Janhunen J. The Mongolic Languages

Подождите немного. Документ загружается.

400 THE MONGOLIC LANGUAGES

although derivative relationships like 3 *gu.r : 30 *gu.t.ï/n would suggest that it may be

secondary in at least some of the items concerned.

It is another question whether the Para-Mongolic set for the teens was also used in

Khitan. The Khitan Small Script would suggest that the teens were expressed in terms of

additive compounds (10 + digit), rather than as lexicalized derivatives. On the other

hand, it is unclear whether the Khitan numerals of the first decade involved the use of the

simple numeral roots, or of derivatives of the Proto-Mongolic type. The current

(Chingeltei) ‘readings’ of the syllabic signs denoting the Khitan numerals are based

on the assumption that all the items were underived and monosyllabic. Most of these

‘readings’ are otherwise roughly compatible with the comparative information:

2 cur (more probably jur or jir), 3 gur, 4 dur, 5 taw (or tau), 6 nir (more probably nil),

7 dol (more probably dal), but the ‘readings’ proposed for 8 naim and 9 is are clearly

anachronistic and unlikely to be correct. Most interestingly, the ‘reading’ for 1 mas

suggests a root completely unknown in the Proto-Mongolic lineage.

Of the higher numerals, the items for the decades 20 *xorï/n and 30 *gutï/n were also

borrowed into Tungusic, as it seems, from Para-Mongolic. The Tungusic item 40 *deki

is likewise commonly assumed to be a Mongolic borrowing, but its connection with

Proto-Mongolic *dö.r.be/n : 40 *dö.c.i/n remains unclear. Among higher numerals, the

‘readings’ 100 jaw (or jau), 1,000 ming, and 10,000 tum have been proposed for Khitan

(Chingeltei), but the evidence remains to be verified.

The Tungusic shapes of the borrowed items for the teens and decades suggest that

most Para-Mongolic numerals, like their counterparts in the Proto-Mongolic lineage,

ended in the unstable */n. Whether this was also the case in Khitan, remains unclear as

long as the decipherment of the Khitan Small Script has not been completed.

OTHER FEATURES

Although little is known of the linguistic substance of Para-Mongolic, it may be pre-

sumed that at least Khitan was structurally very similar to its immediate eastern neigh-

bour Jurchen. Many of the typological peculiarities that distinguish the Jurchen-Manchu

lineage from the other Tungusic languages, are likely to be due to Khitan areal influence,

or to innovations shared with Khitan. On this basis, it may be presumed that Para-

Mongolic, like Jurchen-Manchu, was characterized by, for instance, a somewhat less

abundant (secondarily reduced) system of suffixal morphology, as compared with the

Proto-Mongolic lineage.

In the morphological system, Para-Mongolic nevertheless seems to have possessed all

the basic categories also known from Proto-Mongolic, i.e. case endings, verbal markers

for imperatives, finite indicative forms, participles, and converbs, as well as a system of

denominal and deverbal derivative suffixes. There do not appear to have been any suf-

fixes of personal marking (predicative personal endings or possessive suffixes). The

graphic details of the Khitan Small Script suggest that some suffixes were subject to allo-

morphic variation depending on the rules of vowel harmony (which may have been of

the rotated type).

Lexically, Para-Mongolic may also turn out to have shared many more items with

Jurchen-Manchu than can today be verified on the basis of comparisons with the Proto-

Mongolic lineage. In particular, it is likely that many of the older Chinese loanwords in

Jurchen-Manchu were actually borrowed through Khitan in the Liao-Jin political frame-

work. Such words include, for instance, the nominal type ending in the secondary suf-

fixal element .un, e.g. *pap.un ‘law’ > Manchu fafun, *lam.un ‘blue’ > Manchu lamun,

and possible *gur.un ‘state’ > Manchu gurun (later also borrowed, probably from

Manchu, into Modern Mongolic).

Obviously, much of what can be said today of Para-Mongolic remains speculative and

hypothetical. However, the keys are in our hands, and with more work on, especially, the

Khitan Small Script, the door to the world of Para-Mongolic can, without doubt, be

opened.

REFERENCES AND FURTHER READING

Arapov, M. V. (1982) ‘Leksika i morfologiya tekstov malogo kidan’skogo pis’ma’, in Zabytye sis-

temy pis’ma: Materialy po deshifrovke, Moskva: Nauka, pp. 211–39.

Bazin, Louis (1950) ‘Recherches sur les parlers T’opa (5e siecle après J.C.), T’oung Pao

39: 228–329.

Boodberg, Peter A. (1936) ‘The Language of the T’o-pa Wei’, Harvard Journal of Asiatic Studies

1: 167–85.

Chingeltei [Qinggeertai] (1997a) ‘Guanyu Qidan Wenzi de Tezhan’, in Asia Seominsu ui Munja,

Seoul: Kugyeol Hakhoe & Daehaksa, pp. 101–30.

Chingeltei [Qinggeertai] (1997b) ‘Qidanyu Shuci yu Qidan Xiaozi Biandufa’, Altai Hakpo

7: 143–52.

Chingeltei [Chingeltei] (1997c) ‘Kidan xel xiigeed mongol xelnii arwan xoyor zhil’, Altai Hakpo

7: 156–60.

Chingeltei [Cavggaldai] (1998) ‘Gidav Baqhe vUisug uv Sudulqhav u Bajidal vca (Nuiguburilagu

Ma’teriyal)’, paper presented at the 3rd International Conference of Mongol Studies, Huhehaote.

Chingeltei [Qinggeertai], Liu Fengzhu, and Chen Naixiong (1985) Qidan Xiaozi Yanjiu, Beijing:

Zhongguo Shehui Kexue Chubanshe.

Doerfer, Gerhard (1985) Mongolo-Tungusica [= Tungusica 3], Wiesbaden: Otto Harrassowitz.

Doerfer, Gerhard (1992) ‘Mongolica im Alttürkischen’, in Bruno Lewin zu Ehren 3, Bochum,

pp. 30–56.

Doerfer, Gerhard (1993) ‘The Older Mongolian Layer in Ancient Turkic’, Türk Dilleri

Araªt¦rmalar¦ 1993/3: 79–86.

Franke, Herbert (1969) ‘Bemerkungen zu den sprachlichen Verhältnissen im Liao-Reich’,

Zentralasiatische Studien 3: 7–43.

Franke, Herbert (1982) ‘Randnotizen zu einigen Worten der Khitansprache im Lichte neuerer

Arbeiten’, Acta Orientalia Hungarica 36: 173–82.

Haneda, Toru [Tooru] (1925) [1957] ‘Kittan Moji no Shin Shiryou’, Shirin 10 #1, reprinted in

Haneda Hakase Shigaku Rombunshuu 2: 420–34, Kyouto.

Janhunen, Juha (1993) ‘The Teens in Jurchen and Manchu Revisited’, in Festschrift für Raija

Bartens [= Mémoires de la Société Finno-Ougrienne 215], Helsinki, pp. 169–84.

Janhunen, Juha (1994) ‘On the Formation of Sinitic Scripts in Mediaeval Northern China, Journal

de la Société Finno-Ougrienne 85: 107–24.

Kara, György (1975) ‘Apropos de l’inscription khitane de 1150’, Annales Universitatis Scientiarum

Budapestinensis, Sectio Linguistica 6: 163–7.

Kara, György (1986–7) ‘On the Khitan Writing Systems’, Mongolian Studies 10: 19–24.

Kuz’menkov, Ye. A. (1988) ‘Mongol’skie elementy v man’chzhurskom i dialektnaya baza

staromongol’skoi pis’mennosti’, Voprosy yazykoznaniya 1988/4: 117–22.

Kuz’menkov, Ye. A. (1997) ‘Kidan’skii yazyk’, in Mongol’skie yazyki – Tunguso-man’chzhurskie

yazyki – Yaponskii yazyk – Koreiskii yazyk, Yazyki Mira, Moskva: Rossiiskaya Akademiya Nauk

& Izdatel’stvo Indrik, pp. 87–90.

Ligeti, Louis (1960) ‘Les anciens éléments mongols dans le mandschou’, Acta Orientalia

Hungarica 10: 231–48.

Ligeti, Louis (1970) ‘Le tabghatch, un dialecte de la langue sien-pi’, in Louis Ligeti (ed.)

Mongolian Studies, Budapest: Akadémiai Kiadó, pp. 265–308.

PARA-MONGOLIC 401

Liu Fengzhu (1999) ‘Seventy Years of Khitan Small Script Studies’, in Juha Janhunen and Volker

Rybatzki (eds) Writing in the Altaic World [= Studia Orientalia 87], Helsinki, pp. 159–69.

Luwsandendew [Luvsandèndèv], A. (1997) ‘Syan’biiskii yazyk’, in Mongol’skie yazyki – Tunguso-

man’chzhurskie yazyki – Yaponskii yazyk – Koreiskii yazyk, Yazyki Mira, Moskva: Rossiiskaya

Akademiya Nauk & Izdatel’stvo Indrik, pp. 144–7.

Okada, Hidehiro (1962) ‘Color-names in Manchu’, in Nicholas Poppe (ed.) American Studies in

Altaic Linguistics [= Indiana University Publications, Uralic and Altaic Series 13], Bloomington

and The Hague, pp. 225–7.

Probst, M. A. et al. (1986) ‘Slovar’ kidan’skoi leksiki XI–XII vv.’, in Drevnie sistemy pis’ma:

Etnicheskaya semantika, Moskva: Nauka, pp. 182–231.

Rozycki, William (1994a) Mongol Elements in Manchu [= Indiana University Publications, Uralic

and Altaic Series 157], Bloomington.

Rozycki, William (1994b) ‘Primary *h in Mongol: The Evidence Considered’, Central Asiatic

Journal 38: 71–9.

Starikov, V. S. (1982) ‘Prozaïcheskie i stixotvornye teksty malogo kidan’skogo pis’ma XI–XIII

vv.’, in Zabytye sistemy pis’ma: Materialy po deshifrovke, Moskva: Nauka, pp. 99–210.

Toyoda, Goro (1964) ‘An Analysis of the Major Ch’i-tan Characters’, Memoirs of the Research

Department of the Toyo Bunko 23: 119–35.

Toyoda, Goro [Fengtian Wulang] (1998) ‘Qidan Xiaozi dui Siji de Shenghu’, Minzu Yuyan 1998/1:

78–80 (81).

Viktorova, L. L. (1961) ‘K voprosu o naimanskoi teoriï proïsxozhdeniya mongol’skogo

literaturnogo yazyka i pis’mennosti (XII–XIII vv.)’, Uchenye zapiski LGU 305 [Seriya

vostokovedcheskix nauk 12], Leningrad: Leningradskii Gosudarstvennyi Universitet, pp. 137–55.

Vovin, Alexander (2000) ‘A Modest Proposal on the Decipherment of the Khitan-Chinese Bilingual

Text of 1134 (The Langjun Inscription)’, paper presented at the 43rd Annual Meeting of the

Permanent International Altaistic Conference, Chateau Pietersheim, Belgium.

Wittfogel, Karl A. and Fêng Jia-Sheng (1949) History of Chinese Society: Liao (907–1125)

[= Transactions of the American Philosophical Society, New Series 36], Philadelphia.

402 THE MONGOLIC LANGUAGES

403

CHAPTER TWENTY

TURKO-MONGOLIC

RELATIONS

Claus Schönig

Turkic and Mongolic exhibit many structural similarities and a great number of lexical,

morphological, and syntactic correspondences. These common features are the principal

object of the field known as comparative Altaic linguistics. To explain the situation, two

frameworks have been proposed: the genetic framework and the areal framework. In the

genetic framework, the similarities between Turkic and Mongolic are largely assumed to

be due to an original genetic relationship between the two groups of languages. This

framework, also known as the Altaic Hypothesis, operates with the postulation of a com-

mon Turko-Mongolic protolanguage, termed Proto-Altaic. Apart from Turkic and

Mongolic, other groups of languages, notably Tungusic, but also Korean and Japanese

(Japonic), have been regarded as deriving from Proto-Altaic. By contrast, the areal model

explains the similarities between the ‘Altaic’ languages as being basically due to long-

lasting and intensive contacts.

Since all scholars working in the field of comparative Altaic linguistics recognize the

areal factor, it is mainly a question of what role, exactly, is assigned to the genetic frame-

work. Some scholars see no evidence for a genetic relationship between the ‘Altaic’ lan-

guages, while others tend to explain a large part of the observed similarities as direct

genetic heritage. Still others recognize the possibility of a distant genetic relationship, but

prefer to explain at least most of the similarities as being areally conditioned. Generally,

the more carefully the areal factor has been investigated, the smaller the size of the residue

open to the genetic explanation has tended to become. According to many scholars it only

comprises a small number of monosyllabic lexical roots, including the personal pronouns

and a few other deictic and auxiliary items. For these, other possible explanations have

also been proposed. Most importantly, the ‘Altaic’ languages do not seem to share a

common basic vocabulary of the type normally present in cases of genetic relationship.

In the general context of comparative Altaic linguistics, the Turko-Mongolic relations

are of crucial importance. Not only was the Altaic Hypothesis originally based mainly on

the Turko-Mongolic comparative corpus, but also the potential disqualification of the

genetic framework depends on how this corpus is interpreted. Moreover, whatever conclu-

sion is reached concerning the Turko-Mongolic linguistic similarities, it will have a con-

siderable significance to the understanding of the early ethnic, cultural, and political history

of the entire Altaic sphere (Central Eurasia and Northeast Asia). Many of the lexical

similarities between the ‘Altaic’ languages are, in fact, connected with cultural diffusion.

DATA AND SOURCES

Although the Altaic Hypothesis dates back to eighteenth-century language comparisons,

it was first presented in a coherent formulation by G. J. Ramstedt (1952–66). Apart from

Turkic and Tungusic, Ramstedt connected Mongolic genetically also with Korean, but

not with Japanese. A more or less identical version of the hypothesis was adopted by

Nicholas Poppe (1960, 1965), while more extended (Macro-Altaic, Ural-Altaic, Nostratic)

frameworks, with the inclusion of Japanese (Japonic) and /or a selection of other

language families, have been supported by K. H. Menges (1968), R. A. Miller (1971),

S. A. Starostin (1991), and others. The basic problem common to all of these approaches

is that they tend to underestimate the impact of borrowing, especially between Turkic and

Mongolic, on the one hand, and Mongolic and Tungusic, on the other.

The recognition of the areal framework becomes the more important the further back

we move on the time scale. Contacts between relatively late and historically well-

documented stages of the ‘Altaic’ languages have been studied by even the most ardent

adherents of the Altaic Hypothesis like Poppe (1962, 1969), who (1965: 159) admits that

‘it would not be an exaggeration to say that about twenty-five percent of the Mongolian

[Mongolic] vocabulary is of Turkic origin’. However, the Altaic Hypothesis tends to

ignore the possibility of pre- and protohistorical contacts between the undocumented

(reconstructed) proto- and pre-protolanguages underlying Mongolic, Turkic, and

Tungusic. In reality, most of the Turko-Mongolic comparative corpus can be explained

as the result of early language contacts. This point of view has been supported by, for

instance, Johannes Benzing (1953) and András Róna-Tas (1976).

The early lexical contacts between Turkic and Mongolic have been studied in detail

by several scholars, including, most importantly, Sir Gerald Clauson (1960, 1962),

Gerhard Doerfer (1963–75), L. V. Clark (1977), and A. M. Shherbak (1996, 1997), all of

whom are sceptical of the Altaic Hypothesis. The theoretical premises of the areal frame-

work have been formulated by Doerfer (1988), while Róna-Tas (1973, 1982, 1990, 1998)

has clarified several crucial chronological issues. Some details of structural interaction

are discussed by Marcel Erdal (1991, 1998) and Lars Johanson (2002). In a wider areal

context, there is a vast literature on the diffusion of cultural vocabulary in the Altaic

sphere, as exemplified by the studies of Louis Ligeti (1950) and Volker Rybatzki (1994).

The ethnohistorical and taxonomic background of the Turkic languages, with direct

implications to the issue of the Turko-Mongolic relations, has most recently been

analysed by P. B. Golden (1992) and Claus Schönig (1997–8).

Although the contacts between Turkic and Mongolic have always taken place bilater-

ally, the pre- and protohistorical cultural, political and demographic circumstances seem

to have favoured borrowing mainly from Turkic to Mongolic, while in later times, begin-

ning with the historical Mongols, the predominant flow of influence was from Mongolic

to Turkic. Locally, the patterns of interaction may well have been more complex, and

much detailed work remains to be done in the field of language-to-language contacts in

the various regions of Turko-Mongolic adjacency and cohabitation. Special monographic

treatments of Mongolic loanwords exist for Yakut (Ka)uz

yn´ski 1962), Tuva (Tatarincev

1976), Northeastern Turkic (Rassadin 1980), and Western Oghuz (Schönig 2000).

THE HISTORICAL BACKGROUND

The Altaic Hypothesis involves several problems for the Turko-Mongolic comparisons,

for in spite of their common features Turkic and Mongolic have also profound differ-

ences. To explain these differences, the supposed Altaic protolanguage would have to

have dissolved very early – apparently several thousands of years ago. This, on the other

hand, is not compatible with the categories of vocabulary most typically shared by Turkic

and Mongolic, for much of the shared vocabulary is connected with relatively recent

404 THE MONGOLIC LANGUAGES

cultural innovations, which can even, with some accuracy, be dated and documented

historically and archaeologically.

A more plausible explanation is therefore offered by the assumption of a network of

linguistic contacts, which have united the ‘Altaic’ languages since ancient times up to the

present day. External contacts with Uralic and Indo-European suggest that Proto-Turkic

was once the westernmost member of this network. On the basis of historical informa-

tion it may be concluded that the period of the most intensive early contact between

Turkic and Mongolic coincided with the appearance of the protohistorical ethnopolitical

entities of Xiongnu and Donghu in the regions north of China (Mongolia and

Manchuria). The Xiongnu became politically dominant in the steppes around 300

BC, and

although the linguistic affiliation of the Xiongnu proper is still a matter of dispute, their

political confederation certainly contained a significant Turkic component. By both

ethnohistorical and linguistic considerations this component may in the first place be

identified with the Bulgharic (Bulghar Turkic) branch of Turkic, today represented by the

Chuvash language in the Volga region.

The actual linguistic material deriving from the Xiongnu is scarce and extremely

controversial. Most of the lexical items concerned are cultural terms, which may or may

not be Turkic. Two Xiongnu words subsequently found in almost all Turkic and

Mongolic languages are chengli †tängri ‘god; heaven’ (Proto-Mongolic *tenggeri) and

wolute †ordo ‘palace guard’ (for ‘palace’, Proto-Mongolic *ordu). Neither of these

words has a reliable internal etymology in either Turkic or Mongolic, and they may well

derive from a third (unknown) language. The Turkic component of the Xiongnu is, how-

ever, unambiguously signalled by a number of Bulgharic loanwords in Proto-Samoyedic,

such as *yür ‘hundred’. The Bulgharic (Proto-Bulgharic) speakers are likely to have

entered Southern Siberia, the location of Proto-Samoyedic, not earlier than the last cen-

tury

BC. At the same time, a number of local words, notably *kadï ‘conifer’ (> Chuvash

xïra

„

~ xïr ‘birch’), were borrowed from Proto-Samoyedic into Bulgharic.

In the second century

AD, the Xiongnu were overcome by the Xianbei, who subse-

quently dominated the steppes north of China during several centuries. The Xianbei

derived from the context of the Donghu, who are likely to have contained the linguistic

ancestors of the Mongols. Later branches and descendants of the Xianbei include the

Tabghach and Khitan, who seem to have been linguistically Para-Mongolic. True, a few

lexical items recorded from Tabghach, most notably the actor nouns †kapag.cin ‘door-

keeper’, †bitig.cin ‘secretary’, and †a[sh].cin ‘cook’, may be analysed as basically

Turkic, but they might well have represented early Turkic borrowings in Para-Mongolic.

Although the actor noun suffix *.ci/n is attested in both Turkic and Mongolic, it need not

be originally either Turkic or Mongolic, for actor noun suffixes appear to be universally

liable to be borrowed (as also in many European languages).

Opinions differ widely as to what the linguistic impact of the Xianbei period was.

Some scholars (like Clauson) have preferred to regard the Xianbei and Tabghach as

Turks, or even as Bulghar Turks, with the implication that the entire layer of early Turkic

borrowings in Mongolic would have been received from the Xianbei, rather than from

the Xiongnu. However, since the Mongolic (or Para-Mongolic) identity of the Xianbei is

increasingly obvious in the light of recent progress in Khitan studies, it is more reason-

able to assume (with Doerfer) that the flow of linguistic influence from Turkic (or

Bulghar Turkic) into Mongolic was at least partly reversed during the Xianbei period,

yielding the first identifiable layer of Mongolic (or Para-Mongolic) loanwords in Turkic.

Items with Mongolic roots and/or suffixes and apparently borrowed into Turkic in this

period include Old Turkic balbal ‘statue (of a slain enemy)’

← Mongolic *barï.mal

TURKO-MONGOLIC RELATIONS 405

‘structure’ (from *barï- ‘to grab; to construct’), kertü ‘true’ ← Mongolic *gere.tü ‘evident’

(from *gere ‘light’), qarghu ‘watchtower’

← Mongolic *kara.xu (from *kara- ‘to watch’),

and yalawac ‘envoy’ (< *yala.ba.ci ‘invitee’, from Mongolic jala- ‘to invite’). In many

other cases criteria for the direction of borrowing are missing, as in Turko-Mongolic

*kom (qom) ‘a piece of felt placed under the pack on a camel’.

Assuming that the Xianbei and Tabghach were, indeed, linguistically Mongolic (Para-

Mongolic), it is not immediately clear what contemporary ethnic group represented the

Turks. One possible identification is offered by the Ruanruan, who appeared in Mongolia

at the end of the fourth century. The Ruanruan have conventionally been connected with

the Avar, who appeared in Eastern Europe in the middle of the sixth century, but the con-

nection remains both historically and linguistically unconfirmed. The European Avar

have been identified variously as Turks or Mongols. Most recently, it has been suggested

(by Róna-Tas) that they were Bulghar Turks, which would make them descendants of the

Xiongnu. The information on the Avar language consists of names like Bayan (Mongolic

*bayan vs. Turkic *bay ‘rich’) and well-known titles like tarcan (Mongolic *darkan)

and caganus (Mongolic *kaxan vs. Turkic *kagan), none of which is diagnostic enough to

allow firm conclusions.

The first people to use the title *kagan (qaghan) ‘ruler, emperor’ were the Xianbei.

Later on this title was employed by both the Tabghach and the Ruanruan, as well as by

the Khitan and the historical Turks. The Mongols seem to have adopted it from the Turks.

Although the etymology of *kagan remains unclear, it belongs to a distinct type of nouns

ending in n, many of which may have entered Turkic from Mongolic (or Para-Mongolic).

This group comprises Turkic tarqan, as well as qatun ‘queen; lady’, tegin ‘prince’,

qalqan ‘shield’, atan ‘gelded camel’, toghan ‘[kind of] hawk’, lacin ‘gyrfalcon’, and

colban ~ colpan ‘the planet Venus’. In all of these cases, a foreign origin is also suggested

by the semantics (administrative concepts, hunting birds and other domesticated animals,

astronomical terms). There are also many other Turkic words that for phonological, mor-

phological, and/or semantic reasons might be early borrowings from Mongolic. Possible

cases include: Turkic taluy ‘ocean’ (Mongolic *dala.i), bughday ‘wheat’ (Mongolic

*buxuda.i), turumtay ‘hawking bird’ (with the Mongolic suffix *.tA.i), tuturqan ‘rice’.

In the middle of the sixth century the Turkic group bearing the ethnonym Türk

crushed the Ruanruan and gained control of the eastern steppes for the next few hundred

years. The subsequent Türk empires at times also controlled Mongolic and Para-

Mongolic peoples, including the Khitan, who copied political and organizational terms

from Turkic. During this period, the ancestors of the historical Mongols are likely to have

been contained within the entities known by the names Otuz Tatar (Shiwei) and Toquz

Tatar (Southern Shiwei), located east and southeast of Lake Baikal. West and north of

the lake were the Turkic Üc Qurïqan, the linguistic ancestors of the Yakut. In 742 the

Türk were defeated by the likewise Turkic confederation of the Uighur, who, in turn,

were pushed aside by the Ancient Kirghiz in the 840s. Some Uighur tribes took refuge

with the Otuz Tatar, but most of them withdrew to the oases of Eastern Turkestan. The

Uighur then never returned to the steppes, even when they were invited by the Khitan,

who had overcome the Kirghiz in the 920s.

In the twelfth century, part of the Khitan, subseqently known as the Black Khitan

(Qara Qïtay), migrated westward to Central Asia and became Turkicized. In Mongolia,

the immediate linguistic ancestors of the historical Mongols spread Mongolic (Pre-Proto-

Mongolic) speech to territories previously held by Turkic speaking populations. The

Mongols mainly occupied the basins of the rivers Orkhon and Kerulen, but the closely

related Kereit and Naiman tribes expanded further to the west. Both the Kereit and espe-

cially the Naiman may have contained unassimilated Turkic elements, as is suggested by

406 THE MONGOLIC LANGUAGES

the occurrence of Turkic names and titles among them. East of the Mongols lived the

Tatar, while to the north there were the Merkit, two groups that are likely to have been

linguistically Mongolic (or Para-Mongolic), but that may also have contained a Turkic

component. In an apparent reference to the Turkic elements in Mongolia, the eleventh-

century scholar Maúmûd al-Kâšgarî mentions that the Tatar, as well as the Yabaqu and

Qay, understood Turkic, though they also had a (Mongolic or Para-Mongolic) language

of their own.

The rise of the Mongol empire in the thirteenth century involved political and cultur-

al changes that brought the Turko-Mongolic relations to a new stage. At the same time

as the Mongols (and Para-Mongols) were unified under the relatively homogeneous

Proto-Mongolic language, the Turkic tribes were arranged according to patterns that

anticipated the emergence of the modern Turkic branches and languages. Many Turkic

and Mongolic groups were removed from their earlier homelands, and new and mixed

tribal units came into existence. As a consequence of this mixing, Mongolic (and

Para-Mongolic) ethnonyms are today widespread among Turkic populations, and vice

versa. Mongolic ethnonyms still used as tribal names by the Turks include Merkit

(Turkmen), Qïdat (South Siberian Turks), as well as Kerey(it) and Tatar (Kipchak).

Another wave of Mongolic impact on Turkic, especially on Central Asian and South

Siberian Turkic, was connected with the rise of the Oirat empire in the fifteenth century.

At the beginning of the seventeenth century the Kalmuck tribes reached the Lower Volga

region. In 1678, the Aqtaghlïq Khojas invited the Oirat to Kashgaria in Eastern Turkestan.

Southern Siberia came under Oirat control already in the fourteenth century, soon after the

end of the Yuan dynasty. In the sixteenth century parts of the Sayan region fell under

Khalkha influence. In the seventeenth century the Junghar could reinforce Oirat power,

but the Tuva region maintained its contacts with the Khalkha. Written Mongol remained

the literary language of Tuva until the early 1930s. In the eighteenth century the Junghar

removed large parts of the Siberian Kirghiz population from Southern Siberia.

MONGOLIC AND BULGHAR TURKIC

There are several criteria that allow the Bulgharic loanwords in Mongolic to be identi-

fied. For one thing, Mongolic often has a native word synonymous with a Bulgharic bor-

rowing, cf. e.g. Mongolic *ele.sü/n ‘sand’ vs. *kumaki id.

← Bulgharic *kuma.kï =

Common Turkic *kum (qum). Moreover, the borrowings often show a specialized mean-

ing, whereas the native words have a more general semantic profile, cf. e.g. Mongolic

*xüsü/n ‘hair’ vs. *kilga.su/n ‘hair of a horse’

← Bulgharic *kïlka = Common Turkic *kïl

(qïl) ‘hair’. Other items have an internal etymology on the Turkic side, cf. e.g. Mongolic

*ikire ‘twin/s’

← Bulgharic *ikire = Common Turkic *eki.z, derived (plural) from *eki

‘two’. In still other cases, the Bulgharic original ultimately appears to derive from a third

language, notably Tocharian, cf. e.g. Mongolic *xüker ‘ox’

← Bulgharic *xekür =

Common Turkic *(x)öküz id., Mongolic *jer ‘weapon/s’ < ‘bronze’

← Bulgharic *jer =

Common Turkic *yäz ‘bronze’, possibly borrowed from the cognates of Proto-Tocharian

*xokso ‘ox’ and *yes ‘gold’, respectively.

Two culturally and historically important semantic fields for which Bulgharic

borrowings are abundantly attested are the terminologies of animal husbandry and

metal working. Bulgharic animal names in Mongolic include, apart from *xüker

‘ox’, Mongolic *eljige/n ‘donkey’ (Common Turkic *eshkäk), *biraxu ‘calf’ (Common

Turkic *buzag/u), *koni/n ‘sheep’ (Common Turkic *kony), *ajirga ‘stallion’ (Common

Turkic *adgïr), and others. Relevant metal terms include, apart from *jer

‘bronze’, Mongolic *korgoljin ‘lead’ (Common Turkic *korgashun), *alta/n ‘gold’

TURKO-MONGOLIC RELATIONS 407

(Common Turkic *altun), and *siri- ‘to smelt (ore)’ (Common Turkic *sïz- ‘to melt’).

Some Mongolic metal terms, e.g. *temür ‘iron’ (Common Turkic *tämür) and *jes

‘bronze’ (Common Turkic *yäz) were borrowed from Turkic only later, while others have

been variously reborrowed from Mongolic into Turkic, especially into Northeastern

Turkic. Mongolic also has a number of metal terms of other origin, notably *mönggü/n

‘silver’ and *küril ‘bronze’, but the basic picture is one of intensive and continuous inter-

action with Turkic, starting with the Bulgharic period.

The most important property of the Bulgharic loanwords in Mongolic is that they

carry a number of diagnostic phonological characteristics, which distinguish them from

the later Turkic (Common Turkic) elements. Owing to the relatively large size of the

Bulgharic loanword corpus, it is possible to establish a set of regular correspondences

between Turkic (Bulgharic) and Mongolic. It is largely these correspondences that have

served as the basis for the Altaic Hypothesis in the past. However, a closer look at the

features involved shows that they represent either archaisms or innovations that once

characterized the Bulgharic (Proto-Bulgharic) branch of Turkic. Some of the relevant

features are listed here:

(1) Rhotacism-lambdacism, by which the correspondence Mongolic r l vs. Common

Turkic z sh is understood, as in Mongolic *kuxur [type of lute] vs. Common Turkic

*kopuz (qopuz), Mongolic *töl.ge ‘fortune-telling’ vs. Common Turkic *tüsh ‘dream’. In

some cases, the equivalent of Common Turkic sh appears as lj or lb in Mongolic, as in

Common Turkic *taz+bashï ‘sparrow hawk’ vs. Mongolic *tarbalji, Common Turkic

*kashuk (qashuq) ‘spoon’vs. Mongolic *kalbuga/n. Various explanations have been pro-

posed for these correspondences, but basically they seem to involve a contextually con-

ditioned innovation (*ssh> r l) in Bulgharic. Rhotacism-lambdacism has been inherited

by Chuvash, and it is consistently present in the Bulgharic loanwords not only in

Mongolic, but also in Samoyedic (as well as in Hungarian). A chronological point of ref-

erence is offered by Turkic *izängä ‘stirrup’, a cultural term archaeologically datable to

c.400–300

BC, which appears with rhotacism in Chuvash, possibly also in Mongolic

(*dörüxe). In the Common Turkic branch, rhotacism-lambdacism is generally absent, but

it is occasionally observed in preconsonantal position, which makes the dating of certain

loanwords problematic, cf. e.g. Mongolic *buxas ‘pregnant’ (from Common Turkic

*bugaz id.) vs. *buxar.la- ‘to cut the throat’ (from either Bulgharic or Common Turkic,

cf. Common Turkic *bogaz ‘throat’).

(2) The presence of initial *d *n (*ny) in Mongolic vs. Turkic *y, as in Mongolic

*dayir ‘brown’ > ‘deer’ vs. Common Turkic *yagïz ( yaghïz) ‘brown’, Mongolic *nidur-

ga ‘fist’vs. Common Turkic *yudruk (yudruq). For these features, Chuvash does not dif-

fer from Common Turkic, which means that the Bulgharic language from which

Mongolic received the early layer of borrowings represented a more archaic (Proto-

Bulgharic) stage, still close to the common protolanguage of Bulgharic and Common

Turkic. It may be concluded that Bulgharic came to share a number of innovations with

Common Turkic even after rhotacism-lambdacism had divided the original protolan-

guage (Pre-Proto-Turkic) into two branches.

(3) The presence of stem-final vowels in Mongolic vs. Turkic zero, as in Mongolic

*kona- ‘to spend the night’ vs. Common Turkic *kon- (qon-), Mongolic *balga :

*balga.su/n ‘town’ vs. Common Turkic (*balk >) *balïk (balïq). For this feature, also,

the information from Chuvash is inconclusive, while Mongolic alone preserves the

original (Pre-Proto-Turkic and Proto-Bulgharic) syllable structure. (It goes without

408 THE MONGOLIC LANGUAGES

saying that the reduction and loss of vowels in non-initial syllables in several Modern

Mongolic languages is a separate and much later phenomenon.)

(4) The presence of initial *x in Mongolic as a segmental trace of *p, as in Mongolic

*pürü- > *xürü- ‘to rub’ vs. Common Turkic (*xüz- >) *üz- ‘to tear’. The development *p

> *x is clearly an areal phenomenon shared by both Mongolic (and Tungusic) and the

two branches of Turkic, but on the Turkic side it has been followed by an almost com-

plete loss of *x (with the important exception of Khalaj), while in Mongolic *x is syn-

chronically documented in several languages. The ultimate reconstruction of *x as *p is

based on systemic considerations as well as external etymologies, e.g. Mongolic

*xo( y)ima.su/n ‘stocking/s, boot/s’ = Proto-Samoyedic *päyma id. (both possibly bor-

rowed from a Bulgharic source). It may be noted that in medial position *p is also rep-

resented as Mongolic (Proto-Mongolic) *x, while Turkic has *p (> *b), as in Mongolic

*köper > *köxer ‘proud’ > ‘happy’ vs. Turkic *küpez (> *kübez) ‘proud’, Mongolic

*köperge > *köxerge ‘bridge’ vs. Turkic *köprüg (> *köbrüg).

(5) The correspondence Mongolic *A-A vs. Turkic *A-U, which can most easily be

derived from Bulgharic (and Pre-Proto-Turkic) *A-O, as in Mongolic *altan ‘gold’ vs.

Turkic *alton > *altun. The rules of vowel harmony in both Mongolic and Turkic have

secondarily eliminated the synchronic possibility of the combination *A-O, but the com-

parative data confirm the previous existence of this combination in Turkic. It remains,

however, questionable whether the combination also existed in Mongolic at the time of

the contacts (*altan < Pre-Proto-Mongolic *alton?), or the borrowings were immediately

adapted to the Mongolic rules of vowel harmony (Pre-Proto-Mongolic *altan

←

Bulgharic *alton). It is also unclear what the exact chronology of the development *A-O

> *A-U on the Turkic side was, for the Turko-Mongolic correspondence *A-U vs. *A-A is

also attested in a number of words that may actually belong to a later (Common Turkic)

layer of borrowings, e.g. Mongolic *tusa ‘advantage, benefit’ vs. Turkic *tuso > *tusu.

There are indications that the combination *A-O may still have been synchronically valid

for at least some forms of Old Turkic (as preserved in the Brahmi and Tibetan scripts).

The correspondences listed here (1–5) illustrate the fact that the Bulgharic loanwords

in Mongolic provide information not only on the specific characteristics of the Bulgharic

branch, but also on the preceding stages of Turkic (Pre-Proto-Turkic), which would oth-

erwise remain beyond the reach of diachronic research. There are other correspondences

that do not allow an equally unambiguous interpretation, but that are nevertheless rele-

vant for the understanding of Turkic diachronic phonology. For instance, the representa-

tion of Common Turkic *sarïg ‘yellow’ as *sira in Mongolic has been explained as

suggesting that Bulgharic (Proto-Bulgharic) was characterized by the development

(diphthongization of long vowels) *saarV- > *sïarV- > *syarV-, which is apparently

confirmed by modern Chuvash shura

„

id. (with u < *a). While this may be the correct

explanation, it has to be admitted that there are still many unsolved issues about the

Turko-Mongolic vowel correspondences.

Another detail that cannot as yet be satisfactorily explained is the occasional repre-

sentation of Turkic initial *t as Mongolic *d, as in Turkic *tüsh ‘noon’ vs. Mongolic

*düli. This correspondence is possibly indicative of a very early date of borrowing, but

its phonological basis remains obscure, since both Turkic and Mongolic originally had a

distinction between two kinds of dental stop (strong *t vs. weak *d). In some cases, as

in Mongolic *dala( y)i ‘sea; ocean’ vs. Turkic *taloy > *taluy it might also be a question

of a borrowing from Mongolic into Turkic. In any case, the representation of Turkic *t

TURKO-MONGOLIC RELATIONS 409