder of Altaistics as a science for discovering so many sound laws which are valid to this date“ (Doerfer, 1985: 135) (There is a sea of difference between admiring ingenuity of an inventor, and acceptance of the invention. Acceptance of even one invention that genetically links unrelated phenomena equates with an overall recognition of a genetical pedigree. Unlike the brothers Wrigt's invention, Ramstedt's invention does not need to fly, all it needs is a “recognition of the majority“ of the experts, even when they come loaded with qualifiers and criticisms. The “law“ invented by Ramstedt, like the invented *protoword *pat, can demonstrate how spurious phonetical inventions gain a legitimacy and long life when an anecdotal case is extrapolated into a phantom “law“. Clearly, Dr. Lars Johanson does not side with moving goalposts and sitting on the fences in a clear-cut situation).

Korean and Japanese

The most controversial point in recent discussions has been whether Korean and Japanese (with the closely related Ryukyuan language) should be regarded as members of an Altaic family. G. J. Ramstedt (1939, 1949) was the first scholar to attempt to prove a remote relationship between Turkic-Mongolic-Tungusic and Korean. Though his comparisons have been heavily criticized in more recent studies, N. Poppe considered Ramstedt to have identified at least 150 incontestable Korean-Tungusic-Mongolic-Turkic cognates.

32

Japanese has often been taken to consist of an Austronesian substratum and an Altaic superstratum. E. D. Polivanov (1924) argued that it is of hybrid origin, containing both Austronesian elements and continental elements that are also found in Korean and Micro-Altaic. In an early study, Ramstedt (1924) investigated possible links between Japanese and Altaic without reaching a clear final conclusion. Forty-two years later, S. E. Martin (1966) provided 320 etymologies relating Japanese to Korean on the basis of regular sound correspondences, which allowed him to reconstruct Proto-Korean-Japanese forms. R. A. Miller (1971), who established a set of sound correspondences to the Proto-Altaic phonemes reconstructed by Poppe (I960), clearly claimed Japanese to be one branch of the Altaic family. K. H. Menges (1975) took up a number of Miller's arguments and elaborated further on them. In his book on the Altaic problem and the origin of Japanese (1991), S. A. Starostin established sound arguments and established sound correspondences between Japanese, Korean, and Altaic on the basis of numerous lexical comparisons of Turkic, Mongolic, Tungusic, Korean, and Japanese lexical items. J. Janhunen (1992, 1994), however, pointed out some problems with the Altaic affiliation of Japanese, which he considers premature. He takes Japanese and Ryukyuan to form a distinct family of its own and the Old Koguryo language, once spoken on the Korean peninsula, to be a close relative of Japanese (I.e. there is no genetic link between Japanese-Korean and Altaic families. All lexical correspondences found by Starostin are of diffusion type, and the regular phonetic alterations pertain to the modifications internal to the Manchu family, of which Korean and Japanese are biologically genetic descendents. The lexical loanwords and phonetic transformations, obligatory for the loanwords borrowed from an unrelated language, independently of the extent of borrowing, can not establish a genetical link between languages belonging to fundamentally different linguistic families).

Bibliography