Osawa

SUSA/JSFOu93,2011
TakasblQSAWA(Qsaka)
PevlsltlngtbeOngllnscrlptlono|Mongolla|romtbeSecono
TurklcQaanateontbebaslso|rubblngsbyG.[.Pamsteot
Since its discovery in the late 19th century, the Ongi inscription is well known as one oI
the Old Turkic inscriptions in Mongolia. In the early 20th century, however, the inscribed
stele was badly destroyed and only partly conserved, which makes it diIIicult to research
the original text. It is thereIore noteworthy that the rubbings and photographs taken by
G. J. Ramstedt and S. Plsi in the excavation oI the site in 1909 are still preserved in the
collectionsoItheFinno-UgrianSocietyinHelsinki.Thispaperaimstoreconstructthe
original orientation oI the inscription and to analyse philologically the new texts, and to
resolve historically the purpose and dating oI the stele and by whom it was established,
with reIerence to new materials and the original data oI the author`s Iieldwork.
1. Pre|ace
1.1. Tbeolscoveryo|tbeOnglslteanostele
As is known, the history oI discovery and research oI the Ongi
1
siteandinscription
can be dated back to the period oI the Orkhon Expedition, which was organized and
executed under the supervision oI V. V. RadloII in the last quarter oI the 19th century.
N. M. Jadrincev, a Russian archaeologist and ethnologist oI Irkutsk, also participated
in this expedition along with other members. RadloII himselI stayed in the Khosho
Tsaidam Basin between 28 July and 7 August 1891. Jadrincev, however, continued
to research old kurgans, stone statues and so-called deer stones Irom Erdeni-zuu to
Oyanga som oI Ubur Khangai along the Kh|kshin Orkhon River. On the way Irom
Erdeni-zuu to the Sayin Noyan Buddhist Temple, he was inIormed by Chinese trad-
ers that there was an unknown site and epigraphic stone near the Maanit Steppe. With
help Irom local nomad Mongolians, he then looked Ior the unknown site and inscrip-
tion.Atlaston20August1891,JadrincevdiscoveredthissiteintheMaanitsteppe
along the Tarimal River (Jadrincev 1901: 4244; cI. Kljatornyj 1964: 63; Kljatornyj
2003:7578)
1 ThisnameisbasedonnativeMongolianusage.InthemodernMongolianroadmap,theriverisgiven
as Ongi gol (Mongolia, Road ATLAS gazryn =urag, 2005, p. 42).
Map 2.OloTurklcsltesneartbeeasternKbangalMountalnslnMongolla.
Khr-Asgatsite
Ongisite
KhangaiMountains
(gtkn y)
O
n
g
i

R
i
v
e
r
Tamir River
T
u
u
l

R
i
v
e
r
O
rk
h
o
n
R
iv
er
Erdene-
Mandal
site
Bugutsite
Karabalgasun
site
Khoshoo-Tsaidam
site(KlTiginsite
andBilgQaghan
site)
Ikh-Khanui-
nuursite
Ikh-
Khoshoot
site
Russia
China
Ulaanbaatar
K
h
e
n
t
e
i
M
o
u
n
t
a
i
n
s
K
h
a
n
g
a
i
M
o
u
n
t
a
i
n
s
A
l
t
a
i
M
o
u
n
t
a
i
n
s
Gobi desert
B
a
i
k
a
l
L
a
k
e
S
e
l
e
n
g
e

R
i
v
e
r
O
r
k
h
o
n

R
i
v
e
r
T
u
u
l

R
i
v
e
r
K
h
e
r
l
e
n

R
i
v
e
r
Khoshoo-
Tsaidamsite
Ungetsite
Tunyukuksite
Map 1.
OloTurklc
sltesano
steles
lntbe
Orkbon
valley.
Revisiting the Ongi inscription oI Mongolia Irom the Second Turkic Qaanate 149
1.2. Tbelocatlono|tbeslte
WithregardtotheinvestigationoItheOngisiteandstele,theJapanese-Mongolian
International Joint Expedition oI 221 August oI 2007, deIined the location oI the
centre oI the sarcophagus at the site at latitude 46 20 20.3 North, longitude 102 11
02.5 East and at an elevation oI 2110 metres (see Map 1 & 2).
Generally speaking, this site is positioned at the point where the Tarimal River
pours into the Ongi River on the right, and it is located at a distance oI approxi-
mately 17 kilometres Irom Oyana som in Ubur-Hangai Aymak, southern Mongolia.
This site is 300 m Irom the right bank oI the Maanit River and in the centre oI the
basin oI the Ongi Steppe, which is not large. The site is close to the low Maanit-Ula
MountaininthenorthandXosh-UlaMountainintheeast.Fromtheentrancetothis
site,alineoIthebalbal stones extends eastwards to a small stone circle at an eleva-
tion oI 20302040 metres (sawa & Katayama 1999: 126127, Pl. 2a, 2b, 2d; sawa
1999b: 281282; 1999c: 281281; 2000: 197198; 2008: 164166; MNTSD: 112).
Figure 1. Sketcb o| tbe Ongl slte by N. M. [aorlncev 1891 (Atlas: Pl. X|v).
150 Takashi sawa
Figure 2. Four stone statues o| tbe Ongl slte |rom tbe |ront (eastern) sloe (MSSP: 131, Pl. 80).
Figure 3. Four stone statues o| tbe Ongl slte |rom tbe rear (western) sloe (MSSP: 131, Pl. 81).
1.3. |nvestlgatlono|tbeslteanotbelnscrlptlon
According to Jadrincev`s diary, he Iirst surveyed this site and inscription, aIter which
hesentsketchesandphotographsoIthesiteandrubbingsoItheinscriptionstoV.V.
RadloII. On the basis oI these materials, RadloII published the sketches oI the general
view and some remains in 1893 (Atlas: XIV; cI. Figure 1(a) oI this paper). RadloII
described each side and part in his mouldings oI Ongi as Iollows: the wider side
has 8 lines (O), the narrow side has 4 lines (Oa), this same side having small letters
oI 7 horizontal lines over the part oI (Oa)(Ob) and a balbal stone with the letter
(Oc) (Atlas: XVI, LXXXIII1). In the same year, the Orkhon Expedition under the
supervision oI D. A. Klemenc also visited his site and took many photographs, which
however have not been published until now (cI. Klemenc 1895; ATIM: 244). On 10
September 1909, Ramstedt and Plsi then visited the Ongi monument, surveyed it,
tookphotographsandmaderubbings.SomeoIthesephotographsandasketchoIthe
site were published by H. Haln oI the University oI Helsinki (Figures 2, 3 and 5). It
is especially noteworthy that we can see the Ongi stele lying behind a stone statue on
the leIt side in Figure 3. This photograph veriIies the existence oI the Ongi stele until
10September1909.
Figure 4.Tortolsestoneo|tbeOnglstele(145150143148cm,tbebolels43cmwloeano
28 cm long, c|. MSSP: 130, Pl. 79) (Pbotograpb No. vKK 5382 |rom tbe Pamsteot collectlon,
Natlonal8oaroo|Antlqultleso|Flnlano(N8A).
152 Takashi sawa
They excavated and Iound a tortoise stone and bricks under the central area, and some
sheep bones in the west part oI the mound (MSSP: 6364, 130, 132, Pl. 8083; Figure
4oIthispaper).
In 1926, Kozlov visited the site and recorded that there were three stone statues
and two sheep Iigures oI granite that had been broken and were placed in the tomb, a
IrameoIslabsoIgraniteintheholethathadbeendugintheground,andstonebalbals
in a line running east Irom the mound etc. (Kozlov 1949: 117; Vojtov 1989: 3638).
His photographs, however, have not been published until now. In 1962, Tryjarski sur-
veyedtheplanoIthismonument,measuringitandtakingphotographsoIthreestone
statues and two stone Iigures oI sheep (Tryjarski 1966: 166168, Fig. 1114, 2325;
1971: Fig. 20; 1972: 3739, Pl. 1a, 1b, 4, 5). Namhaidagwa oI the Arbaikheer National
Museum, however, inIormed Tryjarski that Buddhist monks (Lama) had secretly dug
up and destroyed this monument in the period between 1909 and 1962. Silver plates,
the skull oI a horse, a harness and earthenware vessels etc. had been excavated Irom
there. In early period beIore 1962, Namhaidagwa himselI told Tryjarski that he had
discovered 37 sherds oI broken earthenware vessels carved in relieI oI the Ilower pat-
tern under the sarcophagus, and had added them as new acquisitions to this museum
(cI. Tryjarski 1966: 167). At present, there are about eight holes which can be identi-
Iied as traces oI excavations by the grave-robbers at the time. In 1969, Namhaidagwa
discovered two pieces oI a broken inscription, and Rintchen inIormed Tryjarski oI
Figure 5. Layout o| tbe Ongl slte, 1 = sarcopbagus o| granlte, 2 = tbe tortolse stone ano stele,
3 & 4 = calrn, 5 = oltcb, 6 = bank (MSSP: 64, Pl. 10)
this. Academic research, however, had not been done on this site and stele until the
investigationbyVojtovandBajarin1987(Vojtov&Bajar1989).
1.4. Tberubblngso|tbeOngllnscrlptlon
This site and inscription are recognized as a monument oI the 8th century AD,
although even now it seems that there are many arguments on the points oI the date
oI construction, the party originally responsible Ior the site and the stele, and the
historical background. Some researchers have subsequently surveyed the site and
the stele. Since this discovery, many scholars tried to solve the problematic issues
Irom philological, historical and archaeological perspectives. It is especially note-
worthy that V. V. RadloII published the original rubbings by Jadrincev along with
rubbings retouched by RadloII in 1893 (Atlas: XIV, XVI, LXXXIII1), aIter which
RadloII published his interpretation (ATIM: 243256). From this interpretation we
know that Orkun and Malov basically approved his reading (Orkun 1936: 125132;
Malov 1959: 711). G. Clauson, however, criticized RadloII`s retouched rubbings,
and reconstructed the runic text oI this inscription and tried to present new readings,
revisingtheIirstrubbingsbyJadrincev(Clauson1957:177192).Clausonscopies
made by hand have been regarded as the basic sources as well as Jadrincev`s rub-
bings in the Atlas. A Hungarian colleague recently published a critical work on Old
Turkicrunicinscriptionsindetail(Berta2004:209225),butunIortunatelyhecould
not investigate the original rubbings and materials. Regarding the identiIication oI
Ongi Iragments that we researched in the new rubbings in Mongolia, I have reported
the transliteration, transcription and translation and related notes previously (sawa
1999; 1999b; 1999c; 2000; 2007a; 2007b; 2008; 2010). Visiting Helsinki in 2004 and
2005, I was able to study previously unknown rubbings oI the Ongi stele and other
materials in photographs by Ramstedt and Plsi. I then investigated three rubbings oI
OngibyJadrincevinSaintPetersburgin2006.Accordingly,inthispaperIreconsider
the orientation oI the Ongi stele and the original text on the basis oI new rubbings and
our new Iieldwork with Mongolian and Japanese colleagues in 1996 and 2007.
2. Newvlewsontbeorlentatlono|tbetortolsestoneanotbestele
When Jadrincev discovered the site in 1891, he reported that 'the tortoise stone was
turned toward a southern orientation we can see a stone slab installed on the base oI
the stone stele, the stele itselI being oI granite and erected on the stone slab which
can be regarded as in the style oI a tortoise (Jadrincev 1901: 43; Vojtov 1989: 3435).
From this report, we can conclude that the so-called tortoise stone was turned south-
wards, and thus the wide side oI the stone pillar must also have Iaced southwards. But
18 years later when Ramstedt and Plsi visited the site, the tortoise stone was buried
Iacing westwards under the ground (Figure 3 oI this paper). We thus understand that
154 Takashi sawa
the tortoise stone must have changed orientation Irom south to west. This change oI
positionreIlectstheIactthattheremainshadbeenbrokenintothreeIragmentsaround
1900. What then was the original orientation oI the tortoise stone at Ongi?
The Russian archaeologist Novgorodova analysed this problem on the basis oI
L. Jisl and S. Odzhav`s joint archaeological expedition to the K|l Tigin site in 1958.
SheconcludedthatthetortoisestoneoriginallyIacedeast.Thisisbasedontraditional
sun cult belieIs oI Old Turkic peoples, who revered the east, the direction where the
sunrises.Fromthis,Novgorodovaconcludedthatthetortoisehadbeenturnedtothe
west in later periods, but originally had an eastern orientation (Novgorodova 1980:
240; 1981: 210, Pl. 3). There is no doubt that the Old Turkic words randilgr
means 'in the east, eastwards as well as 'in Iront, Iorwards (ED: 189a, 144a; Tekin
1988:165,138).
As Iar as I know, according to the archaeological evidence oI L. Jisl and
S. Odzhav`s excavation at the K|l Tigin site in 1958 and research by the Turkish and
Mongolian joint archaeological expedition at the Bilg Qaan site, the tortoise stones
at both sites were excavated in a position in which the heads were turned to the west-
ern side (Jisl 1963: 392; Pl. 7; TIKA 2003: 391; TIKA 2005: 205; Bajar 2004: 77, Pl.
7). As I analysed, the Chinese texts oI the western side oI the K|l Tigin and Bilg
Qaan sites had been Iirst carved beIore the other parts except other runic texts oI 2
lines on the same side. It can be recognized that the western side can be considered to
be the most important part Irom the viewpoints oI political and cultural relationships
between the Tang Emperor Xuan:ongand the Turkic Bilg Qaan with reIer-
ence to the expression oI the relationship oI Father and Son .This reIlects
the Sinocentric political relationship oI the Tang Dynasty towards Ioreign countries
which can be called symbolically the regime oI Cefeng. The runic text was then
carved secondly aIter Iinishing the Chinese text oI the western side (sawa 2007a:
2329). This position suggests that the tortoise stone and the inscription had been
originally established so that the western side was turned to the soul oI the deceased,
which can return to the stone statues and the sarcophagus or the shrine that were con-
structed in the western part oI the mound.
IthereIoreconsiderthatinthecaseoItheOngisite,thetortoisestoneandthe
inscription would Iirst have been arranged so that the head and the part with the Iirst
line oI the runic letters Iaced in the same direction. In the Orkhon inscription, we can
see that the Chinese part was the most important side oI all Irom the viewpoint oI the
political relationship between both countries. The western side oI the stele is tradition-
ally regarded as the most important side, and accordingly the Chinese text was carved
on the western side oI the stele. The Iact the Iirst line was carved on the western side
canbeconIirmedbytheIactthatintheTunyukukandIkhKhoshootinscriptionsthe
Iirst line had been carved on the western sides (sawa 2007a). According to the gen-
eral position oI the Old Turkic inscriptions oI the Second Eastern Turkic Qaanate,
in the case oI the tortoise stone and the inscription oI Ongi, we can consider that the
Iirst line was carved on the western side also here, and that the tortoise stone was posi-
tioned with the head Iacing west. ThereIore we can say that the wider side oI the Ongi
inscription Iaced westwards, and the narrower side Iaced southwards.
3. Peconstructlono|tbesteleontbebaslso|tbePamsteotcollectlon
anomy|lelowork
Regarding the names oI the sides in the Ongi stele, in the explanation oI Jadrincev`s
squeeze mouldings, RadloII said that on the Iront side oI this stele there was the sign
(i.e. tama) Ior Qan, and on the right side vertical lines that began Irom a location oI
thesameheightasthelinesoItheIrontside.Moreover,intheupperpartoItheselines,
at the same height as one oI the signs Ior Qan (i.e. tama) there were 7 horizontal
lines.InthepreIacetothedecipherment(ATIM1895:244),andregardingposition
oItheselines,henotedthatthisstelehad8linesontheIrontsideand4linesonthe
right side. RadloII also named the side with 8 lines the wide side (O), the side with
4 lines the narrow side (Oa), and the side with 7 horizontal lines the upper part (Oc).
He did not, however, mention where this can be placed on the Ongi stele. Clauson,
Ior example, also named the Iront side (O) the (right) side (Oa) according to RadloII
(Clauson 1957: 177; GOT: 291292). On the contrary, aIter Onogawa named (O) the
east side, (Oa) the south side without doubt, subsequent researchers continued to sup-
port him (Onogawa 1950: 442444; Sawada 1983a: 54; 1983b: 79; 1984: 94). This
name, however, should be corrected. That the letters had been carved in the wide side
andtherightsideoItheoriginalstelecanbesupportedbythestateoIthelargeIrag-
mentasmentionedabove.Moreover,accordingtothedescriptionoIthestonetortoise
plate (MSSP: 130), we can understand that the navel hole oI the stone tortoise is 43
cm long and 28 cm wide, and accordingly it seems undoubted that the wide side oI
the stele Iaced the head or the tail. In his journal, Jadrincev, who inspected this site in
1891 recorded that the stele Iaced southwards (Jadrincev 1901: 43; cI. Vojtov 1989:
43). ThereIore at that time, the wider side oI this stele Iaced southwards, and the nar-
rower side Iaced eastwards. As a result, the south side can be regarded the wider side
oI the stele, and the east side as it narrower side. It seems doubtIul, however, that the
stele had not been altered since its initial construction, because there was a cut gate
in the east side oI the mound, and a shrine and sarcophagus positioned in its western
part. As analysed above, we can maintain in the case oI the Orkhon, Tunyukuk and
Ikh-Khoshoot (Kli or) inscriptions that the Iirst line oI the inscription originally
Iaced west.
It can thus be considered that the inscription had been made to show respect
tothesarcophagusandshrineorthestonestatueoItheburiedpersoninIrontoIthe
shrine (sawa 2007a: 2230). Accordingly, we cannot deny that the beginning part oI
this stele oI 8 lines is turned to the west as its Iront side. This is the same in the Orkhon
inscriptions and Tunyukuk and Ikh Khoshoot inscriptions oI the Second Eastern
Turkic Qaanate periods. From these viewpoints, I researched the journals oI travel-
lers in Mongolia and the squeeze mouldings made by G. Ramstedt and S. Plsi on 9
September 1909. From these materials, I can conIirm that G. Ramstedt and S. Plsi
considered that the beginning part oI the 8 lines can be the western side, although
this was only based on the Iact that the stele was lying on the ground in three parts
in the eastern part oI this mound, and the beginning part oI this stele was turned to
the western side. Regarding the position oI the Iragments oI this stele, I can say that
156 Takashi sawa
Ramstedt`s view on this orientation oI this stele is completely correct. At present, I
would like to suggest that the initial part oI the 8 lines can be the western side and the
tortoise stone was positioned so that it turned its head towards the west and Iaced the
sarcophagus, shrine and the stone statue oI the person buried at this site. The narrow
sideoIthebottompartoIthe4linesisthesouthernside(Figure6).
4. |oentl|lcatlono|tbe|ragmentso|tbeOnglstelelnmy|lelowork
Atpresent,theOngisteleispreservedasonlyIoursmallIragmentsintheArbaikheer
museumandattheOngisite.In1996and2007,Icouldinvestigate(1)aIragmentoI
theheadstoneoIthestelelyingnearapileoIbricksinthemoundoItheOngisite,(2)
a small Iragment, (3) a medium-sized Iragment and (4) a large Iragment.The three
Iragments oI the stele with runic letters are kept in the local museum oI Arbaikheer
in Ubur-Hangai Aymak, while the other Iragments oI the Ongi stele have not been
discovered until now. It seems obvious that we can no longer restore this stele to its
originalshape.
The Iragments were measured in our survey as Iollows:
(1) TheIragmentoItheheadstone:ThereareruniclettersoIsmallscaleononeside.
The Iront side is 40 cm high while the middle part oI the reverse side is 33 cm high. The
bottom part is 40 cm wide and 1718.5 cm thick. A tamadesignconsistingoIaramand
asnakeiscarvedonthesurIaceoItheIrontside(W).Alongitsoutside,theheadstone
has two simple ornaments oI two Iacing dragons in relieI. On the narrower side (S)
there are small horizontal runic letters oI 7 lines. On the surIace oI the reverse side (E)
acoupleoItama designs consisting oI a ram and a snake were carved (Figure 7).
(2) SmallIragment:ThereareruniclettersoIthreelinesononeside.TheIragment
is 11.5 cm long, 16 cm wide and 8.510 cm thick. The Iront side (W) has three letters
Figure 6. Tbe loentl|lcatlon o| tbe beaostone ano tbe small, meolum-slzeo ano large |ragments
o|tbeOnglstele.PeconstructlonbyTakasblQsawa.
Westernside
Southernside
17cm
17cm
23cm
}
Smallrunicletter
oIthe7lines
8.5cm
10cm
6
7
8
15.5cm
18cm
6
7
5
8 16.5cm
22cm
19.5cm
24cm
20.5cm
19cm
2
3
1
4
2
3
1
4
80cm
77cm
80cm
S
W
19cm
22cm
43cm
25cm
in each line. These letters are indicated by three lines in the 6th8th lines. In 2007, we
took rubbings oI the narrow side (i.e. the northern side) oI this Iragment, discovering
a slight trace oI a ruled line.There is thus no doubt that there are runic letters also
on the northern side oI the Ongi stele, but, unIortunately, we cannot attest the letters
because other parts oI this stele were broken and have disappeared.
(3) Medium-sized Iragment:There are runic letters oI Iour lines on one side.The
Iragment is 2430 cm long, 15.516.5 cm wide and 19.522 cm thick. The Iront
side(W)oIthisIragmenthasaruniclettersonIourlinesIromthe5thtothe8thline
(Figure8).
(4) LargeIragment:ThereareruniclettersoIIourlinesononesideandonesoIIour
lines on the other side. The Iragment is 7780 cm long, 19 cm wide and 20.522 cm
thick.TheIrontside(W)oIthisIragmenthasruniclettersonthe1st4thlinesandthe
narrow side (S) has runic letters on the 1st4th lines.
As I have analysed previously (sawa 1999b: 130131; 2007b: 330332), the
small Iragment can be shown to be the top part oI the western side oI the original stele,
the medium-sized one can be identiIied as part oI the western side, and the large Irag-
ment can be identiIied as the bottom part oI the western side and the top part oI the
southernside(Figure9).
Figure 7.Tbebeaostoneo|tbeOnglstelewltbtamaoeslgnso|aramanoasnake.Fromtbe
le|t, west, east ano soutb sloes (Takasbl Qsawa1996).
Figure 8. Small ano meolum-slzeo |ragments |rom Ongl. Arvalkbeer Museum, Mongolla (Takasbl
Qsawa1996).
158 Takashi sawa
(5) A balbal stone with a runic inscription and a tama design oI a ram and a snake:
Asmentionedabove,abalbalstoneoIBilg sbara Taman balbalw with a tama
design oI a ram and a snake was the eleventh oI the balbal stones extending in a straight
lineIromtheeasternborderoIthissite,andthisstoneisthoughttohavebeeninthis
position since it was erected (as based on our comparative Iieldwork at the site in
1997 and 2007), i.e. the balbal can be identiIied with the shape and the position oI the
surrounding mountains landscape towards the north (this is based on a photograph
taken by S. Plsi in 1909). Accordingly, we cannot regard this balbal oI Bilgbara
Taman balbalw as the Iirst balbal oI this site as considered by RadloII, Plsi, Tryjarski,
Vojtov, Xaraubaj, Bazylxan and others (ATIM: 244; MSSP: 63; Tryjarski & Aalto
1973: 417; 418, Fig. 3; Vojtov 1989: 36; Xaraubaj 2003: 143; Bazylxan 2005: 58,
n. 114; Xaraubaj & Myrzataj 2006: 96). This tama is identical on each side oI the
headstone. As I have analysed beIore, this balbal stone was built by the Bilgbara
Taman Tarqan, who can be identiIied as the younger son oI (letmis yabu inW-4
(Figures10,11a,11b).
5. Pubblngso|tbeOnglsteletakenbyPamsteoton10September1909
Ramstedt and Plsi reported as Iollows on the condition oI the stele. The basic, 56
cm-long Iragment oI this stele stood in the eastern mound. The other part was made
up oI two Iragments, a middle part 154 cm in length and an upper part 104 cm long,
which were lying in the mound. The inscription oI which Kozlov made a rubbing had
been buried under the ground, and when Plsi tried to dig up the Iragment sheep bones
were recovered Irom among the bricks. The main stone had been carved in the shape
oI a tortoise, with its Iace to the west (MSSP: 63). From this, we can assume that
Ramstedt and Plsi had taken three rubbings Irom the original stele.
Figure 9. Tbe large |ragment ln tbe Arvalkbeer Museum, Mongolla (Takasbl Qsawa1996).
Figure 10.Abalbalstonewltbrunlclettersontberlgbtsloeanoatamao|aramanoasnake
on tbe le|t sloe at tbe bottom ln a pbotograpb by S. Palsl (MSSP: 132, Pl. 83).
In March 2004, Iollowing the inIormation presented by Haln (1978: 99) and a
note Irom my report on the Ongi stele (sawa 1999: 130), my Japanese colleague Dr.
KousetsuSuzukivisitedFinlandandtriedtoresearchtherubbingsoItheTunyukuk
inscription in Ramstedt`s materials in the National Archives oI Finland. He was able
160 Takashi sawa
Figure 11a.8albal
stoneslntbeeast
parto|tbeOngl
slte.
Figure11b.Tbe11tbbalbalstonewltbrunlcletters(TakasblQsawa1996).
toconIirmtherubbingsoItheinscriptionsoIOngiandine-Us,butcouldnotconIirm
therubbingsoITunyukuk,althoughtheyareregisteredunderthelistoI(343)SUS
2.20inthecollectionoItheFinno-UgrianSociety(Haln1978:99).Atthattimehe
onlyestimatedthesizesoIrubbingsandinvestigatedsomelettersoIseverallines.
2
ItriedtoconIirmtheconditionoItheirrubbings,andaIterreceivingpermission
toconductresearchIromtheFinno-UgrianSociety,IvisitedtheNationalArchivesoI
FinlandandtheNationalBoardoIAntiquitiesoIFinlandinMarchandApril2005and
FebruaryandMarchoI2006.ThereIconIirmedtherubbingsoItheOngiandine-Us
steles made by G. Ramstedt and S. Plsi during their Mongolian expedition with
IinancialassistanceIromtheFinno-UgrianSocietyunderthedirectorProI.O.Donner
in 1909 (cI. Ramstedt 1978; Aalto 1971; Haln 1998: 143147).
3
InthetoppartoIthe
western side oI the Ongi rubbing, we can conIirm the handwritten note 'TAMIR 10/
IX 1908 in black ink. The rubbings consisted oI 13 thin and light brown paraIIin
papers. The orientations are noted W(est)-1, W-2, W-3, S(outh)-1, S-2, S-3, E(ast)-1,
E-2, E-3. We can see additional epitaphs oI the horizontal 7 lines in the southern side
oItheheadstone(Figure12)
2 IamdeeplygrateIultoDr.KousetsuSuzukiIorhispermissionandhelptousehisimportantdata.
3 With regard to my research in Helsinki, I express my deep gratitude to H. Haln, Doctor oI Honour,
and Dr. V. Rybatzki who helped me obtain permission to investigate the materials oI Ramstedt and Plsi
IromtheFinno-UgrianSocietyandtheNationalBoardoIAntiquitiesoIFinland.
These rubbings were made using the western squeeze moulding method by attaching
the stone with adhesive and tearing oII the paper aIter drying. This is not the same
method as the traditional Chinese rubbing as used by Jadrincev, which makes it too
hard to read and understand the real rune shape than when reading the text Irom the
Jadrincev version. In addition, I would like to emphasize that they took rubbings oI
the three sides (west, south and east). The rubbings consist oI the Iollowing:
Westsloe:TheupperpartoIthestonepillarhasruniclettersoI8linesmeasuring
101 cm x 3839 cm. The middle part oI the stone pillar measures 153 cm x 3537 cm
andbearsruniclettersoI8lines.ThebottompartoIstonepillarhasruniclettersoI8
lines, and measures 62 cm x 3536 cm. Overall dimensions oI the west side: 316 cm
x 3539 cm.
Soutb sloe: The upper part oI the stone pillar has runic letters oI 4 lines and
measures 100 cm x 1821 cm. The middle part oI the stone pillar has runic letters oI
4 lines, with dimensions oI 156 cm x 22 cm. The bottom part oI stone pillar has runic
letters oI 4 lines, measuring 63 cm x 2324.5 cm. Overall dimensions oI the south
side: 319 cm x 1824.5 cm.
Lastsloe:TheupperpartoIthestonepillarhasruniclettersoI8linesandmeas-
ures 111 cm x 3839 cm. The bottom part has runic letters oI 4 lines, with dimen-
sions oI 150 cm x and 38 cm. The bottom part oI stone pillar has runic letters oI 8
lines, dimensions 65 cm x 3942 cm. Overall dimensions oI the south side: 326 cm x
3842c.
Figure 12. Sketcbes o| Pamsteot's rubblngs o| tbe west, soutb ano east sloes o| tbe Ongl stele
(measureobyTakasblQsawa).
Westside
Southside
East side
1839cm
101cm
4.5cm
100cm
1821cm
111cm
40cm
153cm
3639cm
156cm
22cm
150cm
38cm
62cm
63cm
65cm
3536cm
2324cm
4239cm
4.5cm
The data oI our survey shows the Ongi stele to be roughly 320 cm high and 3539
cm wide on the west and east sides, and about 1824 cm wide on the south and north
sides. When compared with the K|l Tigin stele (334 cm in height, 124132 cm in
width on the west and east sides, 42 cm wide on the north and south sides), this stele
is the almost same size in height while the Ongi stele is not as wide. II the text is also
conIirmedonthenorthsideasIsuppose,thestyleisverysimilartotheKlTiginand
Bilg Qaan steles.
Moreover, Irom Ramstedt`s rubbing we can conIirm that a pair oI tama design
consisting oI a ram and a snake was carved on the western and eastern sides oI this
stele. This can be Iitted to our investigation in which rubbings were taken Irom the
both sides oI this headstone in 1997. When Ramstedt and Plsi surveyed this site in
1909, this stele remained, but was broken into three Iragments with the head part and
the second part lying down and the bottom part was lodged in the ground. Ramstedt
noted that the tortoise stone was buried near the eastern location starting Irom the bot-
tom part oI stele. As mentioned above, when Jadrincev discovered it in July 1891, this
stele was built on the tortoise stone with the head oriented south. In 2005, I could con-
Iirm Jadrincev`s statement in his own handwriting noting 'the southern side oI Ongi
monument in Russian in the border oI his original rubbing (Oa) that is conserved
at the Institute Ior Oriental Studies oI Saint Petersburg. From this, we can conclude
Figure 13. Pbotograpbs o| parts o| tbe beaostone ano W-1(bottom), L-1(top) o| tbe Ongl stele
|rom tbe Pamsteot collectlon (vKK 5325, N8A).
164 Takashi sawa
that this stele was moved Irom the place where Jadrincev had discovered it to the east
side oI the mound, and the tortoise stone was also buried in the ground in the place
discoveredbyin1909.AsIstatedabove,theOngistelehadbeentheonlyoneerected
since the period oI construction oI this site (sawa 1999a: 129130; 1999c: 284286;
2007b: 329330) although Vojtov, Xaraubaj and Bold mentioned that this site had
two steles on the basis oI Plsi`s suggestion (Vojtov 1989: 44; Xaraubaj 2003: 132,
146; Bold 2006: 102). We cannot, however, conIirm Plsi`s statement Irom his layout
oI this site (Figure 5 oI this paper) or Irom Xaraubaj (MSSP: 64, Fig. 10; Xaraubaj
2003:358,Pl.).
6. Tbe revlseo runlc tets o| tbe Ongl lnscrlptlon
baseoonPamsteotsrubblngs
I would now like to mention Old Turkic texts according to my investigation oI
Ramstedt`s rubbings that I carried out in 2004 and 2005 at the National Archives oI
FinlandandNationalBoardoIAntiquitiesoIFinland,andtheInstituteIorOriental
Studies oI Saint Petersburg oI the Russian Academy oI Sciences in 2006. I added
somelinesbasedonmyassumptionstothenorthernandsouthernsideoIthisstele.In
transliteration and transcription, I Iollowed the system oI Moriyasu & Ochir (1999:
11910).
Figure 14. Pbotograpbs o| tbe rubblngs o| W-1/2(bottom), L-1/2(top) o| tbe Ongl stele, |rom
tbe Pamsteot collectlon (vKK 5322, N8A).
Figure 15.
Pboto-
grapbso|
tberubblngs
lnW-2/2
(bottom),
L-2/2(top)
o|tbeOngl
stele|rom
tbePamst-
eotcollec-
tlon(vKK
5321, N8A.
Figure 16.
Pbotograpbs
o|tbe
rubblngs
lnW-3
(bottom), L-3
(top)o|tbe
Onglstele
|romtbe
Pamsteot
collectlon
(vKK 5323,
N8A).
166 Takashi sawa
Figure 17.Pbotograpbso|tberubblngsontbesoutbsloe(1)o|tbeOnglstele|romtbePamsteotcollectlon
(vKK 5226, N8A).
Figure 18.Pbotograpbso|tberubblngso|tbesoutbsloe(2)o|tbeOnglstele|romtbePamsteotcollectlon
(vKK 5324, N8A).
6.1. Transllteratlon
TablegivingthetransliterationsystemIromOldTurkicletterstoLatinletters:
Vowels
Front Double Back
a(Iora/)
i (Ior / i/ / e)
(Ior/) W(Ioru/o)
Consonants
Front Double Back
b(Iorb/v) B(Iorb/v)
c
ic
dX D
g G
k Q
k, uQ
Q
lY L
lt
m
n N
nc

nt,

p
r R
s(Iors/)I S(Iors/)
(Iors/)
t T!
y Y
z
g
g
168 Takashi sawa
Generalrules
Transllteratlon
# SeveredpartoItheinscription.
: Punctuation.
|aBc| Letters completely restored.
(aBc) Letters partly damaged but restored with certainty.
/// Damagedandillegibleletters.
. . . Illegible letters with some traces. Number known, indicated by points.
Transcrlptlon
# Severed point oI the inscription (placed aIter a word where it extends over two Irag-
ments).
|aBc| Words completely restored.
(aBc) Letters omitted by mistake, but restored by the editors.
/// Damagedandillegibleparts.
Translatlon
|aBc| Words completely restored.
(aBc) Words not in the text but added to improve the English, or explanatory remarks by the
translator.
6.2. Transllteratlonanotranscrlptlon
W-1 c m z : p a m z : Y m a Q G N : t r t : b W L G : (Q s)m s# ( : ) Y i G m s :
YYms:Bsms:WLQN:YWuQ:BWLTuQDa:ksra:ly(i)tms:Q
c (m s) : Q a : / / / / / / / : Lt m s : t g m s : / / / / / : / / / / / / / /# / / / / / / / / / / / / / /
/ / : l b g l r : T B G c :
cmiz apamz : yama qaan : t|rt : buluu : qsm # : ygm : yaym :
basm : ol qan : yoq : boltuqda : kesr : elyitmi : qac(m) : qa : / / / / / / / :
altm : tgmi : / / / / / : / / / / / / / /# / / / / / / / / / / / / / / / / : elbgler : tabac :
W-2 Q G N L d uQ : Q G N n : i c G N i : i d m s : t r k B W D N : # r a : k n :
T W G s Q a : k s r a : k n : B T S Q a : t g i : b r y a : T B G c Q a : Y i R a
y a : y i S Q a : (t)g(i) : / / / / : / / / / / / / / / / / / / / / / #/ : / / / / / : Q = G nt uQ : c
n : WL : Q z W G L n : k r t i :
qaanladuq qaann cn : dm : trk bodun : | # r : kn : tusuqa : kesr
: kn : batsqa : tgi : beriy : tabacqa : yraya : yqa(t)g(i) : / / / / : / / / / / /
/ / / / / / / / / / #/ : / / / / / : qazantuq : cn : ol qz oln k|rti :
W-3 L p r n : B L B L : Q s D i : t r k B W D N : T i Y uQ : B W L W : # B R m s :
r t i : t r k B W D N : (y) i t m z n : t y n : Y W L i R(r)m z n : t y n : z : t
ri:ti:(ms):rg/#/////////////////://////////://///:nta://///
/////:
alp rin : balbal : qd : trk bodun : at yoq : bolu : barm : rti : trk bodun :
yitmzn : teyin : yulr rmzn : teyin : z : tri : temi : rig / # / / / / / / / / /
////////://////////://///:nta://////////:
W-4 Q p G N : l t r s Q G N : l a : Q i L nt m : l t m s : Y B G W : W G Li # : s B R
a : T m G N c W R : Y W G a : i n s i : b l g a : s B R a : T m G N : T R Q N : Y
W G a T i : b s : y t m s : c m : T i m : T / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / /
/ / / / / / / / /#/ / / / / / / / / / : / / / : / / / / / : s : s : W G L i : T m G N :
qapan : elteri : qaan : eli : qlntm : eletmi yabu : ol # : bara : taman
cor : yoa : inisi : bilg : bara : taman : tarqan : yoa : at : be : yetmi :
cim atm : at/ / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / /#/ / / / / / / / / /
: / / / : / / / / / : s : i : ol : taman
W-5 B W : T B G c D a : Y i R y a : T g W G z : R a : y t i r n : Y G i : # B W L m s :
Q m : B G a : t r k n : y n : nt a : Y W R m s : i s g : k cn : b r(m s) : r t m
: (k l)|m | / / / / / / / / / / / / / / / / / / : / / / / / / / / / / / #/ / / / / / / / / / / / / / / : T R D
Ws:BWLTuQDa:
bu : tabacda : yriy : tg ouz : ara : yeti rin : ya : # bolm : qam baa :
trikn : yannta : yorm : iig : kcin : bermi : rtmi : (kl)|mi) / / / / / / / /
//////////:///////////#///////////////:tardu:boltuqda:
W-6 t r k n : k a : i s g : b r t : t y n : Y R L Q m s : s D T G : # nt a : b r m s : B W
L T uQ D a : T W q z : W G z : T g : Y G i : r m s : B d k : r (m s : )Q G N : t g(m
)//://#////////:n////////////////////#//////////:bzbdkbz
:biz:BTbz:bz:
triknk iig berti : teyin : yarlqam : ad at : ant # a : bermi : boltuqda
: toquz : ouz : tg : ya : rmi : bdk : r(mi :) qaan : tg(mi) / / : / / # / /
//////:n////////////////////#//////////:bizbdkbiz:bizbatbiz
:biz:
W-7 Y B z B T b i z : z G k s g : k r t g : i(r)t i : s l t m # : t r r m s : m T i b g l
r m a : t r r m s : b z : z b z : t y n : L # Q N W R : (r t m z) : / / / / / / / / / / / / / / /
/ / / / / / / / / / / / / / /#/ / / / / / / / / / / / / / / : Y W R i Y i N : s : s l y(i)n :
yavz bat biz : az kg : k|rtg : irti : sltim # : ter rmi : amt bglrim a
: ter rmi : biz : az biz : teyin : al # qnur : (rtimiz) : / / / / / / / / / / / / / / / / / / / /
/ / / / / / / / / /#/ / / / / / / / / / / / / / / : yoryn : s : slyin :
170 Takashi sawa
W-8 Q m : S(D) : nc a : t n m s : t r k n : L m(z n) : t y n # ( : ) / / / / / / / / / / / / /
/ / /(B W)D N : . (nt)/(b r)m z ()a : T s W L(r) / / / / : (T c m a) / / / / /d/ / / / / / /
//////////////////////////////////#//////////////////////////
///////////////
qam : (ad) : anca : |tnmi : trikn : almazn teyin # : / / / / / / / / / / / / / /
bodun : anta bermzi : tusul r / / / / / : (atacm a) / / / / /d/ / / / / / / / / / / / / / /
//////////////////////////#/////////////////////////////////
////////
S-1 Q m uQ : B L Q a : (t g d m) : uQ W N L D m : lt m : s s i : k l t i : R Q s # n :
y G D m : b g i : Q c D i : / / G r t i( : )T B G c B W D N i : / / / / / / / / / / / / : T
uQ i D(T)m : y G(D)m : B s(d m : y y d m : ) / / / / (B W = W Q N c a : )/ / / / / /#
/ / / / / / / /m z : r(nc) : s g(k c g) b(r)t(i)|m: | / / / / / / / / /
qamuq : balqqa : (tgdim) : qonuldm : altm : ssi : klti : arqas#n : yadm
: bgi : qacd : / / rti : tabac bodun : / / / / / / / / / / / : toqdtm : yadm :
bas(dm : yaydm : )/ / / / buzuq anca : / / / / / / / / : / / / mz : rinc : iig kcig
bert(im:)//////////
S-2 k l r : (r)t mi z : k i n R a : T g Y G i : B W L m s : t g m c i # m n t y n : S Q nt m
: t r i b l g a : Q G N Q a : nc Q a : s g k c g : b r s g m : B R r m s : r nc : (t)
g d(k)n ( c n) / / / / / / / / / / / / / / / / / / / # b m a : t g d k m : W R l i Q l i : D
R L m (L)m : t y n : Q G N D a : D r L m z : t y n : t g d m :
klir : rtimiz : ekin ara : tg ya : bolm : tgmci#mn teyin : saqntm :
tri bilg : qaanqa : ancaqa : iig kcg : bersgim : bar rmi : rinc :
tgdk(in) (cn) / / / / / / / / / / / / / / / / / / /# bim tgdkm : ur el qa el :
adrlmalm : tyin : qaanda : adrlmaz : teyin : tgdim :
S-3 t g p : i n m a : W G L m a : nc a : t l d m : Q : Y W R p : l t r s Q G N Q a # :
D R L m D uQ : Y L m D uQ : t r i b i l g a : Q G nt a : D R L m L m : z m L
m : t i y n : (nc)a : t l d m : l g r B R G m a : B R D i : b i l g : Q G N : B W
D (N)i / / / / / / / / / / / / : B R D i# : l g n : T Q a s g k c g : b r t m : D R L m
z:tyn:tgdm
tgip : inima : olma : anca : |tldim : qa : yorp : elteri qaanqa# : adrlma-
duq : yalmaduq : tri bilg : qaanta : adrlmalm : azmalm : teyin : anca :
|tldim : ilgr barma : bard # : bilg qaan : bodun / / / / / / / / / / / / / : bard
# : |lgin : atqa : iig kcig : bertim : adrlmaz : teyin : tgdim
S-4 z a : t r i : Q N : l i : y i L Q a : y t i nc : Y : k c l g : L p # r : Q G N m D a
: D R L W : B R D z : b i l g a : T c m : Y W G : uQ W R G N : Q z G nt m
: (i)l y t i : t r(i) : z a : t r i : k n : / / / / / / / / / / / / / / / / / /# / / / /(T B G)c : k
r r : r t m : d g : t c m : z : tri : qan : li : ylqa : yetinc : ay : kclg : alp
r : qaanmda : adrlu : bardz : bilg : atacm : you : qorn : qazantm :
el yetti : tri : z : trikn : / / / / / / / / / / / / / / / / / /# / / /(taba)c k|rr : rtim
: dg : atacm : / / / / /
L-1 ///////////:////////////////////#////////////////////////////
/////////////////////////////////////////////////////////////
/////////////////////////////////////////////////////#//////:
/zW///n///////ir//////://///////////:////////////////////
#////////////////////////////////////////////////////////////
/////////////////////////////////////////////////////////////
/////////////////////#//////:/zW///n///////ir//////://
L-2 ////////////:////////////////////////#///////////////////////
/////////////////////////////////////////////////////////////
/////////////////////////////////////////////////////#:TWuQ
i D m z : c m B W Y L a : / / / / / / / / / / / / / / / / / / / / / / / / / / / : / / / / / / / / / / / / /
///////////#/////////////////////////////////////////////////
/////////////////////////////////////////////////////////////
/ / / / / / / / / / / / / / / / / / / / / / / / / / /# : toqdmz : cim boyla : / / / / / / / / / / / / /
//
L-3 /////////////////grtm:////////////////#/////////////////////
/////////////////////////////////////////////////////////#//B
s T p r g : W D R W : nc D m : / / / :
///////////////////////girtim:////////////////#///////////////
/////////////////////////////////////////////////////////////
/ /# / / basatp rig : uduuru : sancdm : / / / : / / / / / /
L-4 / / / / / / / / / : s g : k c g : k / / : / / / / / / / / / / / / /# / / / / / / / / / / / / / / / / / / / / /
/////////////////////////////////////////////////////////////
/ / / /# / / : n g / / : / / / / / / / / c n : / / / d m :
/ / / / / / / / / / / / : iig : kcig : k / / : / / / / / / / / / / / / /# / / / / / / / / / / / / / / / / / /
/////////////////////////////////////////////////////////////
/ / / / / / /# / / : n g / / : / / / / / / / icn : / / / dim : / / /
L-5 / / / / / / / / / / / / / /D s / / / / / /T c m : / / / / / /# / / / / / / / / / / / / / / / / / / / / / / / / / /
////////////////////////////////////////////////#/m:B//////
/ n : t g : S B R a : T R Q NG :
/ / / / / / / / / / / / / / / / / /D s / / / / / /atacm : / / / / / /# / / / / / / / / / / / / / / / / / / / / /
////////////////////////////////////////////////////#/m:B//
/ / / / / n : tg : bara : tarqan : / / / /
L-6 / / / / / / / / / / G / / / / / / / / / / / / / / / / / / / / / /# / / / / / / / / / / / / / / / / / / / / / / / / / /
//////////////////////////////////////////////////#//////////
///////////////////////////////////////////////////#//////////
/////////////////////////////////////////////////////////////
/////#/////////////////////////////
172 Takashi sawa
L-7 ///////////////////////////////#g////////////////////////////
/////////////////////////////////////////#///////////////////
/////////////////////////////////////////#g//////////////////
///////////////////////////////////////////////////#/////////
////////////////////
L-8 / / / / / / / / / / / / / / / / / / / / / / / / / / / / / /# (l) t (r) s : Q G N : / / / / / / / / / / / / / / / /
///////////////////////////////////////////////////////////#//
////////////////////////////////////////////////////////#elteri
: qaan : / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / /
/////////////////////#////////////////////////////
(N-1~4) The number oI letters is uncertain, but it is possible that they were carved in
theoriginalstele(cI.Bichees68).
Transllteratlono|tbesoutbsloeo|tbebeaostone
1 T(c)m Q a : b i t g : T G
atacmqa : bitig : ta
2 Q i L D m : b i g
qldm : big
3 Q G N m : T c m
qaanm : atacm
4 b l g a : T c m : l
bilg : atacm : l
5 yiLQa:bilga:
ylqa : bilg:
6 W L G : L p r : d g Q N :
ulu : alp r : dg qan:
7 T c m : l t i
atacm : |lti
6.3. Translatlon
W-1 It was said that my ancestor, my IoreIather Yama Qaan compressed, assembled,
enlarged and made a surprise attack on the Iour corners (oI the world). It was
said that when that qaan died, his peoples lost their way and they were scattered.
Anditissaidthatheorganizedthepeoplesandattachedto//////,//////,///
/////////.Itissaidthatleaders(Begs)oItheTurkiccountry
W-2 completely lost the qaan whom the Tang Dynasty had made their ruler. Up to
the Iront (eastern) side where sun rises, up to the back (western) side where sun
sets, up to the right (southern) side where the Tang Dynasty existed, up to the
leIt (northern) side where the wooden Mountains existed, Turkic peoples were
victorious(againstthe///////,and////////////////).Hisdaughtersandsons
(i.e.Turkicdaughtersandsons)servedhim(i.e.theTangDynasty)(literally,he
saw his daughters and sons).
W-3 They constructed brave warriors as Balbal stones. 'The Iame oI the Turkic peo-
pleshasvanished.You,Turkicpeoples,dontgotoruin!andYou,Turkicpeo-
ples, don`t scatter, the God Tri said above. They / / / / / / warriors. / / / / / / / /
//////////////////////,there,/////////:
W-4 I was born in the country oI Qapan, Elteri Qaan. I was a son oI Eletmi Yabu,
and the younger brother oI bara Taman or Yoa, and I had the name oI Bilge
bara Taman Tarqan Yoa. (I and) my 65 uncles younger than my Iather and
nephews (or grandsons), / / / / / / / /,/ / / / / / /,/ / / / /, 'Make the military go ahead
in a hurry. It was said that his son Taman
W-5 became the enemy with the seven warriors among Ouz tribes up to the northern
regions oI this Tabac (China). It was narrated that my Iather went to march tak-
ing side with Baa Trikn, and he served (Baa Trikn) and he has passed. /
/ / / / / / / / / / / / / / / / /,/ / / / / / / / / / / / / / / / /, when he became the Tardu,
W-6 'You served the Trikn it was said that (Trikn) told and appointed him a
title oI ad at that time. It was said that when he became ad, he opposed up to
the Toquz Ouz. It was said (the enemy) was strong. It was said that Qaan went
ahead.//////://///////://///////////////:Wearestrong.Wedonthave
advantage.We
W-7 are in a bad way and worth very little. You have seen how Iew we are and how
many they are. They were annoyed. We marched, he had said. 'Now, my lord
(beg), ah he had said. 'We are a small Iorth, aren`t we he said and 'We were
exhausted: / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / /, I want to march,
I want our army to march
174 Takashi sawa
W-8 said he. And my Iather ad addressed as Iollows, 'Don`t let Tngrikn plunder
saidhe.///////////////:AndMypeoples,you,dontgivehimsomething
Iromhere,ah.PleasebeuseIul!/////////////////////////////////////
/////////:MydearIather,you,ah////////////////////////////////
//////////////////////////////////////////
S-1 I attacked all cities, and I settled down. His army came (against me). Then I at-
tacked Irom behind, the lord (beg) escaped, and they were / / / / / / / / / / / / The
Tabac (Chinese) peoples did / / / / / / / /, I made (my army) attack and I attacked
(them). I overwhelmed and broke into pieces. (They) Iell into / / / / / / / /, then it
iscertainthattheydidnot/////////////Iserved(him)/////,
S-2 We came and we became enemies against them as Iar as to the location between
the two groups. 'I am not willing to attack (them) I thought. I have a desire to
serve Tri Bilg Qaan at least, certainly. They attacked me / / / / / / / / / / / / / / /
/ / / / / / / / I went to my house. 'The people oI sons` generation and Iathers` gen-
eration(oImyIamily),you,dontbeapartsaidandLetusnotbeapartIrom
the Qaan said and I marched.
S-3 I attacked, and I advised my little brothers and my sons as Iollows: 'My Iather
marched ahead and he did not make a mistake against Elteri Qaan. 'Let us
not be apart Irom Tri Bilg Qaan and not leave him said I and also said as
Iollows: 'Ones who went ahead went. The peoples oI Bilg Qaan went away, / /
/ / / / / / / Regarding death in battle as an honour, I served him. Let us not be apart
(Iromhim),Imarchedahead.
S-4 Over, Tri Qan, you went away separating Irom my Qaan that was a strong and
brave warrior in June oI the Li (Dragon) year, my dear and wise Iather I could
holdIuneralservicesIoryouandIgotyourpasture.Heorganizedhiscountry.
Over the sky Trikn and the China were subjugated. / / / / / / / / / / / / I subju-
gatedtheChina.MydearandgoodIather!
L-1 ///////////:////////////////////#////////////////////////////
/////////////////////////////////////////////////////////////
/////////////////////////////////////////////////////#//////:/
zW///n///////ir//////://
L-2 ////////////:////////////////////////#/////////////////////////
//////////////////////////////////////////////////////////////
//////////////////////////////////////////////////#:Weattacked.
Myuncle,Boyla:///////////////
L-3 / / / / / / / / / / / / / / / / /He entered (or He was subjugated): / / / / / / / / / / / / / / / /# /
/////////////////////////////////////////////////////////////
/ / / / / / / / / / / / / / / /# / / I made attack (the enemy ?), and then ran aIter him and
spearedhim:///://////
L-4 /////////(I)served:k//://///////////#////////////////////////
/////////////////////////////////////////////////////////////
/#//:ng//:///////Ior...:///Idid...:///
L-5 //////////////Ds//////mydearIather://////#///////////////////
//////////////////////////////////////////////////////#/m:B
/ / / / / / / n : : bara : tarqan such as . . . : / / / /
L-6 ///////////////////////////////#g////////////////////////////
/////////////////////////////////////////#///////////////////
//////////
L-7 ///////////////////////////////#g////////////////////////////
/////////////////////////////////////////#///////////////////
//////////
L-8 / / / / / / / / / / / / / / / / / / / / / / / / / / / / / /# elteri : qaan : / / / / / / / / / / / / / / / / / /
/////////////////////////////////////////////////////////#///
////////////////////////
(N-1~4) The number oI letters is uncertain, but it is possible they were carved in the
originalstele.
Aooltlonallnscrlptlon
1 FormydearIather,stoneepitaph
2 I constructed. Eternal,
3 my Qaan, my dear Iather
4 wise and my dear Iather, in the
5 L (Dragon) year, wise,
6 great and brave warrior, good Qan,
7 mydearIatherdied.
176 Takashi sawa
6.4. Pbllologlcalanoblstorlcalconsloeratlons
W-1:yama qaan:
When RadloII Iirst read this word, he transliterated it as Y m i (Yami), not Y m a
(Yama).
4
However, as Iar as I investigated in Ramstedt`s rubbing in the National
Archives oI Finland in Helsinki in the winter season oI 20042005, this word should
bereadYma, and this could also be conIirmed by one oI Jadrincev`s rubbings when
IvisitedtheInstituteIorOrientalStudiesoISaintPetersburginthesummeroI2006.
As to the identiIication oI Yama 4aan, I would like to identiIy him with Shemo
hehanIromanOldChinesesourceandZhamaoItheOldTibetandocument
(Pelliot Tibetan No. 1283 oI Tunhuang Documents collection; cI. Moriyasu 1977: 3, n.
5,6).
5
TheChineselettersShemocanbepronouncedas*D:ia,*MuainMiddle
Era Chinese (MEC) (GSR: No. 807a, 17e). Furthermore, according to Pulleyblank,
the Iirst character can be also reconstructed as *ia or *fia and the second one is
reconstructedas*muainLMC(Pulleyblank1991:279,364,217).Ataglance,there
is a diIIerence between the Yama oI Ongi and the *D:ia-Nua oI the reconstructed
phonology oI EMC. How can we explain this diIIerence? I would bring your attention
toavariantoIthetitleoIYehu,*ip,*uo (GSR: No. 633d, 783k). This suggests
two kinds oI sounds such as yabuandfabu in the Turkic Qaanate period. Especially
regardingthesecondsound,fabucanbeattestedinavariantoIyabusuchas
She/yi, po, hu, *aie/tsie, *bua, *uo (GSR: No. 3l, No. 25q, No. 41i) Irom Old
ChinesedocumentsunearthedIromtheAstanatombs(No.329)oIthe6th7thcen-
turiesAD oI the TurIan Basin oI Xinjiang Uigur regions oI China (TAM60: 23/1,
23/2; TFCW III: 342343; TFCWF I: 461; cI. sawa 2002: 84, 87, n. 41). As another
example, we can point out that the same phonology is Iound on Old Turkish coins in
the 67th centuriesAD unearthed Irom Tashkent, Samarkand and other regions oI
Sogdiana in modern Uzbekistan, in which the title oI yabu was carved in the Iorm
oIcapuinOldSogdiancursiveletters.Thisc was used in the place oI f,becausein
Sogdian letters there were no letters reIlecting f, and p was used in the place oI b.
ThereIore, we can reconstruct fabuasthecorrectphonologicalIorm(Babayar2007:
1922). From this, we can say oI the title yabu that there were two phonetic vari-
ants such as yabu and fabu in the Turkic Qaanate period.
6
Thus we can assume
that there was another variant that had a ysoundintheheadpartintheplaceoIthe
pronunciationoI*D:ia-muainOldTurkicperiod.ThisIormcanbeattestedasYama.
4 Boldreadsthisasyumi qaan(Bold2006:111,116).
5 On the basis oI RadloII and Clauson`s reading oI Yami 4aan on the Ongi stele, V. Rybatzki
consideredthatthisnamecanbeattestedas:a-ma-moanintheTibetandocumentsoINo.1283oIthe
Pelliot Collection (Rybatzki 2000: 209). The Chinese researcher RuipresumesthatYami qaanonthe
OngisteleisShemo qaan oI a Chinese romance (YY==-4) oI the Tang period (Rui 1998: 297). But both
researchers` identiIications are hard to accept as there is a diIIerence between i and a oI the last vowel oI
this name. Regarding this proper name, I have proposed a new view (sawa 2009: 401, n. 2).
6 This phenomenon can be conIirmed by the Iact that the title oI ad~ has other phonological Iorm
such as cad~ in the runic letters oI the Terkh (Tariat) inscription and ine-us inscriptions oI the early
Uigur nomad Qaanate in Mongolia.
Fromthis,IcansuggestthatYama 4aan can be identiIied with the *D:ia-mua Qaan
oI Old Chinese sources and the Zhama oI Old Tibetan documents. We cannot thus
support the view that the YamioIOngicouldbeistemiasClausonhassuggested.Who
was he then? As to this problem, I can say only that he can be identiIied with Bumn
4aan oI the Orkhon inscription, as almost all researchers such as RadloII, Marquart,
Onogawa and Sawada have assumed (ATIM 1895: 253; Marquart 1898: 39; Onogawa
1950: 444, n. 1; Sawada 1984: 102103, n. 10). Regarding this problem, I have pre-
sented my view oI the cultic-cultural relation between the Sogdians and Turks Irom
philological and historical points oI views (sawa 2007b).
W-1 qsm.
Theverbqs-isavariantoIq- to compress, to squeeze` (ED: 665b-666a).
W-1 : il bgler : tabac : W-2: qaanladuq qaann cn : dm :
This last part oI W-1 can be read in Ramstedt`s rubbing. This can be directly con-
nectedtotheIirstlineoIW-2.ThephraseIlbglrmeanshigh-rankingaristocratic
peopleoItheOldTurkictribes.InOrkhonTurkic,a-hasaIunctionoIanaspectual
verb and means to do something completely` (Erdal 2004: 247), so wcwnw wa-means
lost (trans.) completely` (Erdal 1991: 257). As to the sentence involving qaanlaauq,
we can reIer to the same expression Irom the K|l Tigin inscription reading as Iollows:
(D6)begleri . boaunw . t:s: cn . tabac . tebligin . krlgin! cn . armaqcwsin
. cn . inili . ecili . kiksrtkin . cn . begli . boaunlw . yoasurtuqwn . on . trk
. boaun . illeak . ilin . wcwnu . wamws(D7)qaanlaauq . qaann . yitr . wamws .
tabac . boaunqa . beglik . urw owlin . qul bolt . esilik . qw: owlin . kng boltw . trk)
begler . trk . atin . wttw . tabacg . begler . tabac . atin . tutpan . tabac . qaanqa
(D8)krmus, elig ywl . isig kcg birmis (KT-E: 68; Tekin 1988: 1011).
I can also mention the same contents in the Tunyukuk inscription as Iollows:
(2) Tr()k boaun . qanin bulmayin . tabacaa . aarwltw . qanlantw . qanin qoaup .
tabacqa . yana icikai . (T: 2; Tekin 1994: 23).
In my view, this sentence indicates the historical situation in which their
Qaanate was subjugated by the Tang Dynasty in 630. As a result, the Turkic peoples
served theTang emperor as his subjects. In 679 the Ashite tribal leaders Wenchuan
andFeng:hiandtheAshinastriballeader*Ni:hukbgmountedarebellionagainst
the Tang Dynasty, which, however, was suppressed by the Tang army. Accordingly,
this part oI this stele can be dated to AD 630679. As is well known, the First Turkic
Qaanate was ruined in 630. They then moved into the regions oI the Ordus and the
Hunag-hua River oI northern China and served the Tang emperor in a traditional
tribal system under the Tang Dynasty. This sentence is written with the evidential
marker-mi, meaning that this is hearsay knowledge reIlecting an independent war
against the Tang Dynasty. Who was 'the qaan that the Tang Dynasty made ascend
to the throne? From the historical point oI view, this can remind us oI AshinaSima
that was called Lisima as its Chinese surname. According to
178 Takashi sawa
sources (TD: 197: 5415, cI. XTS 215: 6039), at Iirst when Qimin kehan (qaan)
was subjugated to the Sui Dynasty, the Turkic tribes oI the northern steppe
(i.e.Mongolia)supportedandappointedAshinaSima to be their qaan. Qimin qaan,
however, returned to his country and Simaabandonedhistitle.
7
ThenaIterthedestruc-
tion oI the First Turkic Qaanate, Sima surrendered to the Tang Dynasty with Illig
qaan. But the Tang emperor praised his loyalty to his lord. Thus, in 639 when the
TurkictribesroseinrebellionintheOrdusregion,TaizonggaveSimathetitleoIYimi
Nishu, silibi, hehan*iet,*mfie,*niei,*:iuk,*a:i,*lfi,*biet
(Yama? Ni:huk, Iltbr) qaan (GSR: No. 505a, 359m, 563d, 1026a, 976m, 519a,
405g)inthegovernment-generaloIDouaufuoIDingxiangthathadbeen
organizedearlierinthenorthernregionsoItheHunag-hua River. Sima was expected
tocontroltheTurkictribes,buthedidnothavetheabilitytoorganizetheTurkicpeo-
ples that wanted him to be a qaan independent oI the Tang Dynasty. He Ieared the
attackoItheSiyantou tribes oI Tokuz Ouz in Mongolia. He escaped to the
capitaloItheTangDynastyandservedTai:ong. However, he died Irom a battle wound
intheLiaotongregion(cI.Suzuki2005).TheTangDynastythengavetheSiyantous
triballeaderthetitleoIYinc Bilg qaan and awarded him control over the Tokuz
Ouz peoples oI Mongolia. However, they also Iailed to organize, and many Tokuz
Ouz escaped to the Tang Dynasty. These historical accidents oI Old Chinese chroni-
cles must be reIlected in this inscription (JTS-194: 51635166; XTS-215: 60396043;
KT-E: 510; BK-E: 610; T: 14; Tekin 1988: 25, 3841; 1994: 23).
W-2: trk bodun : |r : kn : tusuqa : kisr : kn : batsqa : tgi : beriy : tabacqa
: yraya : y(qa) : (tgi : ) / / / / / /
From the context, as I mentioned above, this sentence can be identiIied with the his-
toricalIactthatTurkicpeoplescouldnotassistincampaignsasapartoItheChinese
army under the control oI the Tang Empire aIter the destruction oI the First Turkic
Qaanate. This phrase beriy tabacqa . ywraya yisqa tgi in the south as Iar as the
7 AccordingtotheChineseepitaphoIAshinaSimastombnamedaatangquyouwuweiaafiangfun:eng
bingbushanshuyiyueshunLifunmu:himingbingxu
(that is, Epigraphic text and introduction on the Sir Li that was appointed Great Commander oI
the right wing oI the Tang army and secretary oI the Tang court), the Iather oI the grandIather oI Ashina
Sima was yili kehan (Illig qaan) and Sima`s grandIather was aaba *at bwt
(GSR: No. 271, 276h) kehan ( Qaan), that can be identiIied with Tatpar Qaan that was the
Iourth qaan oI the First Old Turkic Qaanate. Sima was appointed as bositeqin(ParsTigin).
Thenhesuddenlysatonthethroneasfulukehan (Kllg Qaan) and he controlled the tribes
oIXueyantuo,Huihe(Uigur),Baogu(Bogug),Tongluo(Tongra)andsimilar
ones. But then he was deIeated by Jielikehan*Illig qaan, and then he was captured by the
Sui Dynasty. The Emperor YangaigrantedIreedomIromSuicontrolandgavemanygiItssuchas
500tanoIsilkcloths,andmadehimtocomebacktohiscountry.Thentheshibi*Shipi Qaan
gavehimthetitleoIfiabiteqin*kia bt (Pulleyblank 1991: 253; GSR: No. 405g) Tigin that
canbesameasgabi*kp pt (GSR: No. 630a, 407b) Tigin might be reconstructed as *Qapr
Tigin in my view, and aIter this qaan`s death, when Jieli Qaan sat on the throne, he (Sima)
was appointed Luoshi teqin, *las t (GSR: No. 6a, 402a) *Yari (?) Tigin. (The remaining
sentence is omitted.) (=LBS 1993: 112; Suzuki 2005: 4550.)
Tang Dynasty, in the north as Iar as the wooded mountains` can remind us that Turkic
tribes were settled in northern China and engaged in campaigns oI war in surround-
ing areas. In Iact, as mentioned above, since 644, theTang Dynasty conquered the
Tokuz Ouz peoples as Iar as the Mongolian Steppe near the Hangai Mountains, and
reorganizedthenomadpeoples,givingthetriballeadersChineseadministrativetitles
suchastotoq,sici,casw underthecontroloIthegovernment-generaloIDouaufu
oIDingxiangandthegovernment-generaloIDouaufuoIYun:hong.
Moreover, when the Tokuz Ouz peoples rose in rebellion in 660, the Tang Dynasty
sentamilitaryIorcetotheSelengeSteppeandsuppressedit,andestablishedanadmin-
istrative centre named the protectorate-general Douhufu oI Hanhai in
theOrkhonSteppeandtheprotectorate-generaloIYun:honginInnerMongolia
in 663. The Iirst one can be identiIied with Tuo balwq (i.e.Douhuchengmeans
Protecting General administrative oI the City) (KT-N: 4; BK-E: 30). Under the Tang
policy that controlled Ioreign peoples with Ioreign soldiers, the Turkic peoples were
appointed to a high-ranking oIIicial with military titles serving the Tang emperor. At
that time under the Tang commanders, Turkic peoples engaged in campaigns oI war
to the west as Iar as the northern steppe oI the TianshanMountainsintherebellion
oItheAshinaHelo oI the Western Turkic Qaanate in 640 years and eastwards as Iar
asthenorthernKoreanareasintherebellionoItheGaogoulipeoplesin650,
northwards as Iar as the Mongolian Steppe in the rebellion oI Tokuz Ouz. So this y
istobereadastknYi(TheHangaiMountains).Asimilarsituationisnarratedin
the Orkhon Turkic text, as noted in W-1 (KT-E: 8; Tekin 1988: 1011).
W-2: : qazantuq : cn : ol qz oln k|rti : W-3: alp rin balbal : qd:
This part oI W-2 can be attested to only in Ramstedt`s rubbing, being connected to
the beginning part oI the next line. BeIore the word qa:antuk. cn, we cannot read
the letters Irom Ramstedt`s rubbing, but Irom the context it is possible that they stand
Iorilig, which is comparable with the expression with same details in KT-E (910).
As I mentioned above, these incidents are closely related to the campaigns against
surrounding enemies. But it is remarkable that these two sentences are narrated in the
style oI the direct past suIIix -ti/-a. In my view, this verb Iorm also has the Iunction
oInarratingahistoricalIact.ThisIactmusthavebeenashameIulandunIorgettable
memoryamongtheTurkicpeople.
Secondly, regarding the word rin, RadloII read it as rin as derived Irom r
(man, warrior) an accusative suIIix -in (ATIM 1895: 247), but Clauson read rn
as consisting oI r (man, warrior) a plural suIIix -n (Clauson 1957: Pl. V, 182;
Tryjarski&Aalto1973:419).ButinOldTurkic,thetermbalbalisusedasasingular
Iorm with an accusative suIIix -(i)n, -(i)g or without a suIIix: alp rin |lrp balbal
ql bertim (BK-S: 7), qu sng balbal bertim (BK-S: 9), qrz qaan balbal tikdim
(KT: E-25). Thus, in this case, the Iorm r can be also a singular Iorm. Here, I would
like to analyse -in as an accusative suIIix as done by RadloII.
8
8 Regarding this problem, Bold also analysed this as r+-in(Bold2006:112,117).
180 Takashi sawa
Regarding q-, Clauson considered it to mean to thrust` (Clauson 1957: 182).
This verb is used in the Tunyukuk inscription, meaning birlikte yapmak, etmek,
klmak` (T: 5, 6, 11, 21; Tekin 1994: 61). It is, however, hard to say why the 'balbal
stone is mentioned. This Iact can be dated to the periods under the control oI the
Tang Dynasty. From the context, alp r (brave warrior) was a tribal leader oI their
enemykilledbyTurkicpeoplesincampaignsaroundthemintheperiodAD630679.
AndthisdescriptionindicatesthatevenunderthecontroloItheTangDynasty,Turkic
peoples maintained their traditional burial customs since the First Turkic Qaanate.
This has been conIirmed in Old Chinese chronicles (=S-50: 910; SS-84: 1864), and
archaeologically conIirmed in the Bugut site, the Tsetsuuv site (Idel site), and the
Bayin-Tsagaan Kh|ndii site (Gindin-Bulak I site) oI the First Turkic Qaanate period
(Vojtov 1996: 2730, 3334; Moriyasu & Ochir 1999: 121; sawa & Suzuki 2008:
8485,8889,129131).
W-3: trk bodun : at yoq : bolu : barm : rti : trk bodun yitmzn : teyin : yulup
rmzn : teyin : z : tri : temi : rig / / / /
This part can be considered to be the words oI Tri (a god), and the Iirst sentence can
be compared with the Iollowing expression: trk.begler.trk.atin.tt (KT-E:
7; Tekin 1988: 1011). Here the God Tri ordered the Turks to take a new ruler to be
independent Irom Tang rule. A similar expression can be conIirmed as Iollows: :.
trk. trisi. trk wauk yiri. subw. anca tmis. trk. boaun. yok. bolma:un. tiyin.
boaun. bolcun tiyin. (KT-E: 11; Tekin 1988: 1011). This is Iollowed by the enthrone-
mentoI(lteris qaanandElbilgqatun (KT-E: 11; Tekin 1988: 1011). Thus in this
context, it is expected that a Turkic qaan`s name (possibly Elteri qaan) that was
admitted by the God Tri was carved in the space that Iollowed, which is hard to
read now.
W-4: qapan : elteri : qaan : eli qlntm :
This part has keywords to resolve the date oI this stele. First, elia hasthebasicmean-
ing in the country` (cI. GOT: 132). As Iar as we know, the usage oI /// eli canbe
also conIirmed in T: 1, 58; SU-N: 4. Gabain regarded eli as Ior the country`, and as
a result Clauson translated it as I grew Ior (i.e. as a subject oI) the realm oI Qapan
and Elteri`. However, Bazin insisted that in Old Turkic the dative suIIix -qa/-kahas
the basic Iunction oI date, but this expression can mean royal country` and the reign`
with elempire` (Bazin 1991: 159). Giraud (1961: 66, 120) thus analysed that a ruler`s
name+-eli means in the reign oI (ruler)`. Tekin also translates eli oIT(1)as
y|netimi srasnda` (in the reign oI`) (Tekin 1994: 23), which means that this part
can be I was born (and grew up) in the reign oI Qapan Elteri qaan` (GOT: 291;
Moriyasu2009:4748).Butthelastphrase4apan (lteris qaan isuncertain.Asis
well known, 'Qapan and 'Elteri are titles oI rulers oI the Second Turkic Qaanate,
and Clauson thus translated it as I grew up in the reign oI Qapan Elteri qaan`. But
as Bazin explained, the word qlnthas the basic meaning oI he was born` or he grew
up,butIoraperson,theIirstoneispreIerred.Accordingly,IselectthemeaningoI
he was born`.
9
InthiscaseasBazintold(Bazin1991:159160),apersoncannotbe
born in the reigns, oI two qaan, so we must consider either one oI the two titles as an
epithet. It thus means I was born in the reign oI Elteri qaan who was a conqueror`
or I was born in the reign oI qapan qaan who organized the people`. Bazin consid-
ered qapan to be an epithet oI Elteri,buttherearenohistoricalsourcestoattesthis
view (Bazin 1991: 160). Regarding this problem, Japanese colleagues Onogawa and
Sawada have pointed out that Mochuo , B|g or also had the title oI elteri,
Xieaielishi*xfiat, *tiat,*li, *i inLMC(Pulleyblank1991:341,79,188,
282; Onogawa 1950: 435436; Sawada 1983b: 8085).
10
Qapan qaan was given
the title Xieaielishi, aanyu, libaoguo, hehan by the
Chineseempress(Wu:etian) in 696 (==TJ-205: 6510; JTS-194: 5167; XTS-
215: 6044; TD-198: 5434). Contrary to this, Elteri qaan did not bear the title oI
qapan.Thus,inthiscase,4apan (lteris qaan eli qwlwtwm can be considered as I
was born in the reign oI Qapan qaan who organized the tribal peoples`. This means
that the writer oI this sentence shown with a personal suIIix -mwas born in the reign
oI Qapan qaan (691716).
W-4: eletmi yabu : ol : bara : taman cor : yoa : inisi : bilg : bara : taman :
tarqan : yoa : at : be : yetmi : cim atm :
Here are names oI (letmis yabu`s Iamily. His son was named sbara Taman Cor
Yoa and his younger brother was named Bilg sbara Taman Tarqan Yoa.Here,this
writer oI the Ongi site stele introduced his Iamily. In my investigations oI Ramstedt`s
rubbingandJadrincevsoriginalrubbingin2005,IcouldnotconIirmtheletterssuch
asYBGW.insiandY M G L G. (. . L) t m s. asClausontransliterated(Clauson
1957:Pl.7,182).
As to this title (letmis yabu, as Sawada stated, it can be a proper name hav-
ing the basic meaning organizer oI tribe`, and the proper name (lteris qaan basi-
cally means qaan that assembled the tribal people`. Regarding the historical back-
ground oI this naming, I have to state the related historical Iacts. As we know Irom
the Tunyukuk inscription (T: 5; Tekin 1994: 34), Qutlu launchedaninde-
pendent war against the Tang Dynasty; he had a aa.AIterheascendedtothethrone
as(lteris 4aan with the help oI his adviser Tunyukuk, he settled in a camp named
Baiaaochuan (White road steppe) oI the northern steppe oI theYinshan Mountains
9 Rybatzki considered qln-toderiveIromql-+-n-; usually meaning in a passive sense to be made,
created` and the like, but in the early period also (oI a child) to Iorm its own character, to grow up` (Ry-
batzki1997:75,n.216).
10 P.PelliotreconstructedMochuoIrom the Chinese expression as Bg cor onthebasisoIOldTibetan
sources.ButthistitleshouldberevisedasBg cor oIDru-gu,notasbg cor (Moriyasu1977:3,9,n.
17).This4aan canbeidentiIiedasBg 4aan that can be attested with Qapan qaaninT:34,50
(Tekin1994:1415,n.44,2021).
182 Takashi sawa
(uay y). In KT-E (1314), it is mentioned as Iollows: ecm apam . trsince .
yaratmws . bosurmws . tlis . taraus (boaun . anta etmis) yabu . saaw . anta bermis
(Tekin 1988: 1013). From here, when on the basis oI the traditional law Elteri qaan
couldconquertheeasternregionsnamedtlis, and the western regions named tarau,
heappointedayabu tobetheleaderoItheIormerregions,andaaatobeanother
leader oI the latter ones (Tekin 1988: 1013).According to inIormation Irom Old
Chinese chronicles, Elteri qaan gave the title oIaatohisyoungerbrotherMochuo
(B|g or), later 4apan 4aan and gave the title oI yabu to another younger
brothernamedDuoxifu,*tut,*siet,*biuk(GSR: No. 496h, 1257a, 933m),
*TursBg (JTS-194: 5166; XTS-215: 6044).
11
Fromtheserecordsitappearsthattlis
yabu as told in the inscription was Duoxifu,andtaraus yabu was his elder brother
Mochuo.The(letmis yabu oI the Ongi text was thus the tlis yabu that Elteri qaan
appointed to organize the eastern regions oI the Second Turkic Qaanate. Then aIter
Elteri qaan died in AD 691 and Mochuo became a Turkic ruler as his successor,
Mochuo qaan gave the appointment oI tlisaatohisyoungerbrotherDuoxifu,and
tarau aa to his nephew, and the Iirst son oI Elteri qaan Mofu , later Bilg
4aan (JTS-194: 5169; XTS-215: 6046). But according to the Orkhon text, Moju was
appointedtarauaaat the age oI 14 (KT-E: 17; BK-E: 1415), thus gaining this title
inAD697.ItisgenerallypresumedthatDuoxifubecametlisaaatthebeginning
oItheenthronementoIMochuo (Onogawa 1950: 440441), but he may have gained
thistitleinthesameyear,697,inadditiontoappointingMojutobetarauaa.Inthis
case,itispossiblethatDuoxifuwas tlis yabu duringtheperiodsAD687to697.
Then in 716 Qapan qaan was killed by a warrior named Xie:hile oI the
Bayrku tribe oI Tokuz Ouz in the steppe along the Tula River oI Mongolia. His
head was sent by a Chinese oIIicer named HaolingchotothecapitalcityoItheTang
Dynasty (Changan) in June 716 (JTS 8: 176; XTS-215: 6049; TD-198: 5439).
When the lesser qaan Inl, son oI Qapan qaan was about to become a great
qaan, K|l Tigin, the second son oI Elteri Qaan arose in rebellion against Inl qaan
and murdered him, his younger brothers and relatives (JTS-194: 5137; XTS-215:
6049; TD-198: 5439). In the conIusion, Inl`s relatives and the Tokuz Ouz tribes
escaped Irom Mongolia to the northern Chinese regions along the Huang-hu River
(XTS-215: 6052; TFCW 986: 11583; cI. Iwami 1998: 226278), and he was head oI
theShierXingTufue (Union oI twelve Turkic tribes) (Haneda 1957: 374
375; Moriyasu 1977: 3, 1314; Iwami 1998: 217). But we cannot Iind Duoxifuamong
Turkic peoples that escaped into northern China. Thus we can only admit that Duoxifu
11 The (letmis Yabu oI this inscription can be attested as the man buried at the site and named
Duoxifu, who had been appointed tlis yabuduringthereignoIelteris 4aanasaIounderoItheSec-
ond Old Turkic Qaanate. The third letter *biuk can be attested with the Turkic title bg,buttheIirst
and second words seem hard to reconstruct until now. Hamilton reconstructed the Iirst Chinese letter as
-dr-~ in the Old Turkic periods (Hamilton 1988: 168). The second letter siisreconstructedbyKarlgren
as*siet, whereby the Old Turkic name auoxifucanbereconstructedas*Tursbg while Marquart consid-
eredittobeTusikbg(Marquart1898:37).
must have been killed by K|l Tigin as one oI the close relatives oI Inl qaan`s Iamily
in716.
The word Bilg can be an Old Turkic title meaning counsellor` (Clauson 1957:
188), but not the adjective wise` as regarded by RadloII, Onogawa (ATIM 1895:
248; Onogawa 1950: 442). This title can be attested in many instances in Old Turkic
periods as 'Tun Bilg (oir inscription), bara Bilg Kli or` (K-W: 4) etc. (cI.
Rybatzki 1997: 75, n. 215).
Onogawa regarded the personal name Yoa oI sbara Taman Tarqan Yoa to
mean a son oI Qapan qaan (Onogawa 1950: 440), Yangwoshi teqin,
*Yoaci Tigin who had been sent as an ambassador to the Tang Dynasty in AD 713
and had died there according to Old Chinese chronicles (==TJ-210: 6686; XTS-215:
6047). Sawada (1983b: 88) doubted this, but I cannot agree with him because this
partcanbepositionedastheintroductionoIsonsoI(letmis Yabu,i.e.theyounger
brother oI Qapan, Duoxifu,andhisIamily.Moreover,Yoa lacksthelastonomastic
suIIix ShioI*Yoaci.Yoa isalsorecordedashisyoungerbrothersnameasbara
Taman Tarqan Yoa, who was the second son oI (letmis yabu.Fromthissituation
it appears that this word can be a personal name or an Old Turkic title that has never
beenattested.AstothetitleoITamgan,itcanbeattestedintheIirstlineoItheeastern
side and other lines oI the Kh|r-Asgat inscription, and it seems that it means having
a tama` (sawa 2010: 2223, 36).
W-4: : s: i : ol : taman W-5: bu : tabacda : yriy : tg ouz : ara : yeti rin :
ya : bolm :
This part oI W-4 can be read only in Ramstedt`s rubbing and it is connected with the
beginningoIlineW-5.Itranslateda s assi consistingoItheverb-meaning to
amble` (ED: 255b) inIinitive suIIix oI the second person singular -.Itseemspos-
siblethatolw oIolw Taman indicates son oI (letmis yabu.TheTaman canmean
(bara) Taman (Cor Yoa) or his younger brother (Bilg bara) Taman (Tarqan
Yoa)in line W-4. But it is questionable that he would have been the younger brother
who was also the writer oI this stele, because their campaign against their enemy
between northern China and Ouz can be dated to the period oI AD 686687 in the
reign oI Qutlu, Elteri qaan. This situation does not suit the Iact that his younger
brother was born in the reign oI Qapan qaan. The Taman oIthispartcanthusbehis
elderbrother(bara) Taman (Cor Yoa). This view can be supported by the usage
oIthe-mistyle.ThisuseoI-miindicates that this writer happened to hear this Irom
anotherperson,possiblyhisIatherorIromrelatives.ThispersonhavingthetitleoI
taman must have been more than 15 years old during the reign oI Elteri qaan, and
thusmusthavebeenbornintheperiodAD630667,becauseaboycouldbearthe
adultnameratattheageoIatleast15atthattime(Bazin1991:107108).
Regarding the word bu,althoughClausonregardsB (s a)(basa) then`, I cannot
conIirm this letter. From Ramstedt`s rubbing, we can read BW,asreadinJadrincevs
originalrubbing.ThisistobereadasademonstrativeadjectiveoITabac.
184 Takashi sawa
Until now, the passage T g has been discussed and the proposed hypothesis
among philologists is a problematic issue as Clauson and Sawada have pointed out
(Clauson 1957: 185186; Sawada 1983a: 6061).
12
TheIirstletterTisabackconso-
nant with a back vowel as T W G (tu) engel (barrier)` in T: 26 (Tekin 1994: 1213;
cI. Alylmaz 2005: 212). However, in this stele, the Iollowing letter is shown with the
Irontconsonantgthat is contrary to the vowel harmony oI the Old Turkic language.
As Iar as we know, there is no evidence oI such a combination in old runic inscrip-
tions. According to the context, however, it must be a kind oI postposition because the
preceding word has a dative suIIix such as -(y)a oIyr.Someresearchershavethus
translatedtgandthepostpositiontapaasup to` (Onogawa 1950: 446, n. 13). This,
however, is still a problematic issue. This can be a mistaken Iorm in runic spelling,
asMoripointedoutthattherearesomemistakenspellingsorIorgottensoundsinthe
runic inscription (Mori 1992: 515, n. 15, 587595). In this text, this curious usage oI
Tgis conIirmed three times. In my view, this can be considered as the postposition
tg(Orkun1936:128).Inthiscase,thebackconsonantTmightbeusedintheplace
oItheIrontconsonanttto coincide with the back vowel sound /a/oIyraya.Inmy
view, iI this word is transcribed as tg up to`, this can be regarded as a variant oI
the postposition tgi up to; as Iar as` (ED: 477; Tekin 2003: 154, No. 358) oI Old
Turkic with diIIerences between Turkic dialects or the periods. For example, it is well
known that there are some kinds oI tgisuchasaegin,aek,aegi,aak-e kadar` in Old
AnatolianTurkish,aein,aek,aak-e kadar` in Osmanli Turkish, aein,aek-e kadar`
intheTurkmenlanguage,-aan/-an/-aak/-ak -e kadar` (cI. Li 2004: 471472) and
tek kadar, dein` in the dorum dialect (TDK, Derleme S:l XII, 1982: 4747)
in colloquial Turkish. Because until now this word had not been analysed Irom this
viewpoint, I would like to suggest that this word is a variant oI tegithatreIlectsthe
colloquialpronunciationatthattime.
Withregardtoytirin, RadloII read ytirnasthesubjectoIthissentence,and
Clauson and Sawada concurred (ATIM 1895: 248249; Clauson 1957: 182; Sawada
1984: 98). Onogawa, on the other hand, regarded it as an instructive Iorm, reading it as
by seven warriors` (Onogawa 1950: 443). I agree with Onogawa`s view, because, as
is well known Irom old runic inscriptions, it is seldom that a certain numeral is added
toapluralnoun,andthusyetirn (seven warriors) cannot Iit this rule, although n
might be an Old Mongolian loan word with a plural suIIix. At present, I would like to
consider this Iunction oI the suIIix -in as the comitative Iorm with`.
This sentence as a whole can thus be interpreted as my son, (Bilg bara)
Taman Tarqan (or Yoa) became enemy with seven warriors among Ouz tribes
as Iar as the northern regions oI this Tabac (China)`. This situation reminds us oI
expressions such as qawm qaan . yiti yegirmi rin . taswkmws . tasra yorwyur. tiyin
. k esiaip . balwqaaqw . tawqmws . taaaqw . inmis . tirilip yetmis r bolmws (KT-E:
12 Bazylxan (2005: 56, n. 110) made transcriptions such as at eg in the W-5, 6 and 10 lines oI this stele
andinterpretedtheethnosashavinghadhorsesandanimalsIortransport,althoughthereisnoevidence
oIthisinhistoricalsourcesoItheOldTurkicpeoples.
1112; Tekin 1988: 1011) and yeti y: . kisig . uau:uma . uluu . saa rti . aywl
tiai . aywmasi . bn rtim . bilg tunyuquq (T: 45; Tekin 1994: 23; Onogawa 1950:
440). From this description, we can regard that this incident can be dated to the years
AD 682686 at the beginning oI the reign oI Elteri qaan, and the enemy oI that time
could be the Tang Dynasty.Accordingly, this description points to the rebellion oI
TurkictribesagainsttheTangDynasty.
W-5: qam baa : trikn : yan : anta : yorm : iig : kcin : bermi : rtmi :
Regarding qawm, Bazylxan`s transcription as aqawm isgenerallyaccepted,consider-
ing aqa as the cognate oI ake Iather` in the Kazakh language (Bazylxan 2005: 56,
n. 111). But this analysis entails a problem because he neglected the existence oI the
sound, so I cannot agree with him.
AsmentionedaboveinthenoteonW-4,qawm (my Iather) can be identiIied with
(letmis yabu that was the Iather oI two sons. Baa trikn,inturn,mustbetheone
thatappointedtlis yabu, i.e. Qutlu, Elteri qaan, as suggested by Onogawa (1950:
437).
Concerning this, V. Rybatzki mentioned that 'the hitherto accepted dating oI the
Ongiinscriptiontotheyear720or730mightnotbecorrect,asIorpalaeographicas
well as onomastic reasons the inscription might belong to the beginning oI the period
oI the Uigur Steppe empire (Rybatzki 2000: 209). Here, he considered that the title
trikn could not have been Irom the Second Turkic Qaanate period although we
can conIirm this word in the runic part oI the headstone oI the Qarabalgasun inscrip-
tion (Moriyasu & Ochir 1999: 219) and other Old Uigur documents (Rybatzki 2000:
210211). Rybatzki`s viewpoint, however, is disproved by the existence oI a tama
consisting oI a ram and a snake. As well known, this tama belonged to the collateral
line oI the Ashinas Iamily oI the Second Turkic Qaanate as conIirmed in the oir
inscription (Kljatornyj 1971: 250253; Alylmaz 2005: 20, photo 13), both sides oI
the tortoise stone oI the Mukhar inscription (sawa 1999d: 146, Pl. 7ac) and the
second Qarabalgasun inscription (sawa 1999e: 133134, Pl. 6; Alylmaz 2005: 20,
photo 14). Moreover, the word trikn canbeattestedintheHangitaHatinscription
on a rock near Dashinchiren som at Bulgan aymak in Mongolia.This rock inscrip-
tion had been investigated by Rintchen, Bazylxan, Shinekh, Kljatornyj and TIKA
(Rintchen 1968: 37; Bazylxan 1968; Shinekh 1979; Kljatornyj 1975; TIKA 2003:
340342; cI. Battulga 2005: 75). I also researched this rock in August 2008. There I
couldreadaphrasesuchasBa: 4an Olw Tri ucmws (It is said that) the son oI Baz
qan went away to the sky (i.e. died)` on the second line and Bg r, Triknim . . .
(Bg r, my Trikn . . . on the Iourth line. A ram-styled tama Ior the Ashinas
royal Iamily was carved evidently on the leIt over the inscription. Moreover I could
conIirm the Old Turkic runic text Irom the eastern, northern and cover boards oI
sarcophagus oI the Olonnuur Kh|ndii site oI Galuut som oI Bayankhongor aymak in
Mongolia in 2008. This site has the typical Ieatures on the Second Turkic Qaanate
periodasevidentlysuggestedbythreestonestatues,balbalstonesandsarcophagus
186 Takashi sawa
with runic text. From the top oI the eastern sarcophagus, I could read z triknim,
a. ya: yrim, a (My Tngrikn, ah My brown Earth, ah`) (sawa 2009: 197198).
From this, we can date this inscription to the Iirst halI oI the 8th century, that is, to
the Second Turkic Qaanate as suggested by Kljatornyj. We can thus say that this
Ongi stele can be dated to the Second Turkic Qaanate. But until now it is unresolved
whether trikn canbeattestedasTiannan, son oI TriorTri 4an (cI.
Haneda 1957: 372374; Sawada 1983a: 71, n. 29; Rybatzki 2000: 210211; Moriyasu
et al. 1999: 220). At present, it is clear that this title was a special epithet oI qaan
Irom the similar uses oI this title in the runic inscriptions oI the Old Uigur Qaanate
and Old Uigur documents oI Koco Uigur. I do not discuss this etymology any Iurther
here, but, as Onogawa presumed (Onogawa 1950: 441442), this word must have
been a special title Ior indicating respect to a ruler, i.e. Qutlu, Elteri Qaan oI the
Second Old Qaanate..
Historically, the campaign against Tokuz Ouz mentioned in this sentence can be
datedto686687IromtheTurkicOrduinthesteppenamedBaiaaochuan(Whiteroad
steppe)oIKaraQum(Blackdesert)inthenorthernsteppeoItheYinshenMountains
(uay y) to the steppe around the Hangai Mountains (gtkn y; gtkn yer) oI
Mongolia across the Gobi Desert and the Ongi River (Iwasa 1936: 130131).
W-5: : tardu : boltuqda : W-6: triknk iig berti : teyin : yarlqam : ad at :
anta:bermi:boltuqda
This part oI W-5 could be Iirst conIirmed in Ramstedt`s rubbing, and it is connected
with the beginning part oI line W-6.
The word tarau means Iirstly the western territory oI the Turkic Qaanate`,
andtarauboltuqaa accordingly when the western regions (tarau) were Iormed (or
organized).ThisphraseremindsusoIasimilarphrasesuchastlis . taraus (boaunu
. anta etmis) yabu . saaw . anta . bermis (KT-E: 1314). So in this place, taraucould
beanomissionoItaraus (boaun).Whothenorganizedtarauboaun(thepeoplesoIthe
western regions) at that time? According to Old Chinese sources, it was Mochuo(later
Qapan qaan), or Mofu (later Bilg qaan) who became ad oI tarau (the western
regions oI the Qaanate), not Duoxifu,i.e.(letmis yabu.InChinesesources,Duoxifu
was never given the title oI tarauaa. As a result, this view does not Iind support.
Is a diIIerent translation possible? We can oIten read tarauasapersonalname
derivingIromtheoriginalmeaningsuchastarau,kltarau,tarauaaetc.Fromthis
usage,thistaraucanbepartoIapersonalnameoratitle.IIso,tarauboltuqaameans
when a man named taraubecame a ruler` or when he became tarau(ad).Inthe
Iollowing place, trikn givesthetitleoIaato(letmis yabu,i.e.Duoxifu.Onthe
basis oI Chinese chronicles, it was 4apan qaan who gave the title oI aatoDuoxifu
in699.Thus,trikn can be identiIied with Mochuo, Qapan qaan. According to the
Iirsttranslation,amannamedtaraucanbeMochuo who became a ruler in 691, not
Mofu (later Bilg qaan). In the latter translation, he was Mofu who became tarauaa
in 697. I cannot say which translation is better. At this point, I would like to translate
it literally as the peoples oI the western regions oI the Qaanate were organized`, but
thismeanstarauaa(i.e. Mochuo)becamearuler.Thispartcanthusbedatedto
691, when Mochuo became Qapan qaan (Sawada 1983a: 5759).
The one who 'yarlqam (commanded) can be Qapan qaan although it is also
possible that this is 'Tri (God) iI we compare it with the phrase Tri yarlqaduqn
cn (KT-S: 9; BK-N: 11; T: 40; Tekin 1988: 45, 2829; 1994: 1617).
W-6: : toquz : ouz : tg : ya rmi : bedk rmi : qaan : tg(mi)
TgiscarvedasTgand it seems problematic because there is no dative suIIix in the
preceding word ou:.IItgis a postposition, it requires a dative suIIix -qa/-ainthe
preceding word. Clauson, Tekin and Sawada considered it to be a tribal name that
cannot be attested in other Turkic inscriptions. At present, I would like to consider it
tobeavariantoIapostpositiontgiup to` as I mentioned above in the note oI W-5.
This scene can be dated to the reign oI Qapan qaan as I mentioned above.
So this accident can be related to the campaigns against Tokuz Ouz in the reign oI
Qapan qaan.
W-6: biz bdk biz : biz bat biz : biz : W-7: yabz bat biz : az kg : k|rtg :
This part oI W-6 can be read only in Ramstedt`s rubbing and can be connected with
theIirstpartoIthelineW-7.Thispartcanbedividedasbi:bakbi:,bi:batbi:,bi:
yab:batbi:,a:kgkrtg. From this, we can read that Qapan qaan`s warriors
were psychologically aIraid oI marching against the strong Touz Ouz army because
oI their lack oI warriors. But (letmis yabu who had the title oI tlisaacommanded
them not to be aIraid and marched to Tokuz Ouz.
W-7:irti:sltim:terermi:
AccordingtoClauson,theverbir- means to mope, Ieel lonely or bored` (ED: 194a).
I regard this verb as an expression oI the Turkic peoples` being despondent possibly
because their army had Iew warriors. However, the subject oI sltim(Imarched)can
be(letmis yabu and Irom this we can understand that he wanted to tell oI his great
contributiontohiscountry.
W-7: amt bglrim a : ter rmi : biz az biz : teyin : alqonur : (rtimiz) :
These addressing words can be also attested in the Orkhon inscription when Bilg
qaan ascended to the throne in gtkn y in 716. Tekin regarded it as matmeaning
royal, IaithIul` with a question mark (GOT: 355). But he also translated amtas now`
in KT-S: 3; KT-E: 9; BK-N: 2; GOT: 302. He also translated the same spelling mat
asroyal, IaithIul` in (BK-S: 13, 14), while he changed this word to amtasnow` in
the same place (Tekin 1988: 5455). In my view, mticanberegardedasapartoI
188 Takashi sawa
tlis yabu`s address to his soldiers who were despondent Irom marching, and I would
like to regard it as the word amt,meaningnow`. The word acanbetheinterjection
a (Tekin 2003: 161), not a dative suIIix as stated by RadloII, Onogawa and others
(ATIM 1895: 249; Onogawa 1950: 443). The word alqon-canbeconsideredasavari-
antoItheverbalqn- to use oneselI up, exhaust oneselI; to be used up, exhausted`
(ED: 138b). In the context, this part indicates that Qapan`s army exhausted them-
selves on this campaign. In the next place, it is possible that three or Iour letters were
carved but there is no evidence oI what letters were inscribed. rtimi:, however, is to
be expected.
W-7: : yoryn : s : slyin : W-8: qam : (ad) : anca : |tnmi : trikn : almazn
teyin:
This passage oI W-7 can be attested only in Ramstedt`s rubbing, and it is connected
with the beginning part oI the next W-8 line. 4awm saa(myIatherad)indicatesthat
theIatheroIBilg sbara Taman TarqanhasthetitleoIaa.Thissituationsupports
the view that this incident can be dated to the reign oI Qapan qaan as I clariIied in
thenotetoW-6.Thistrikn can be Qapan qaan.
W-8: bodun : anta bermzi : tusul r
ThispartcanbelongtoDuoxifu`s saying. And the word Iollowing berm:inisan
interjection, not a dative suIIix as other researchers have considered until now.
W-8: (atacm) a
Regarding atacwm, I discussed this identiIication in connection with the Iirst line oI the
horizontalinscriptionoItheOngistele.
S-1: qamuq : balqqa : (t)gdim :
Clauson transliterated Q M / Irom RadloII`s Atlas, and he interpreted the uncertain
thirdletterasL,notasuQ,andreadqamul;haminamed yiwu*i-nguo (MEC)
(GSR : No. 604a, 58I) as an oasis country oI Chinese Turkestan Irom the context
(Clauson 1957: Pl. VI, 182, 186). However, Irom Ramstedt`s rubbing we can read
QmuQ, which I could also conIirm Irom Jadrincev`s original rubbing at the Institute
IorOrientalStudiesoISaintPetersburgin2005.Sothispartshouldbereadqamuq
asavariantoIkama all` (ED: 627a). According to Clauson, the Iorm qamuqwith
labial vowel attraction does, however, occur as early as the Turkic inscriptions. As to
this sentence, I can assume that it can be considered as a historical Iact thatTurkic
peoples went on warring campaigns against enemies in surrounding regions between
Kaarkan y the Daxinanling Mountains` in the East and Tmir 4apw The Iron
Gate` in the West. There were campaigns against the Tang Dynasty and the Qrz as
mentionedintheOrkhoninscription(KT,BK,andT).(letmis yabu, however, had
thetitleoItlisaawith control over the Eastern territories oI the Turkic Qaanate,
which suggests the possibility that he conquered Chinese cities.
S-1: qonuldm :
Regarding this verb, RadloII read aqunlaam to mean I attacked`, and Onogawa read
uqunlaam as I plundered` (Onogawa 1950: 443, 447, n. 15). In historical terms, I
can remember an expression such as tknyerigkonm (they settled down in the
gtkn country`) (T: 17; ED: 632b). This description tells us that the Turkic army
occupied the north or northwestern regions under the Tang Dynasty and laid siege to
allthecitiesoItheseregions,asImentionedinthenoteonthesamelineoItheOngi
stele.
S-1: mz : rinc : iig kcig bert(im) S-2: ekin ara : tg ya : bolm : tgmci teyin :
saqntm :
This part oI S-1 can be read in Ramstedt`s rubbing, and it is connected with the begin-
ning oI line S-2. In this, the one that 'served can be the writer oI this stele, Bilg
sbara Taman Tarqan. He is written with a pronoun suIIix oI the Iirst singular -m.
The one whom he served can be Bilg Qaan, the son oI Elteri Qaan. This can be
suggested by the context oI the next lines. The word Tgisregardedasthevarianttg
oIthepostpositiontgi (up to`) as in line W-5.
S-2: tri bilg : qaanqa :
Tri Bilg 4aan can be the elder son oI Elteri qaan, Muju, Bilg Qaan as
Onogawa and Bazin considered (Onogawa 1950: 438439; Bazin 1991: 153). Bilg
Qaan was indeed recorded as Tri Bilg 4aan,asDobrovitspointsout(Dobrovits
2000: 150). In this inscription, however, as Trikncanbesuggestedastheepithet
Ior Elteri Qaan, it is possible that the title Tri is used as the epithet Ior Bilg
Qaan who was the son oI Elteri Qaan as an honoriIic title (Onogawa 1950: 438). In
my view, the epithet tri is a shortened expression oI tri t . tria bolmws . trk
bilg qaan (BK-N: 1; Tekin 1988: 2829).
S-2: ancaqa : iig kcg : bersgim : bar rmi : rinc :
Theadverbancaqwa can be read in Ramstedt`s rubbing. I would like to express my
gratitude to my Japanese colleague Kousetsu Suzuki who advised me oI the possibil-
ity oI this word in 2005. It means a very little` (ED: 175a). Until now, everyone have
read anca like this` since RadloII (ATIM 1895: 250251). This reading makes the
unresolvedmeaningoIbrsigimbar clear (sawa 1999a: 135136). Erdal considered
brsigim impossible Irom the viewpoint oI Turkic grammar, because this Iorm can
190 Takashi sawa
be attested in an Old Uigur document (OTWF: 527, n. 147). In general terms, this
sentence expresses with hesitation that Bilg sbara Taman Tarqan intends to join
the army oI Bilg Qaan, because (letmis Yabu who was the Iather oI Bilgbara
Taman Tarqan, the builder oI the Ongi site and stele was killed by K|l Tigin, the
younger brother oI Bilg Qaan in June 716. However, he Iinally decided to go on
a campaign on the side oI Bilg Qaan`s army. From this, we can understand that he
was in a complex political condition under Bilg Qaan.
S-2: bim tgdkm : url qal : adrlmalm : tyin : qaanda : adrlmalm : teyin :
tgdim.
This part can be read in Ramstedt`s rubbing and is connected with tgip . inima olwma
. anca tlaimoIS-3.The-loIurlw qalwisregardedasthecoordinatingconjunc-
tive suIIix -l with, and` (GOT: 124). We can conIirm the same Irom examples such
asini-li ci-li younger and elder brothers` (KT-E: 6), bgli boaunlw the lords and
people` (KT-E: 6), toruqbuqa-lsmi:buqa-l lean bulls and Iat bulls` (T: 5), tn-li
kn-li night and day` (BK-SE). This means that urlw qalwcanbeliterallyinterpreted
as son and Iather`. But as I stated, the Iather oI the writer sbara Taman Tarqanhad
alreadydied,andthisqalw can be understood as meaning the people oI the Iather`s
generation.Furthermore,url is also understood as the people oI the son`s genera-
tionbelongingtohisIamily.
It is also noteworthy that the last letter a oI the word qaanaaisinthelastposi-
tion in S-3, and the Iollowing phrase aarlmalm.teyin.tgaimgoesIromtheright
to the leIt and then reaching the leIt margin in S-4. Reaching the leIt margin oI the
surIace, the row oI letters curves upwards, making a complete about-turn and running
eventually in the opposite direction and turning the letters upside down. In my view,
this part narrates that when sbara Taman Tarqan returned to his camp along the
Ongi River, he requested his companions and his Iather`s companions to join the side
oI Bilg Qaan`s army.
S-3: qa yorp : elteri qaanqa : adrlmaduq : yalmaduq :
This part is related to the historical contribution oI Bilg sbara Taman Tarqans
Iather(lteris yabu to the construction oI the Second Turkic Qaanate under Qutlu,
Elteri Qaan. This part also indicates that this builder, Bilg sbara Taman Tarqan,
wanted to maintain a political position in the Bilg Qaan`s government.
S-3: tri bilg : qaanta : adrlmalm : azmalm : teyin : anca : |tldim :
Tri Bilg 4aan can be Bilg Qaan (716734 AD) as I mentioned above in the
notetoS-2.Thissentencetellsusthatsbara Taman TarqanadvisedhisIamilyto
Iollow Bilg Qaan and to guard his political position.
S-3: ilgr barma : bard : bilg qaan : bodun
ThispartcanbeIirstreadinourrubbingoIthelargeIragmentoItheOngistelethat
is kept at Osaka University (sawa 1999a: 136). This reading can also be conIirmed
in Ramstedt`s rubbing.
S-3: : bard : |lgin : atqa : iig kcig : bertim : adrlmaz : teyin : tgdim
This part can be read only in Ramstedt`s rubbing, and it explains that the builder
oI this stele also served Bilg Qaan to guard his political position. The expression
lginatqa, meaning literally to die, Ior his Iame`, tells that the writer oI this stele
Bilg sbara Taman Tarqan served Tri Bilg Qaan to deIend his political stand,
and he went to the Iront prepared to die. This phrase can be compared with the expres-
sionlyit |lesiye yitesiye oaltm` (KT-E: 27; Tekin 1988: 14).
S-3: adrlmaz : teyin : tgdim
This part is reversely carved, i.e diIIerently Irom the normal order oI runic letters,
in the Iollowing order Irom leIt to right and joining the Iormer sentence in the last
circularcornerasinmagt.nyt.:mLRD..Thisistobereadinreverseorderas
DRLm:.tyn.tgam.
S-4: z : tri : qan : li : ylqa : yetinc : ay : kclg : alp r : qaanmda : adrlu :
bardz :
Regarding : tri qan, as Onogawa (1950: 439) analysed, tri qanshouldbeiden-
tiIied as the subject shown in the verbal suIIix oI the second person plural -ww:oIthe
verbaarwlu baraww:,thatis,(letmis yabu. As I mentioned and as Iar as we know,
Duoxifunamed(letmis yabudidnotbecomeqaan. Accordingly, this verbal suIIix
-ww: should be interpreted as a respectIul expression reIerring to his Iather Eletmi
yabu by the writer oI the stele. From this we know that Eletmi yabu was also
addressed with Tri 4an as a title oI respect. The same use oI qan/qaan can be
conIirmed as big qaanm in the 23rd lines and bilg klg r dg qan in the
56thlines.
13
Concerningliylqa which have been regarded as uncertain words by Clauson,
these letters are, however, evidently read in Ramstedt`s rubbing and Jadrincev`s origi-
nal rubbing conserved in St. Petersburg. The word liIromtheChinesecharacter
13 The words : tri qan were considered by Sawada to mean To the sky where Tri Qan lived`,
andhealsointerpretedTri 4anastermoITrikn. He considered the verbal suIIix -ww:asaplural
Iorm, i.e., he interpreted that these peoples were the Turkic peoples Iollowing the DuoxifussonsIamily
that became the enemy Iorces oI Bilg Qaan. This, however, should be not interpreted in such a complex
manner as suggested by Sawada. This passage should be read more simply as (letmis yabudiedinthe
reignoIBilg 4aan.
192 Takashi sawa
long, *liwong in MEC (GSR: No. 1193a) means dragon`, which is Ieatured in the
traditionalChinesecalendar.ThistermisalsorecordedasWLWYilQ(uluylqa)
in the Terkh inscription (W-2) oI the Old Uigur Qaanate. On this part oI the Ongi
stele, Bazylxan read it as li (Bazylxan 2005: 57, n. 112), but there is no letter in
thisstele.ItisthereIoretobereadliterallyasliylqa.Withregardtohistory,Ican
deIinitelysaythat(letmis yabu,theIatheroIBilg sbara Taman Tarqan who was
the writer oI this stele died in June oI the Year oI the Dragon (AD 716) when K|l Tigin
launched a coup d`tat against the successor, nl Qaan, the son oI Mochuo, Qapan
Qaan and his relatives as assumed by RadloII, Marquart, Onogawa and Sawada
(ATIM 1899: IX; Marquart 1898: 38; Onogawa 1950: 438; Sawada 1983a: 6869),
although Bazylxan considered the 'Year oI the Dragon to have been 703 between
681and716.Astorinkclg alp r qaan, this letter can only be read in Ramstedt`s
rubbing. This qaan can be identiIied with Bilg 4aan whom sbara Taman Tarqan
servedaIterthedeathoIhisIather(letmis yabu.
S-4: bilg : atacm : you : qorn : qazantm : el yetti :
This passage can be read only in Ramstedt`s rubbing. It indicates that the son, bara
Taman TarqanheldaIuneralceremony,receivedhisIathersterritory,andorganized
hiscamp.Itindicatesthatsbara Taman Tarqan was publicly admitted as the legiti-
matesuccessoroIhisIather(letmis yabu.
S-4: tri : z : trikn : tabac : k|rr : rti :
This part can be read only in Ramstedt`s rubbing, and tri . : . trikncanbe
identiIied as Qutlu, Elteri Qaan as mentioned above. This part indicates that around
AD 686687 in the reign oI the Iather oI Bilg Qaan, Turkic peoples could be inde-
pendentoItheTangdynasty.
S-4: (taba)c k|rr : rtim : dg : atacm :
This part can be read only in Ramstedt`s rubbing. It indicates that Bilg sbara Taman
Tarqan under the rule oI Bilg Qaan engaged in a warring campaign against Northern
China although it seems that there is no record oI battles between Turks and the Tang
DynastyinOldChinesechronicles.Here,thesonsbara Taman Tarqanheldtheter-
ritoryoIhisIather,(letmis yabu,andkepthisorauneartheOngiSteppeinsouthern
Mongolia.
L-2: toqdmz : cim boyla :
As is well known, the word cimeansjuniorpaternaluncleorelderbrother,andBoyla
is a high oIIicer`s title oI the Turkic Qaanate. For example, we can remember the title
Boyla Baa TarqanthatTunyukukhadinthelastperiodoIhisliIeasmentionedin
BK-S 14; T. 6 (Tekin 1988: 5455; Tekin 1994: 45) and Boyla 4utlu Yaraninthe
secondlineoItheSujiinscription(Mori1992:158,162).Thispartmayindicatethat
his uncle or his elder brother was a high oIIicer bearing the title oI Boyla.
L-3: basatp rig : uduuru : sancdm :
This phrase reminds us oI the scene oI a warrior and it can be supported by the sen-
tencessuchasbir rig oqun urtw . eki . rig uausuru sanctw (he shot one man with an
arrow and speared two men sending them one aIter the other) in KT-E: 36; KT-N: 2.
InthetranscriptionoIWDRW, I Iollow Clauson, although T. Tekin transcribed this
word as ua(u)ru (Tekin 1988: 82, n. 99; ED: 73b).
L-5: n : tg : bara : tarqan :
The Iirst element -n can be an animal name such as arslan, ylan that symbolizes
braveryandmanlinessbecauseaIterthepostpositionaltermtg (like`), a high-rank-
ingoIIicerbaraTarqanismentioned.Accordingly,////ntgshouldbeanadjective
modiIieroIamanhavingthetitlebaraTarqan.Asmentionedabove,thistitlecan
be identiIied with a young son having the Iormal title Bilg sbara Taman TarqanoI
(letmis yabuinW-4oIthisinscription.ThispersoncanbealsoidentiIiedashaving
constructedthesiteandasapersonhavingthetitlebaratarqaninthe11thbalbal
stone oI this site, with the short phrase BilgbaraTarqanbalbalandacoupleoI
tama designs oI a ram and a snake (sawa 1999b: 278282).
L-8: elteri : qaan :
Thisnamecanbe4utlu,(lteris qaan oI the Second Turkic Qaanate, but the title is
alsoanelementoIthetitleoI4apan elteris qaaninW-4.IIso,elteris 4aanmay
beatitleoIMochuohehan*muk,*tyatinLMC(Pulleyblank1991:218,63).
Notesconcernlngtbeborlzontalllnesontbesoutbsloeo|tbebeaostone
(1) atacmqa :
The word atacwm can be conIirmed in W-8 and W-4. RadloII read it as the proper name
Tacam (ATIM 1895: 252), but A. V. Gabain has suggested a new reading (Gabain
1953: 543). Clauson, Tekin and others accepted her view (Clauson 1957: 187; GOT:
256), although Bold read the word as atacwm. Gabain`s view can resolve the rela-
tionship between Eletmiyabuandatacwmasthesameperson,andthisreadingcan
be supported by the historical situation and relationship between the Iather (Eletmi
yabu)andhissecondson(Bilg sbara Taman Tarqan) in this runic text. BrieIly put,
inW-4atacwm was an expression with which the writer addressed his Iather (Eletmi
yabu), and the Ongi site was constructed to honour his Iather by his younger son
194 Takashi sawa
Bilg sbara Taman Tarqan.Moreover,asImentionedabove,Bilg sbara Taman
TarqanhadtherunicinscriptionoIbaratarqanbalbal carved in the twelIth line oI
balbal stones oI this site. This balbal stone had a tama consisting oI a ram and a snake
thatissameastheoneontheheadstone.ThisindicatesthathebelongedtotheEletmi
yabu Iamily oI the Ashina`s tribe (sawa 1999c: 278282).
(2) big
As is well known, with reIerence to the Orkhon Turkic Iorm b g intheOrkhon
inscription (KT-S: 8, 11, 12, 13; BK-N: 6, 8, 15; Tekin 1988: 47, 3035), there is
no similar example oI evidence oI the letter i between andgthatcanbeconIirmed
Irom the rubbings oI either Ramstedt or Jadrincev. However, Irom the Manichean
documents oI the Qoco Uigur period we can conIirm that mig vechnyj (eternal)`
(cI.DT:342a)canbeconsideredasavariantoIbig. I cannot say, however, why
this spelling was carved in this stele, and it may also be a mistaken spelling. As well
known, there is Iorm b k (bk) that was a variant oI bg in the Sulek inscrip-
tionoIKhakassia,andintheoldTurkicepitaphsoItheupperandmiddleregionsoI
the Yenisei Basin (E201; E4814, 15; E489, E278; cI. Kormuin 1997: 285; 2008:
318).Inaddition,b (b) is also attested in the Yenisei inscription (E42, 3; E30,
2; Kormuin 1997: 285; 318).
14
In my view, Old Turkic peoples did not have the same
orthography.TurkicspellingreIlectsthelocalpronunciationanddialectsatthattime.
This Iorm can be considered such an example.
(3) qaanm :
This word can be conIirmed in both Ramstedt`s rubbing and Jadrincev`s original
rubbing in Saint Petersburg. It seems strange that DouxiIu was addressed as qaan
althoughthereisnorecordassuchthathebecameqaan.ThisshouldthereIorebe
regarded as unusual usage. In my view, it is worth noting that the term tri qancan
be used as a respectIul expression on the part oI his son as I mentioned in connec-
tion with S-4. In this case, we should understand that this term was written Ior the
dead Iather in honour oI whom this stele was erected. In my view, qaan and qan
were respectIul expressions denoting the same amount oI signiIicance. As stated by
M. Mori, since the middle oI the 7th century, there were very Iew leaders that had the
title 'lesser qaan in the Turkic Qaanate, which can be explained by the Iact that
political matters were centralized under a great qaan at that time. So the title oI aa
was used in the place oI 'little qaan (Mori 1967: 274277, 374379). Accordingly,
qaanwm cannot show that he became a qaan as a matter oI Iact. Instead, this should
be the respectIul expression reIerring to his dead Iather.
14 Regarding this variety, my colleague Hiroaki Fujiie oI Osaka University has remarked that today`s
Karachay pronounce an unnoticed sound g between b and although they recognize the Iorm b
at the phonological level. From this, Fujiie supposes that this may be caused by the diIIerence between
phonologyandpronunciation.
(4)(5) l ylqa :
This year can be attested with the expression liylqa(theYearoItheDragon)oIS-4.
Previously,ClausonassumedthatthemanburiedatthissitediedinQonylqa(the
Year oI the Sheep). I cannot, however, read the letters such as Qny in the main text or
the additional text oI Ramstedt`s rubbing and Jadrincev`s original rubbing conserved
in St. Petersburg. As a result, we cannot accept the date oI construction oI this stele
thatClausoninsistedon(Clauson1957:Pl.VI,183).
7. Concluslons
From philological points oI view, these runic texts oI the Ongi stele can be generally
described as Iollows.
7.1. Arrangement o| tbe contents ln tet
(W-1) Yama qaan establishing the First Turkic Qaanate, his successIul campaigns in
theIourdirectionsandtheruinoIhiscountryaIterhisdeath(552630).
(W-2) Deserting their qaan, Turkic aristocratic people served the Tang Dynasty
as warriors (630679).
(W-3) The God Tri addressed the Turkic peoples who had lost his own qaan
andcommandedthemtobeindependentIromtheTangDynasty(679681).
(W-4) The statement on Iamily lineage oI the writer Bilg sbara Taman Tarqan.
(W-5) The war against the Tang Dynasty under Elteri qaan and participation in
this war (682691 AD).
(W-6)ThegrantingoIthetitleoIaaIorcontributingtothecampaignsinthe
reign oI Qapan qaan.
(W-7,8)TheIather(letmis yabuscontributiontothecampaignsagainsttheir
enemies in weak condition during the reign oI Qapan qaan.
(S-1)(letmis yabuandhisyoungersonsparticipationinthecampaignagainst
allcitiesundertheTangDynasty.
(S-2) The Bilg Qaan`s enthronement, and the writer`s addressing his Iamily to
serve Tri Bilg Qaan, and hesitation to participate in campaigns under Bilg Qaan.
(S-3) The writer commanding his sons and his younger brothers to join Bilg
Qaan`s army.
(S-4)ThedeathoI(letmis yabuandhisIuneralceremony,andthesuccession
oIhissonBilg sbara Taman Tarqan.
(E-1a8) The writer commanding his son and younger brother to serve Tri
Bilg Qaan and to strengthen his contribution to his country.
I can generally say that the Ongi text has many characteristic expressions as
well as other Orkhon inscriptions (KT; BK, T and K). Similarities with other runic
inscriptions indicate that this text was carved in the Second Turkic Qaanate.
196 Takashi sawa
7.2. Tbellneageo| Eletmi yabu anotbewrltero|tblsstele
As mentioned above, Duoxifu *Turs Bg named (letmis yabu was the youngest
brotheroI4utlu, Elteri qaan and Mochuo, Qapan qaan. Duoxifuhad two sons
namedsbara Taman Cor Yoa andhisyoungerbrotherBilg sbara Taman Tarqan
Yoa. The second man also had a son and a younger brother. The person Ior whom
this stele was built can be Duoxifu,i.e.(letmis yabu,and the writer oI this stele can
behisyoungsonandsuccessorBilgbaraTamanTarqan. Regarding Bilgbara
Taman Tarqan Yoa, as I mentioned above, it is doubtIul that he was the one who
erected the eleventh balbal stone with the phrase baraTarqan and a tamas design
consistingoIramandasnakesign.HecouldthusbenamedbaraTarqan.According
to inIormation Irom old Chinese sources, this named person can be identiIied with
a high-ranking oIIicer named under the Bilg Qaan government as Iollows: on the
dateoIBoshin(the7th),JulyoItheKaiyuan11thyear(AD723),Turkicambassadors
oI*k--nk it-puatakt kn in MEC
15
,thatis,baraTarqanoI
theAshinasclanandhis32companionsvisitedtheTangdynasty.TheTanggovern-
mentgavehimthetitleoIgreatgeneral,andgavehiscompanionsthetitlesoIgeneral
Iollowers, and then the Tang dynasty let them return to their native country, i.e. the
Turkic Qaanate.
16
From this, we can know that baraTarqan made Bilg Qaan one
oItheelitemembersoIhigher-rankingoIIicialsin727AD.Thisattestationcansup-
port my Iollowing suggestion regarding the date oI construction oI the Ongi site and
theageoItheinscription.
7.3. Dateo|constructlono|tblsslteanostele
Bazin Iirst analysed this text Irom philological and other viewpoints, and created
a hypothesis. According to him, the Turkic expressions oI this text are almost attested
in Orkhon Turkic (T: K, KT, and BK). The syntax has archaic Ieatures, permitting
adateearlierthanthatoItheTunyukukinscription(AD725726).Fromthecontents
oI this text, the deceased oI this stele was buried in the Year oI the Sheep, as based
on Clauson`s reading, i.e. AD 719 or 732. This text has many words oI the God Tri
ortheIather(letmis yabu advising his younger brother and sons to join with Tri
Bilg Qaan, not to part Irom this qaan, and not to betray him. Bazin thought this
reIlected the killing oI the son oI Qapan qaan and his intimate relatives by K|l Tigin
in July 716. For these reasons, Bazin regarded that this stele was built in 719 (the Year
15 GSR: No. 1m, 975a, 350a, 411a, 275d, 271b, 157a. Concerning the reconstruction oI this name, cI.
sawa 2010: 5051.
16 Intheoriginalsource,itisrecordedas723AD.:
(On the BoshinoINovem-
berinKaiyuanperiods,thegreattriballeaderoITujenamedAshina-IshbarTarqanbroughtatribute
to the Tang dynasty with his thirty-two companions. Tang emperor giIted the title oI General in chieI to
Ishbara Tarqan and the title oI the general oI palace horsemen to his attendants, and then the Emperor let
them to go back to their country`.) (CFYG-974: 3875.)
oI the Sheep) or in the Iollowing year (720) (Bazin 1991: 159161). In Bazin`s view,
as I mentioned above, 'the Year oI the Sheep cannot be conIirmed Irom Ramstedt`s
and Jadrincev`s rubbings. This should be the 'Li year, whereby (letmis yabumust
havebeenburiedintheYearoItheDragon.Inthe8thcenturyduringtheSecond
Turkic Qaanate, the Dragon year could have been 704, 716, 728 or 740. In the con-
tents oI this stele, as Bazin pointed out, the text reIlects the killing oI the son oI
Qapan qaan and his intimate relatives by K|l Tigin, making it possible that it was
carvedin716,728or740.AsImentionedabove,IromChinesesources,itseemsthat
Qapan qaan`s younger brother (letmis yabu was also killed by K|l Tigin, which
canbesupportedbythedescriptiononthisstelenoting:You((letmis yabu) went
away Irom my strong and brave warrior and wise qaan in the June oI the Year oI the
Dragon(li)(S-4).AccordinglytheYearoItheDragoncansuitthetimeoIdeath,
July 716, when the head oI the dead Qapan qaan was brought to the Tang Dynasty.
From this, K|l Tigin`s coup d`tat can be dated to May or June oI 716, aIter Qapan
qaan`s death. Thus, (letmis yabu also died in July as a result oI the battle between
the successors oI Qapan (Inl qaan) and Qutlu (K|l Tigin). (letmis yabusburial
ceremonyandIuneralceremony(Yo)mustthenhavebeenheldinthemonthoIJuly
in 716 or 717. From this stele, however, we can conIirm that Bilg sbara Taman
Tarqan was 'Eletmi yabu`s successor serving with much hesitation Bilg Qaan
who was an elder brother oI K|l Tigin. He Iollowed Bilg Qaan and went on cam-
paign against the Tang Dynasty. This hesitation tells us that Bilg sbara Taman
tarqan was suspected by Bilg Qaan`s relatives oI serving Bilg qaan, because his
Iather was killed by K|l Tigin in 716. He thus had to strengthen and emphasize his
service to Bilg Qaan to deIend his political position in the Second Turkic Qaanate
under Bilg Qaan. sbara Taman Tarqan was then appointed as a high-ranking
oIIicer by Bilg Qaan as a result oI serving the latter with loyalty. This meant that he
could build the stele to honour his Iather in the reign oI Bilg Qaan.
Atpresent,IcannotIindthekeytoresolvethedateoIbuildingthisstele,because
we no longer know what was carved on the eastern and northern sides. The stele was
carved Irom the western side, to the southern, eastern side and northern side, in the
order WSEN. From the viewpoint oI the direction oI lines, the Ongi text was
carved Irom the bottom to the top on all sides. This writing method is diIIerent Irom
that oI theTunyukuk inscription (ca.AD 725), KlTigin inscription (732), that is,
W (Chinese) S E N. From the viewpoint oI the direction in which the line
proceeds, the Ongi text can be read Irom the bottom to the upper part. This is similar
to the Kl Tigin inscription (made in July 732).As a result, Irom a chronological
viewpoint, the stele can be placed between the Tunyukuk inscription and K|l Tigin
inscription, placing it in the periods Irom 726 to 732.This assumption can be sup-
ported by the contents oI this stele reIerring to campaigns in the reign oI Bilg Qaan
(716734). From the palaeographical viewpoint, the verbal suIIix -mi consisted oI
the two letters oImsinT,notoIm as well as KT, BK. At present, in my view the
Ongistelecanbedatedto725732.Theseven-lineinscriptionoIthisheadstonecan
beregardedasasupplementaryinscriptiononthisstele.Thereisalsoasupplementto
the side oI the tortoise stone oI K|l Tigin`s stele and on the upper side oI the western
198 Takashi sawa
(Chinese) side oI the same stele. On the western and eastern sides oI the headstone
oI this stele, there are a tama design consisting oI a ram and a snake. The Iigures oI
dragonsoIthisheadstonecanberegardedassimpliIicationsoIthoseontheKlTigin
and Bilg Qaan steles, which can be evidently conIirmed in the rubbings oI Ramstedt
and Jadrincev. This stele was also built on the tortoise stone like KT and BK. Dragon
ornamentsoItheheadstoneandatortoisestonecouldbeusedIorthemosthigh-rank-
ing aristocratic people oI the Turkic Qaanate (sawa 1999a: 128; 2007a: 2930).
This symbol, however, has Chinese cultural inIluences, as the traditional stele oI the
Tang Dynasty was built Ior high-ranking oIIicers or high-ranking warriors oI the Tang
emperors, never Ior the emperor himselI. Regarding this diIIerence, we can say that
Turkic peoples created the new style oI inscription Ior their Qaans and the most
high-ranking aristocratic rulers oI the Ashinas royal Iamily. As mentioned above,
theseIeaturesindicatedthatthemanburiedatthesitecanbeattestedtohavebeenan
aristocratic individual such as a qaan or qan oI the Ashinas` royal Iamily. This can be
supportedbytheIactthatDuoxifunamed(letmis yabu was called 'qan or 'qaan
onthisstele.
Pe|erences
Aalto, Pentti 1971: Oriental Stuaies in Finlana 18281918. Helsinki: Societas Scientiarum
Fennica.
Alylmaz, C. 2005: Orhun Ya:tlarnn Bugnk Durumu.Ankara:Kurmay.
ATIM RadloII, Wilhelm 18941899: Die alttrkischen Inschriften aer Mongolei. St.
Petersburg:Imp.AkademiiNauka.
Atlas RadloII, Wilhelm 18921899: .
. -: .
Babayar, Gaybulla 2007: Kktrk Kaanl Sikkeleri Katalou.Ankara:TIKA.
Bajar 2004 , 2004: -
. 4,20:7384.
Battulga 2005 2005: (Corpus
Scriptorum Tomus I). : r .
Bazin,Louis1991:Les systmes chronologiques aans le monae turc ancien.Budapest:Aka-
dmiaiKiad.
Bazylxan 1968 , 1968: .
64:9697.
Bazylxan 2005 , . 2005: ( )
. II.
: - . 5258.
Berta,rpd2004:S:avaimat Jol +allfatok, A trk es ufgur rovasirasos emleekek kritikai
kiaaasa.Szeged:JatePress.
Bichees sawa, Takashi & Suzuki, Kousetsu & M|nkhtulga, Rinchinkhorol 2009:
II, 2006
. : SoIex Co. Press.
BKBilg 4aan inscription.
Bold 2006 , 2006: II. :
Press.
CFYG Wang, Qinruo & Yang, Yi (eds) 1960: Cefuyuangui.Beijing:ZhonhuaPress.
Clauson, Gerald 1957: The Ongin Inscription. Journal of the royal Asiatic Society 3, 4:
175192.
Dobrovits, Mihaly 2000: Ongin Yaztn Tahlile bir deneme. Trk Dili Arastrmalar Yll
Belleten2000:147150.
DT , & , & ,
& , A 1969: -
. : .
E EniseiInscription.
ED Clauson, Gerard 1972: An Etymological Dictionary of Pre-Thirteenth-Century Turk-
ish. OxIord: OxIord Uni. Press.
Erdal, Marcel 1991: OlaTurkicWoraFormation,2vols.Wiesbaden.
Erdal, Marcel 2004: GrammarofOlaTurkic.Leiden/London:Brill.
Erdal, Marcel 2010: Ongin Yazt. hlk delik avk (ed.) 2010: III. Uluslararas Trkiyat
Aratrmalar Sempo:yumu 2629 Mays Bilairiler Kitab. Ankara: Reserching Insti-
tuteIorTurkishStudiesoIHaccetepeUniversity.363372.
Gabain, Annemarie von 1953: Inhalt und magische Bedeutung der alttrkischen InschriIten.
Anthropos48:537556.
Giraud, Ren 1961: LinscriptionaeBanTsokto.Paris:Adrien-Maisonneuve.
GOT Tekin, Talat 1968: AGrammarofOrkhonTurkic.Bloomington:IndianaUniversity
Press.
GSR Karlgren, Bernhard 1972: Grammata Serica Recensa. Stockholm: Museum oI Far
Antiquities.
Haln,Harry1978:HanabookofOrientalCollectionsinFinlana.ScandinavianInstituteoI
AsianStudiesMonographSeries31.LondonMalm:CurzonPress.
Haln,Harry1998:BiliktuBakshi.TheknowleageableTeacherG.J.RamsteatsCareeras
aScholar.MmoiresdelaSocitFinno-Ougrienne229.Helsinki:TheFinno-Ugrian
Society.
Hamilton, James Russel 1988: Les Ouwghours a lepoque aes Cinq Dynasties aaprs les
aocumentschinoise.Paris:CollgedeFrance,Institutdeshautestudeschinoises.
Haneda, Tru 1957: Tou ko sanjyssei kaghan kijyo ana shi no boshi. Tyshi Kenkykai
(ed.) 1957: +aneaa +akase Shigaku Ronbunnsh, 2nd Vol, Kyto: Kyto University.
365384.
Iwami, Kiyohiro 1998: T no +opp monaai to Kokusai chittsufo. Tokyo: Kyko Shoin.
Iwasa Seiichir 1936: WADA, S. (ed.) Iwasa Seiichiro Iksh.Tokyo:IWASADenichi.
Jadrincev 1901 , 1901: -
1891 . -
V. -. 154.
Jisl, Lumir 1963: Kl-Tegin Antnda 1958`de yaplan Arkeoloji Aratrmalarnn Sonuolar,
Belleten27:387410.
JTSLiu,Xu&Chang,Zhaoyuan&Wang,Shen(eds)1975:JiuTangshu.Beijing:Zhong-
huaPress.
Kc Kli Cor inscription(IkhuKhoshootinscription).
Klemenc 1895 , 1895: -
1894 . -: . . .
Kljatornyj 1964 , 1964:
. : .
200 Takashi sawa
Kljatornyj 1971 , 1971: -
. L. Ligeti (ed.), StuaiaTurcica.Budapest:AkadmiaiKiad.249258.
Kljatornyj 1975 , 1975: -
. 1. , -, -, .
1975:151159.
Kljatornyj 2003 , 2003:
. -: .
Kormuin 1997 , 1997: -
, . : .
Kormuin 2008 , 2008: -
, . : .
Kozlov 1949 , 1949: 19231926 .
: . -. . -.
KT-E KlTigininscription, East Side.
Li, Yong-Sng 2004: Trk Dillerinae Sontaklar. stanbul: Smurg.
Malov 1959 , 1959: -
. : .
Marquart,JoseI1898:DieChronologieaerAlttrkischenInscriften.Leipzig:Dieterichsche
verlags.
MNTSD - (ed.) 1999:
. : -
.
Mori,Masao1967:Koaaitorukomin:okushinokenkyuuI. Tokyo: Yamakawa shuppansha.
Mori,Masao1992:Koaaitorukomin:okushinokenkyuuII. Tokyo: Yamakawa shuppansha.
Moriyasu, Takao 1977: Chibetto go shiry ch ni arawareru Hoppminzoku- Dru-gu to Hor-.
Afia Ahurika gengo bunka kenky14:148.
Moriyasu, Takao & Ochir, Ayudai 1999: Provisional Report oI Researches on Historical
Sites and Inscriptions in Mongolia Irom 1996 to 1998, Toyonaka: Osaka University,
The Society oI Central Eurasian Studies.
Moriyasu,Takao&Suzuki,Kousetsu&Saitou,Shigeo&Tamura,Takeshi&Bai,Yudong
2009: ine-us Hibun yakuch. StuaiesontheInnerAsianLanguages24:192.
Moriyasu,Takao&Yoshida,Yutaka&AkioKatayama1999:Qara-BalgasunHibun.Takao
Moriyasu&AyudaiOchir(eds)1999:ProvisionalReportofResearchesonHistorical
Sites ana Inscriptions in Mongolia from 1996 to 1998. Toyonaka: Osaka University,
The Society oI Central Eurasian Studies. 209224.
MSSP Haln, Harry (ed.) 1982: Memoria Saecularis Sakari Plsi. Auf:eichnungen von
einer forschungsreise nach aer nralichen Mongolei im Jahre 1909. Travaux ethnogra-
phiquesdelaSocitFinno-OugrienneX.Helsinki:TheFinno-UgrianSociety.
Novgorodova 1980 Nowgorodowa, Eleonora 1980: AlteKunstaerMongolei. Leipzig: E. A.
SeemanVerlag.
Novgorodova 1981 , E A1981: -
. -
1977:203218.
Onogawa, Hidemi 1950: Ongin hibun yakuchuu. Tyshi Kenkykai (ed.), Haneaahakase
shfukinen tyshi rons. Kyto: Kyto University. 431451.
Orkun, Hsein Namk 1936: (ski Trk Ya:tlar. stanbul: Devlet Basmevi.
sawa, Takashi 1999a: Ongi Hibun. Takao Moriyasu & Ayudai Ochir (eds), Provisional
Report of Researches on +istorical Sites ana inscriptions in Mongolia from 1996 to
1998. Toyonaka: Osaka University, The Society oI Central Eurasian Studies. 129136.
sawa, Takashi 1999b: Problems on the Ongi inscription oI the Old Turkic period based
on my Iield survey in Mongolia. Cht Isuramu Bunka no shos to gengo kenky,
IK(DA Osamu Sensei gotaikan kinen ronbunsh.Osaka:OsakaUniversityoIForeign
Studies.275298.
sawa, Takashi 1999c: Moolistan`daki Eski Trk Ant ve Yaztlarnn 19961998 Yllar
arasndaki yeni inceleme sonuolar, Japon-Mool ortak oalmasnn |n raporu olarak.
Trk Dili Aratrmalar Yll Belleten2000.Ankara:TDK.277286.
sawa, Takashi 1999d: Tortoise-shaped Socle Irom site oI Mukhar. Takao Moriyasu &
AyudaiOchir(eds):ProvisionalReportofResearchesonHistoricalSitesanaInscrip-
tions in Mongolia from 1996 to 1998.Toyonaka:OsakaUniversity,TheSocietyoICen-
tral Eurasian Studies. 146.
sawa, Takashi 1999e: Qara-Balgasun Inscription II. Takao Moriyasu & Ayudai Ochir
(eds): Provisional Report of Researches on Historical Sites ana Inscriptions in Mon-
golia from 1996 to 1998. Toyonaka: Osaka University, The Society oI Central Eurasian
Studies.143145.
sawa, Takashi 2000: Moolistandaki Eski Trk Ant ve Yaztlar zerine Yeni Aratrmalar
(1), 19961998 Japon-Mool ortak dalmalarnn gn Raporu. Trk Dilleri Aratr-
malar10:191204,235247.
sawa, Takashi 2002: Bat G|ktrk Kaanl`n Ainasl bir kaan`n eceresine ait bir
kaynak. Hasan Celal Gzel & Kemal dioek & Salim Koca (eds), Trkler2.Ankara:
Yeni Trkiye Yaynlar. 7988.
sawa, Takashi 2007a: The archaeological excavation oI the Bilge Kaghan site in recent
years and the political relations between the Old Turkic Kaghanate and the Tang
Dynasty.Shih39:1438.
sawa, Takashi 2007b: Ongi site and inscription as the historical sources on the ancient
Turkish Kaghanate Based on the International joint researches between Mongol and
Japan.Proceeaings of the 46th meeting of the Permanent international Altaistic Con-
ference on the 22na27th June 2003.Ankara:TDK.323338.
sawa, Takashi 2008: New research on historical aspects and interpretation oI the Ongi site
andinscriptions.Chronica78.Szeged:SzegedUniversity.161177.
sawa, Takashi 2009: The Cultural relationship between Old Turkic kingship and Deer
image. Jan Bemmann & Hermann Parzinger & Ernst Pohl & Damdinsren Tseveen-
dorzh(eds),Current Archaeological Research in Mongolia, Jor- una Frhgeschichteli-
cheArchologieRheinnischeFriearich-Wilhelms-UniversittBonn.Bonn:University
BonnPress.401416.
sawa, Takashi 2010: Kh|r-Asgat Hibun Saik. StuaiesontheInnerAsianLanguages25:
173.
sawa, Takashi & Katayama, Akio 1999: Ongi Hibun. Takao Moriyasu & Ayudai Ochir
(eds),ProvisionalReportofResearchesonHistoricalSitesanaInscriptionsinMongo-
lia from 1996 to 1998. Toyonaka: Osaka University, The Society oI Central Eurasian
Studies.126129.
sawa, Takashi & Suzuki, Kousetsu 2008: Mongoru chsa hkoku 2007. Kichi Matsuda
(ed.),ResearchonUrbani:ationanaTrafficinMongoliaBaseaonInnerAsianSour-
ces, News Letter 03.Osaka:OsakaInternationalUniversity.103138.
202 Takashi sawa
OTWF Erdal, Marcel 1991: OlaTurkicWoraFormation.2vols.Wiesbaden:OttoHarras-
sowitz.
Pulleyblank, Edwin George 1991: LexiconofReconstructeaPronunciationinEarlyMiaale
Chinese, Late Miaale Chinese, ana (arly Manaarin.Vancouver:UniversityoIBritish
ColumbiaPress.
Ramstedt, GustaI John 1978: Seven Journeys (astwara 18981912. Tr. Irom the Swedish &
ed. by J. R. Krueger. Bloomington, Indiana: Mongolian Society.
Rintchen, Biamb 1968: Lesaessinspictographiquesetlesinscriptionssurlesrochersetsur
les stles en Mongolie. Corpus Scriptorum Mongolorum Instituti Linguae et Littera-
rum Academiae Scientiarum Reipublicae Populi Mongolici 16, 1, Ulaanbaatar: Mongo-
lianAcademyoISciencePress.
Rui, Chuanming 1998: Gutufuepimingyanfiu.Shanghai:ShanghaigujiPress.
Rybatzki, Volker 1997: Die Tonyuquq-Inschrift. Studia Uralo-Altaica 40. Szeged: Szeged
University.
Rybatzki, Volker 2000: Titles oI Trk and Uigur Rulers in the Old Turkic inscriptions.
CentralAsiaticJournal44,2:205292.
Sawada, Isao 1983a: Ongin hibun ni kannsuru ichiksatsu- Sono setsuritsumokuteki to set-
suritsu nendai wo chshin to shite. Tyshikenky41,4:5273.
Sawada, Isao 1983b: Ongin hibun toomen dai yongyoo no kaishyaku ni tsuite. Masao
Mori(ed.),Nairikuafianishiafianoshakaitobunka. Tokyo: Yamakawa shuppansya.
79110.
Sawada, Isao 1984: Ongin hibun yakkai. ShunaaiShigaku41:94110.
Shinekh, Maamdai 1979: .
13,8:669.
SSWei,Zheng&Changsun,Wuji&Yan,Shigu&Kong,Yingda(eds)1973:Suishu.Bei-
jing:ZhonghuaPress.
Suzuki,Kousetsu2005:OnthegenealogicallineoItheTurkicAshinaSimo.Ty Gakuh
87,1:3768.
TTunyukukinscription.
TAM TurIan Astna Monjo (Old Chinese manuscripts unearthed Irom the Astna tombs in
theTurIanBasin).
TAM60: 23/1, 23/2 The Iirst & second
parts oI the 23rd Iragments oI the 60th TurIan

Astna manuscript.
TDDo,You(ed.)1988:TongDian,5vols.Beijing:ZhonghuaPress.
TDKTrkDilKurumu(TurkicLanguageInstitution).
Tekin,Talt1988:Orhon Ya:tlar.Ankara:TurkishHistoricalAssociation.
Tekin,Talt1994:Tunyukuk Ya:t.Ankara:Simurg.
Tekin,Talt2003:Orhon Trkoesi Grameri. stanbul: Simurg.
TFCW The ancient Philological Research Institute oI Chinese & The Philological rese-
archgroupoItheNationalCulturalAntiquityManagementBureau&XinjiangUyghur
Autonomous region Museum, School oI the Historical research oI Wuhan University
(eds)19811991:Tulufanchutuwenshu.DocumentsunearthedIromtheTurIanBasin
Vol. 110. Beijing: Wenwu Press.
TFCWFInstituteIortheChineseCulturalrelics&XinjiangUyghurautonomousMuseum
&TheSchooloIthehistoricalresearchoIWuhanUniversity(eds)19911997:Tulufan
chutu wenshu. Documents unearthed Irom the TurIan Basin Facsimiles compared
with transcription Vol. 14. Beijing: Wenwu Press.
TIKATurkishInternationalCooperationAdministration.
TIKA 2003 Turkish International Cooperation Administration (ed.) 2002: Work for the
Profect Turkish Monuments in Mongolia in year 2001.Ankara:TIKAPress.
TIKA2005TurkicInternationalCooperationAssociation(ed.)2005:WorkfortheProfect
Turkish Monuments in Mongolia in year 2003.Ankara:TIKAPress.
Tryjarski, Edward 1966: Die heutige Mongolei und ihre alten Denkmler, The Present State
oI Preservation oI Old Turkic Relics in Mongolia and the Need Ior their Conservation.
Ural-altaischeJahrbcher38:154173.
Tryjarski, Edward 1971: On the Archaeological Traces oI Old Turks in Mongolia. Eastana
West,N.S.21,1,2:121135.
Tryjarski, Edward 1972: Some Remarks on the Monument oI the Orkhon Turks. Milli
Etim Bakanl (ed.), Trk Kltr (l Kitab. stanbul: Milli Etim Basmevi. 3543,
418422.
Tryjarski, Edward 1974: =ur neueren Geschichte des Ongin-Denkmals. Gy|rgy Hazai &
Peter Zieme (eds), In Sprache, Geschichte una Kultur aer Altaischen Jolker. Berlin:
Akademie-Verlag.629630.
Tryjarski, Edward 1991: Zwyc:afe pogr:ebowe luaow rureckich na tle ich wier:e. War-
szawa: Wydawnictwo Naukowe PWN.
Tryjarski, Edward & Aalto, Pentti 1973: TwoOlaTurkicMonumentsofMongolia.Mmoires
delaSocitFinno-Ougrienne150.Helsinki:TheFinno-UgrianSociety.413420.
TTKTrkTarihKurumu.
Tujue:ThenameoIoldTurkicclan,tribeandcountryunderruleoItheroyalIamily
oItheAshinas.
Vojtov 1989 , 1989: ,
. 1989,3:3450.
Vojtov 1996 , 1996:
- JIJIII . : -
.
Vojtov & Bajar 1989 , & , 1989: -
. .
16:5967.
Xaraubaj 2003 , 2003: . : Kul
TEGIN.
Xaraubaj & Myrzataj 2006 , & ,
2006: . : Kul TEGIN.
XTSOuyang,Xiu&Song,Qi&Song,Minqiu(eds)1975:XinTangshu.Beijing:Zhonghua
Press.
YY== Duanchengshi (?863) 1985: Youyang:a:u. Chongguo shixue congshu 35, 2nd vol.
Taibei:Xueshengshuju.
ZLBS Zhaoling Museum & Chang, Pei (eds) 1993: Zhaoling beishi. Xianyang: Sanqin
Press.
ZSLinghu,DeIen&Fang,Xuanling&Cen,Wenben(eds)1971:Zhou-shu.Beijing:Zhon-
ghuaPress.
==TJ Sima, Guang (ed.) 1956: Zi:hitongfian.Beijing:ChonghuaPress.
Takashi sawa osawatak#world-lang.osaka-u.ac.jp~

Osawa

Uploaded by

Copyright:

Available Formats

You might also like

Osawa

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Osawa

Uploaded by

Copyright:

Available Formats

SUSA/JSFOu93,2011

parts oI the 23rd Iragments oI the 60th TurIan

You might also like