JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
Akadémiai Kiadó is collaborating with JSTOR to digitize, preserve and extend access to Acta Orientalia
Academiae Scientiarum Hungaricae
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
Acta Orientalia Academiae Scientiarum Hung. Volume 51 (1-2), 85-130 (1998)
Robert L. Fisher
(Toronto)
A compound noun can be defined as two free morphemes, at least one of which is a n
have been combined into an accentual and semantic unit, such that no morphological affix
attached to the first element. In the oldest attested Altaic languages, such compounds are
finding points to a relatively isolating, accusative proto-language with the relationship
words shown primarily by word order and particles. Altaic' s rich derivational morpholog
the need for compounds. The lack of possessive compounds, and other features, hint at
active/stative type for Proto-Altaic. The development of compounds is due mainly to c
Indo-European languages and Chinese.
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
86 R. L. FISHER
"opaque". An oft-quoted ex
is blue", referring to a perso
In addition, Bluebeard is
a bahuvrīhi , literally, "muc
is obvious, as in bluebird ,
compound reflects a reduc
example is English pickpo
"[he] art makes" + nomina
pound derives from a com
sukrta "well-made, well-do
return to our English exam
a compound may be "additi
and ten. In Sanskrit these ar
nation.
Distinguishing compound n
of precise knowledge conc
inscriptions, as Tekin4 not
between colons":
ttümpamrb'umn'qYn'nsYmiqYn'mlVms: (KT E 1)
äcüm apãm bumïn qayan ištāmi qayan olurmiš
The practice of using colons was not, however, consistent. The only other direct
source of knowledge about accent is whatever can be inferred from the syncope of
vowels.
In Classical Uighur texts using the Brahmi script, word groups, von Gabain5
observes, are often written together:
On the other hand, in most manuscripts the plural suffix and case endings are written
separately.
In the Old Turkic texts that I examined, namely the Orkhon Inscriptions and
Hamilton's edition of Le conte bouddhique du mauvais et du bon prince (1971),
I found only one true noun + noun endocentric, descriptive compound: bäqgü taš
"memorial stone", literally "eternity (or: eternal) stone". The first element, bäqgü,
can be considered either a noun (eternity) or an adjective (eternal) since Altaic does
not distinguish these two grammatical classes (another example is Orkhon ädgü
"good" in general, but ädgü-q "your benefit" (KT E 24 = BK E 20).
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 87
Uteris < *īltēriš "the one who gathers the tribes" < il, or el, "tribe" + ter- "to gath
+ -s/š deverbal suffix, literally, "(the) tribe-gathering (one)", and ultim
derives from a clause "he gathers the tribes",
eletmis < *ēlētmiš "one who has organized the tribes" < el "tribe" + et- "to fo
organize" + -mis, a suffix forming verbal nouns denoting agents and act
(cf. igidmiš "one who has fed", qalmiš "those who have survived"),
eltäbär "one who has put the tribes in order" < el- "tribe" + tab- "to put in or
-är participial suffix.
Compound Gloss
yir/yer sub territory, land (literally, land + water)
bäd'iz yaraťíyma picture designer
barq itgtici house builder
kün batsïq sunset; west
kiin tuysïq sunrise; east
bitig tas inscription stone
bitig tas itgiiči stone inscriber
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
88 R. L. FISHER
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 89
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
90 R. L. FISHER
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 9 1
Mongolian
The oldest stratum of Mongolian is attested in The Secret History of the Mongols.
The following list of compounds is taken from Haenisch.20 The number in parenthe-
ses refers to the line number in de Rachewiltz.21
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
92 R. L. FISHER
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 93
23 Ibid., 74 §284.
24 Prof. T. Tekin, personal communication.
25 Brockelmann, Osttürkische Grammatik, 144ff.
26 F. Lessing, Mongolian-English Dictionary. Berkeley and Los Angeles 1960, s.v. mor
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
94 R. L. FISHER
The following list of Mongolian nominal compounds is taken from the Penta-
glot (Sanskrit-Tibetan-Mongolian-Manchu-Chinese) Dictionary of Buddhist Termi-
nology :30
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 95
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
96 R. L. FISHER
The sense of yurban čag derives from Skt trikālajnah (three + time + knowing
"omniscent", that is, knowing the three times (past, present, future).31
In the last example, the Mongolian expression translates a Sanskrit compoun
which is not given in the Pentaglot Dictionary: adhimukticaryâbhûmih (confi-
dence, propensity + performing + world), "one of the ten Bhūmis". The Mongolia
agrees with the Chinese equivalent: -f ~ iļk ^ (ten + world + name) in that it tran
lates the concept by giving the rank of that world: "name of the tenth world".
The remaining twelves entries involving numerals are more in the nature o
mnemonics: "the five powers" (tabun kūčtin), "the five organs" (tabun erketen
"the three doctrines" (diirben iineker) and the like.
It should be noted that Mongolian follows the general Altaic pattern in regard
to additive compounds:
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 97
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
98 R. L. FISHER
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 99
This evolution toward increased compounding may in part be due to Russian influ-
ence, yet it is clearly evident that Turkic has followed the same path from limited,
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
100 R. L. FISHER
transparent compounding in
them opaque, in the middle an
From Poppe's The Twelve
tavistara , the following comp
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 101
Manchu
From Nowak:47 The Tale of the Nišan Shamaness: A Manchu Folk Epic:
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
102 R. L. FISHER
Sibe
The Manchu compound adun morin "horse herd" presents the same un
pected word order, for an OV language, as the Classical Uighur sürüg ud "c
herd", discussed above. First of all, it is interesting to note that both the element
the Manchu compound are loanwords from Mongolian: adun < Mongolian adu
and morin has exactly the same form in both languages. In Mongolian, how
aduyun means "horses, herd of horses" and in neither The Secret History n
Lessing is the word used with morin. Instead, it is often collocated with ide'
ideyen) in a coordinate compound (dvandva) "horses (or herds) and food". It w
appear, then, that the borrowed Mongolian word for horses, adun, was sim
reinforced with another word for horse, itself ultimately a loanword from Mong
Note that yonin adun "sheep herd" follows the regular OV word order.
Considering the length of the texts {The Tale of the Nišan Shamanness , the
heroic epic), the number of compounds attested is quite small. If the same stories
been written in Homeric Greek or Sanskrit, the percentage of compounds
have been much higher. To illustrate this point, I counted the nominal compounds
48 S. Kaluzy ñski, Die Sprache des Mandschurischen Stammes Sibe aus der Gegend
Kuldscha. I. Band 1. F. Muromskis Sibenische Texte; 2. Wörterverzeichnis. Warsaw 1977
Orientalistyczne 25).
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 103
the first fifty lines of the Odyssey and found, including proper names, seven c
pounds. In the first seventeen lines of the Classical Sanskrit epic Naia and
yantï (an episode in the great Hindu epic, The Mahãbhãrata ), there were no less t
twenty.
From The Sino-Jurchen Vocabulary of the Bureau of Interpreters ,49 the fol-
lowing compound nouns have been gathered. Although it is not always easy to
decide, caiques on Chinese expressions have been excluded. This present glossary
was compiled in the Ming Dynasty and represents either late Jurčen or early Man-
chu, "a form of Manchu dating long before that language was first written in Mongol
script in 1599."50
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
104 R.L.FISHER
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 1 05
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
106 R.L.FISHER
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 107
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
108 R. L. FISHER
53 S i nor, ibid.
54 Cf. Kane, Sino-Jurchen Vocabulary , 119.
55 Cf. C. Watkins, The American Heritage Dictionary of Indo-Europe
1985, s.v. aus-1.
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 109
from the shamanic point of view.56 In contrast to Manchu, the Turkic directions,
summed up in the Kül Tigin Inscription (South 2), are quite different: ilgärü
tuysiqa birigärü kün ortusiqaru, quriyaru kün batsïqîqa yirïyaru tün
ortusïqaru "eastwards to the sunrise, southwards to the midday, westwards as far as
the sunset, and northwards to the midnight".57
Hezhen
From this sister language of Manchu, spoken by some 2000 people in Heilongjiang
Province of China, come the following compounds:58
The expression dau gaska apparently refers to the migratory habits of the wi
goose, a pattern of behavior it shares with the salmon.
Tungus59
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
110 R.L.FISHER
It is interesting to note
compounds of noun + ve
a rule are rare in Altaic.
Korean
Japanese
The list of compounds below has been compiled from the first two hundred pages of
Pierson' s Manyoshu Dictionary ,62 The Manyõshu is an anthology of songs dating
from the seventh to eighth centuries and is therefore the most ancient large corpus in
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 1 1 1
Old Japanese. The Manyöshü is rich in compounds, but in this list compounds show
ing nigori (voicing of the first consonant of the second element) have been autom
cally disqualified since the voicing is the residue of the earlier genitive (or adn
nal) suffix *-n-. This suffix violates the constraint in the definition of compound
no affix appear between the two members. (Sometimes a compound occurs wit
without nigori , for example, asa-duyu "morning dew"63 and asa-tuyu "Wem
Also excluded are Chinese loanwords (riku-shi "wrestler" < Chinese lìshì "str
man"). In general, compounds containing a color term or adjective are not listed, b
those with a verbal element have been included (sidari-yanagi "weeping willow
sidaru "to hang down").
The numbers in parentheses under "Location" refer to the page number
Pierson' s dictionary and his translation and commentary on the Manyöshü ;6
Roman numerals followed by a dash and a number indicate the book and the numb
of the poem within the book; the number after the slash (/) is the consecutive nu
of the poem. I have changed Pierson' s v' s to p's.
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
112 R.L.FISHER
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 1 1 3
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
114 R.L.FISHER
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 1 15
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
116 R.L.FISHER
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 1 1 7
attested in the earliest Altaic languages, in which two contiguous nouns rarely form
single accentual unit.
Syromiatnikov75 notes that in Old Japanese, simple nouns had no word-
forming suffixes. I would submit that this poverty in derivational suffixes is
reason why the language was rich in compounds. (See more below in the Conc
sion.) In any event, the vast majority of the compounds in the above sample a
endocentric (transparent), a fact that accords with the general Altaic pattern at
oldest layers. Some exocentric (opaque) compounds occur, and in greater numbe
than in the rest of Altaic: sori-gupi "beard stubble" (literally, "cut post"), koto-a
"prayer" (< koto "thing, matter, affair" + age "raising"), kumo-wi "clouds" (< ku
mo "cloud" + wi "seat"), and from Syromiatnikov:76 pi-ziri "ruler, sage" (literal
"the sun-knowing"), kanuti "blacksmith" (< "metal-beating"), utagë "drink
party" (literally, "hand-clapping"), mï-nasi-go "orphan" (< "without-body-child
abumi "stirrup" (< asi-pumi, literally, "foot-stamping").
Conclusion
At this juncture it may prove fruitful to take a closer look at the definition of com-
pound. Scholars working in the various Altaic languages have analyzed compound
mostly from the semantic point of view. They categorize compounds according
the meanings of their components: synonym + synonym, pairs of antonyms, pairs of
rhyming words.77 The approach taken here, however, leans more heavily on stru
turalism. In this vein, Street78 has pointed out that even from a structuralist point o
view, the concept of compound (which he terms "semi-compounds") is slippery: "
is difficult to give a real definition of semi-compounds: there is no sharp dividin
line between true compound stems, semi-compounds, idioms, and normal syntact
constructions... Formally, such semi-compounds fall somewhere between morph
ogy and syntax; semantically, somewhere between idioms and non-idiomatic coll
cations".79 In fact, compounds are best understood as a point along a spectrum o
nominal expressions, stretching from the single noun to compounds to syntagm
(word groups that are syntactically interchangeable with words [single or compound]
as elements in the clause or sentence).80 It should be noted that some authors use the
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
118 R.L.FISHER
1 . Genuine compounds
rare in the oldest attested
compounds, of which a m
tériš "the one who gathe
the tribes"; and eltäbär "o
Conte bouddhique du mau
Secret History of the Mong
while Twelve Deeds of th
the Nišan Shamanness ha
teen.The sole exception to
there are good reasons fo
not single units but rather
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 1 19
attest compounds as two words). Schmidt87 for Mongolian and Ramstedt88 for A
in general, had already noted this paucity of compounds. What Altaic does a
these ancient texts is syntagma composed of two, more rarely three or more, n
These nominal phrases (syntagma) can be endocentric, in other words, the
which is always the second, or last, noun, occurs within the compound. T
compounds are transparent, that is, their total meaning is clearly derivable from
constituents. In exocentric compounds the head is outside the compou
consequently the meaning cannot be derived from the constituents (presumably,
is what Street meant by "idioms"). In all compounds the typical OV word o
Altaic languages is shown in the pattern of the modifier preceding the h
examination of the lists collected above suggests that we are dealing with n
phrases and the occasional idiom.
2. Although the early Altaic languages are poor in compounds (or bet
nominal phrases), over time they have all developed huge inventories of com
as a perusal of the modern dictionaries will demonstrate.89 Two contiguous l
families, Indo-European and Sino-Tibetan, have followed a parallel track in
regard. Lehmann90 points out that compounds were relatively rarer in the oldes
(although certainly not a "virtual absence" in Hittite);91 a condition that he
back into Proto- and Pre-Indo-European. For Chinese, Dobson92 states that
sample of Late Archaic Chinese (fourth and third centuries BC), the "propor
compound words to single or free words... was never higher than three per
but rises perceptively in Han usage (second cent. B.C. - second cent. A.D.),
very much higher in Modern Chinese." In fact, the great Qing Dynasty dictiona
compound nouns (the Lián mián cídiàn )93 for Chinese of all periods runs to
volumes. Lessing94 thinks that the Buddhist terminology in Mongolian actually
closer semantic and structural affinity to Tibetan than to Sanskrit. In Ti
compounding is the principle means of word building and compound nouns form
largest group of substantives.95 There are at least three reasons for this rapid g
in compounds.
First of all, compounds have to be seen in the context of derivation, of which
they form a part, along with the other major means of creating new words: affixation,
loanwords, loan translations (caiques), and blends (like Manchu sifulu < sike +
fulhu, discussed above). The derivational affixes of the early Altaic languages
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
120 R.L.FISHER
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 1 2 1
Sha-ťo, with the Chinese noting intermarriage between Turks and Sogdians with
term za-hu "half-breeds" (also used of Sogdian merchants under Uighur prote
the eighth-century alliance of Sogdians and Uighurs who plundered Khotan
Kansu and the partial sedentarization of the Uighurs under Sogdian influence; also
the eighth century, the Kol Tegin inscription describing barčakar "emigré
Sogdian *prcykr)ì refugees from Bukhara, who would flee yet again to Penjikent
thence to the Türkíi;106 Sogdian artists were employed in Eastern Turkestan
ninth century. Of linguistic relevance is the fact that Sogdian was used as a
franca in Central Asia and Eastern Turkestan from the fourth century, possibly
the second, with bilingual Sogdian and Turkic (Orkhon and Uighur) inscrip
attesting to the equal political status of these languages in the eighth and
centuries;107 Manichaean, Christian and Buddhist Sogdian texts from the eig
the eleventh centuries, along with loanwords in Uighur and Mongolian, show
cultural importance of this Iranian language over a vast area of Centra
Furthermore, with the gradual demise of the Sogdian language and the turkificat
of its people, the West Iranian languages of Farsi and Dari became promine
Central Asia as the vehicles of Islam.108 Even with this sketchy summary of Iran
Turkic-Mongolian history it is evident that these peoples and their lang
experienced prolonged and intimate contact, leaving as their mark the use o
Sogdian alphabet, in modified form, as the writing system adopted by both Uigh
(who depended on Sogdian scribes and administrators) and Mongols (and even
by the Manchus).
Especially pertinent for our study are the "open compounds" of Sogd
term coined by W. B. Henning and defined as "a collocation of a group of words,
proaching the status of a compound, without having reached its stability", ex
of which are Spyry'h ptyrnyy "opposed to writing"; zßnd pc'w'kryy "causin
comrades to quarrel"; prôyztjmncyq w'r "watering at the time suitable for orchar
mrc ßnd'm yw'nkryh "a death-sentence sinner, a sinner punishable with deat
These Sogdian expressions recall the Mongolian phrases, cited above, which r
original Sanksrit compounds.
With the advance of Islam, Persian, in addition to Arabic, greatly influ
Turkic generally. In Modern Uighur more than half the vocabulary is compo
foreign words, and compound words are formed mainly from Arabic elemen
Gabain1 1 1 gives the example of the abstract compound burxan quťí "Buddhascha
which she compares to Tokharian A puttisparäm "dignitas Buddhae" < Skt Bu
and is plus Tokharian paräm "dignitas"112 and Sogdian pwt'n'k.
Another wave of influence on Turkic came from Russian. Not unlike Altaic
itself, Russian, and Slavic generally, was at its earliest stages poor in compounds,
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
122 R.L.FISHER
In the east, Chinese played an analogous role with Korean and Japanese, bot
of which langauges absorbed, in addition to the writing system, vast quantities
vocabulary (in the case of Korean, about 50% of its total vocabulary). Gabain118
wonders if an tony mie compounds forming abstracts are built on Chinese models (-R
chángduãn "length" < "long" + "short"). Modern Uighur, being split between
two empires, has developed a dual modern vocabularly, one Russian and one Ch
nese.119 Despite its late attestation and long intimacy with Chinese, Manchu stil
presents a picture not much different than Old Turkic and Classical Mongolian: some
nominal phrases, but mostly relying on derivational affixation, which has remain
extensive.
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 1 23
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
124 R.L.FISHER
tabyač atlïy süsi bir tümän artuql yeti big siig ilki kün ölürtüm (BK SI )
Chinese horseman army-its one myriad excess-its seven thousand troops-ACC
first day I-killed
"On the first day I killed 17,000 Chinese cavalrymen."
Two -i/-ï third person singular possessive suffixes are used to group nouns and a
final accusative suffix (-g) is added to the end of the string, with no inflection on the
adverbial phrase.
In fact, one is struck by the fact that in early texts such as the Orkhon inscrip-
tions and the Secret History of the Mongols , nouns have few modifiers and, further,
subjects are only explicit expressions for the subjective elements in verb forms, that
is, subjects are only included when a specific meaning was to be expressed, thus
leading to the conclusion that apposition was a basic charactetistic of Proto-Altaic
syntax.131 This appositional characteristic is supported by the looseness of Altaic
noun + noun combinations: in the majority of cases the two nouns do not form an
accentual unit and the constituent members retain their individual meaning, with the
first noun modifying in some general way the second noun. Furthermore, the
dvandva compounds in Altaic point to poorly developed coordination constructions
in the proto-language. In this regard, it is interesting that even in compounds denot-
ing the teens, there is no reconstructible system. Even for two such closely related
dialects as Žiirčen and Manchu the order is opposed: Žiirčen unit + ten (*omso
"eleven", *žir-xuan "twelve", *yuor-xuan "thirteen", etc.)132 vs. Manchu ten + unit
(juwan emu "eleven", juwan juwe "twelve", juwan ilan "thirteen", etc.),133 but
they agree for the number fifteen: *to-bu-xuan and tofohon, respectively. Old
Turkic and Modern Turkish contrast, too: Yeniseian (Old Kirghiz) yeti: yegirmi
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 1 25
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
126 R. L. FISHER
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 127
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
128 R.L.FISHER
genitive/possessive case).
search, but in the meant
thoughts dally with false s
References
Baek, E. J. (1993), A Historical Study of Korean Noun Compounds. Korean Studies 17, pp. 105-
116.
Bazin, L. (1968), Structure et tendences communes des langues turques (Sprachbau). Philologiae
Turcicae Fundamenta , Vol. I. Ed. by J. Deny, K. Gr0nbech, H. Scheel, Z. V. Togan.
Wiesbaden.
Benzing, J. (1955), Lamutische Grammatik. (Franz Steiner Verlag.) Wiesbaden.
Brockelmann, C. (1954), Osttürkische Grammatik der Islamischen Litteratursprachen Mittel-
asiens. Leiden.
Chen, Ching-lung (1989), Concepts regarding numbers, colors, and the cardinal points among the
Turkic peoples. Proceedings of the XXVIII Permanent International Altaistic Conference,
Venice 8-14 July 1985. Giovanni Stary (ed.). Wiesbaden.
Clauson, G. (1972), An Etymological Dictionary of P re-thirteenth-century Turkish. Oxford.
Cleaves, F. W. (1982), The Secret History of the Mongols: for the first time done into English ou
of the original tongue and provided with an exegetical commentary. Volume I (Transla-
tion). Cambridge, Massachusetts.
Dobrovolsky, M. (1992), Toward a Lexical Phonology of Chuvash Inflection as Derivation. The
Non-Slavic Languages of the USSR: Linguistic Studies. Second Series. Chicago, pp. 99-
112.
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
COMPOUND NOUNS AS EVIDENCE FOR EARLIER STAGES OF ALTAIC 1 29
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms
130 R.L.FISHER
247-259.
Wackernagel, J. (1908), Genetiv und Adjektiv. In: Mélanges de linguistique offerts à M. Ferd
de Saussure. (Librairie Ancienne Honoré Champion.) Paris.
Watkins, C. (1985), The American Heritage Dictionary of Indo-European Roots. Boston.
Whitney, W. D. (1964 [1889]), Sanskrit Grammar. Tenth issue of the second edition. Camb
Massachussetts.
This content downloaded from 130.194.20.173 on Sun, 09 Oct 2016 14:32:13 UTC
All use subject to http://about.jstor.org/terms