G eeks, I anians, and Kušān in Tokha is ān. Bac ian
Iden i ies and he Kušān Empi e
S e an Hä el
ABSTRACT
The Bac ian epig aphic and numisma ic ma e ial om he Kušān pe iod, despi e i s sca ce, eclec ic and
sca e ed na u e, con ains aweal h o in o ma ion abou cul u al and e hnic iden i ies in Bac ia a he ime.
The Kušān in e ac ed wi h I anian and G eek legacies while adding hei own he i age o he mix. Kušān
insc ip ions and seals documen he p esence o indi iduals wi h names om all h ee sphe es in he Kušān
impe ial hie a chy, while he numisma ic pan heon is assembled om dei ies om hese a ious eligious
backg ounds. The e a e se e al indica o s showing ha he Kušān possessed ac i e awa eness o his cul u al
he e ogenei y and used i o shape ade ini ion o Bac ia o hei own pu poses. This pape examines his
e idence and concludes ha he Kušān saw he I anian and G eek elemen s as equally indigenous o Bac ia
and ha hey did no employ adelibe a ely an i -G eek policy.
KEYWORDS
Kushan, Bac ia, éli es, language, adminis a ion, iden i y.
INTRODUCTION
The G aeco -Bac ian pe iod ended when Bac ia was in aded in he mid-2nd cen u y BCE by
nomads om Cen al Asia known in Chinese sou ces as he Yuezhi. Almos wo cen u ies
passed be o e one o hese Yuezhi clans, he Kušān, used Bac ia as abase o o m hei own
empi e.1 Bac ia became he hea land o he Kušān Empi e, which con olled Bac ia and
Gāndhā a, empo a ily ex ending deep in o he Indian Subcon inen . As he new ule s o
Bac ia, he Kušān encoun e ed alegacy o med by he I anian na i es o he land, and he
cul u e o G eek o e lo ds who had con olled he e i o y since he campaigns o Alexande
and he Seleukids. The Kušān added hei own he i age o he mix, and his aises he ques ion
o how hese e hnic and cul u al iden i ies in e ac ed in Kušān Bac ia and how hey wen on
o shape he empi e as awhole. The ollowing in es iga ion ocuses on he epig aphic ma e ial
om Kušān Bac ia o y and ind answe s o his ques ion, wi h ull awa eness o he ac ha
due o he agmen a y na u e o Kušān sou ces, we migh ne e be able o see he ull pic u e.
The i em which a i s glance seems o con ain he mos in o ma ion on e hnic ela ionships
in Kušān Bac ia and has done he mos o g ab he a en ion o schola s is he Raba ak In-
sc ip ion.2 He e in pa icula , one s a emen seems o be ins uc i e (lines 3–4):
οτηια ι ιωναγγο οασο οζοαστο ταδηια αριαο ωσταδο
1 On he a ying opinions on he geog aphy o Yuezhi and ea ly Kušān Bac ia, c . G ene 2006;
Hill 2015/I, 329–361; Falk 2018 and now Sinisi 2022.
2 The Raba ak Insc ip ion is ci ed ollowing Sims -Williams 2008.
STUDIA HERCYNIA XXVIII/2, 108–121
109STEFAN HÄRTEL
Li e ally, he ph ase ansla es as ‘and he [Kaniška I] led ou he G eek speech and placed
A yan’. As αριαο is an adjec i e,3 i will e e o he e m οασο, hus p o iding he meaning
‘and placed (i in) A yan’ o ‘placed (an) A yan (speech)’. Some schola s ha e unde s ood his
o mean ha Kaniška Iabolished he o icial unc ion o he G eek language in Bac ia and
eplaced i wi h ‘A yan’, agene ic name o I anian languages and iden i ies.4 This in e p e-
a ion was pa icula ly a ac i e because i seemed o ie in wi h he swi ch o he language
o Kušān coin legends om G eek o Bac ian ha can be obse ed ea ly in Kaniška’s eign,5
and mo e gene ally wi h G. Fussman’sp oposal ha he Kušān ac i ely ejec ed he G eek
cul u al legacy in a ou o an indigenous I anian one.6 Howe e , an e ymological and syn-
ac ical analysis o his s a emen sugges s ha i should much a he be aken o mean ‘and
he made a[= he ollowing] p oclama ion in G eek and ansla ed i o Bac ian’, ameaning
ha was al eady sugges ed by N. Sims -Williams.7
In b ie ,8 he mos ele an wo d is οασο, de i ed om Old I anian *wāč ‘wo d, speech’.9
I s seman ic meaning can be ga he ed om i s cogna es A es an ačah- ‘speech, exclama ion’,10
Middle Pe sian/Pa hian wʾc ‘wo d, speech’
11
and Sogdian p wʾʾc ‘denig a e, slande ’.
12
The
wo d is ne e used o e e o as uc u ed language, bu a he o acon en ha is being spo-
ken, ap oclama ion.13 This p oclama ion is hen ci ed o pa aph ased in he ollowing pa
o he insc ip ion and e e s o he conques o India by he Kušān.
The e a e wo conc e e pieces o e idence o sugges ha wha ollows he ph ase οτηια
ι ιωναγγο οασο οζοαστο ταδηια αριαο ωσταδο is a leas apa aph ase o ap e -exis ing ex .
Fi s , he ollowing sen ence is no in oduced by aconnec ing pa icle, as would be expec ed
in auni ied ex composi ion, bu ea u es he wo d φροαγδαζο, apas pa iciple φροαγδ- con-
nec ed wi h a e b αζο, which Du kin -Meis e e ns demons a ed o be an exp ession o he
impe ec , he only ime his occu s in he ex an Bac ian co pus.14 This sugges s ha his
segmen o he insc ip ion is sou ced om a ex wi h ahighe deg ee o li e a y s ylism, as
3 The su ix -αο appea s o be iden ical o Sogdian -ău, Kho anese -au and Pa hian -āw, c . Sims-
-Williams– C ibb 1995-1996, 90; Ge she i ch 1961, 164 (§ 1076) and Du kin -Meis e e ns 2014,
163 (§ 314); in Pa hian in pa icula , he su ix is used o deno e languages.
4 This was i s exp essed by J. C ibb apud Sims -Williams– C ibb 1995-1996, 110–111, whe e
C ibb’sunde s anding o he sen ence con adic s ha o N. Sims -Williams. C ibb’sin e p e a ion
was ollowed by Mukhe jee 1998, 6 and Gnoli 2009, 153 among o he s. Fussman 1998, 596–598
sugges s unde s anding he s a emen as pe aining o some hing G eek, hen es ablishing i in
Bac ian. No ably, Panaino 2009, 339, n. 59 (and Panaino 2015, 96) suppo s he in e p e a ion
gi en by N. Sims -Williams in Sims -Williams– C ibb 1995-1996, 82–83 and ha is also ollowed
he e.
5 Discussed in Göbl 1960; c . also Fussman 1998, 596.
6 c . Fussman 1977.
7 Sims -Williams– C ibb 1995-1996, 82–83.
8 This is discussed in g ea e de ail in Hä el 2024b, chap e 4.2.
9 Sims -Williams– C ibb 1995-1996, 93.
10 Ba holomae 1904, 1340.
11 Du kin -Meis e e ns 2004, 333. No e also he Pa hian e b wʾc-, ibid.
12 Gha ib 1995, 291 (§ 7258).
13 This is unde sco ed by he only occu ence o he wo d in which i does e e o alanguage, namely
in he Middle Pe sian Šābuh agān, in which wʾc is e e ed o he ‘speech’ o he ‘Mazan misb eeds’
(c . Hu e 1992, 88, 92–93), and he con ex sugges s ha wha e e i is hey speak, i is no alan-
guage in he s uc u ed sense.
14 Du kin -Meis e e ns 2007, 43–44.
110 STUDIA HERCYNIA XXVIII/2
would be expec ed o he p oclama ion o an empe o .
15
Second, ano he insc ip ion published
by Sims -Williams 2019 also appea s o ci e his p oclama ion, as some pa s o he ex echo
Raba ak a he closely and e e o he same Indian campaign.
Wha a e he u he implica ions o his sen ence? I shows ha he Kušān we e conscious
o wo linguis ic iden i ies in Bac ia, namely G eek (ιωναγγο = Ionian) and I anian (αριαο
= A yan). A hi d iden i y, υνδοοαο ‘Indian’, is la e e e ed o in he gloss be ween lines 9
and 10, and one o wo di ine names a e ansla ed acco dingly, as will be discussed u he
below.16 While his should ha dly be su p ising o pos -Hellenis ic Bac ia, i may be wo h
examining he o he e idence o hese wo g oups in hope o each an app oxima ion o how
exac ly he Kušān unde s ood hese g oups and wha hei posi ion was in Bac ia.
THE PREVALENCE OF GREEK IN KUŠĀN BACTRIA
The ea ly Kušān coin legends may help o ge a i s glimpse a he s a us o he G eek lan-
guage and i s speake s. Since he Hellenis ic pe iod, coins issued in Bac ia had monolingual
G eek legends. Coins issued by Indo -G eek so e eigns wi h bilingual G eek and Gāndhā ī
legends seem o always ha e been min ed sou h o he Hindukush.17 The ea ly Yuezhi ule s o
Bac ia issued G eek imi a ion coins, bu once o iginal coin designs eappea ed, he legends
also became dynamic in con en again, and hey emain monolingual in G eek. The Kušān
in oduced i les on hei coins ha a e inno a i e o Bac ia,18 such as βασιλεύς βασιλέων
and σωτήρ μέγας, and e en τυραννοῦντος (gen. sg. o τυραννεύω), possibly in ended as he
G eek equi alen o he i le yabghu, as a i le o non -mona chic bu ne e heless absolu e
powe .19 Meanwhile, in Gāndhā a he G eek legends degene a ed unde he Indo -Scy hians
and Indo -Pa hians while hose in Gāndhā ī emained dynamic.
The Kušān we e appa en ly also no in e es ed in adop ing he Gāndhā ī adi ions o ei he
g oups in hei coin issues, as Kujula Kadphises on his bilingual issues ansla es he G eek
i le δίκαιος wi h sachadh ama hida o jus dh ama ida a he han adop ing he es ablished
15 Sims -Williams 1998, 84 sugges ed ha he expec ed ταδο migh be omi ed by haplog aphy om
he p eceding ωσταδο, bu also admi s ha he expec ed o m would be ταδι.
16 Panaino 2015 a gues ha αριαο e lec s he o iginal sense o ‘A yan’ and hus encompasses bo h
I anian and Indian iden i ies (Panaino 2015, 96–97), and ha he use o υνδοοαο he e is eligious.
I ind his a gumen di icul in he ligh o he e idence p o ided by Sims -Williams (also e e ed
o by Panaino) ha i is o med on he same basis as αριαο; he p oblem o dialec ology add essed
by Panaino does no need o ha e posed i sel o he Kušān, who will ha e likely called ‘Indian’ any
language spoken by hose people hey de ined as ‘Indian’. Any dis inc ion be ween he indi idual
Indian languages o dialec s (inso a as hey we e pe cei ed by Kušān impe ial au ho i ies in he
i s place) would ha e been i ele an o he Raba ak con ex .
17 Bopea achchi 1991, 57–59.
18 Al hough βασιλεύς βασιλέων was commonly used by he A sakids, Indo -Scy hians, and Indo-
-Pa hians, i was ne e used by he G aeco -Bac ians o Indo -G eeks ( o alis o i les used by
he la e c . Bopea achchi 1991, 387–389). Σωτήρ μέγας appea s o be agenuine Kušān ex ension
o he common σωτήρ i le. Some coins o Gondopha es -Sases con ain he sequence Maha ajasa
Maha asa T a a asa on hei legends (examples in Senio 2001/II, 168–171), bu maha a also appea s
on i s own, e.g. on coins o Gondopha es (e.g. Senio 2001/II, 158, no. 222.4–6).
19 Fo an al e na e explana ion c . Falk 2018, 5. Howe e , Falk does no conside ha he use o
βασιλεύοντος e e ed o he e is pa o he A sakid da ing o mula o he A oman documen s,
i.e., no meaning ‘ac ing as king’ bu ‘in he eign o he king o kings’, c . Hä el 2024b, chap e
5.1. On he use o i les by Kujula Kadphises c . C ibb 2018.
111STEFAN HÄRTEL
ansla ion dh amika in use since he Indo -G eek pe iod.20 The same i le was la e ansla -
ed in o Bac ian using he wo wo ds ραþτογο and λαδειγο which encompass he mo alis ic
and legalis ic meanings o he wo d ‘jus ’ as ep esen ed by δίκαιος, hus e lec ing aca e ul
ansla ion p ocess.21 All his indica es ha he e was s ill awo king knowledge o G eek in
Bac ia a he disposal o he ea ly Kušān.
The e is u he e idence o his. A agmen a y insc ip ion ound in he Kušān dynas-
ic sanc ua y o Su kh Ko al in he Bac ian hea land con ains aG eek colophon ollowing
aBac ian ex : διὰ Παλαμήδου.22 Since he sanc ua y was ounded unde Kaniška Iand he
Bac ian insc ip ion p eceding i ma ches pe ec ly wi h he epig aphic e idence om he
Kušān pe iod, he mos easonable explana ion is ha his is acolophon w i en in G eek by
aG eek sc ibe named Palamedes. Colophons a e common in Kušān insc ip ions, bu his is
he only p ese ed example o one in G eek. Howe e , he ending o he Raba ak insc ip ion
(line 23), hough la gely illegible, con ains he pe haps incomple e wo d ατιδνοσ, which may
be aG eek name, and hus also indica e aG eek sc ibe.23 I so, howe e , he name is no in
geni i e case as ha o Palamedes, and he colophon would he e o e p obably be in Bac ian,
no in G eek.
The de elopmen o he coinage unde Vima Kadphises and Kaniška Ip o ides u he
e idence. Vima Kadphises issued bilingual coins wi h G eek and Gāndhā ī legends, al hough
by his ime, Bac ian had al eady been educed o w i ing as he insc ip ion o Daš -iNawu
shows. He uses bo h he i les βασιλεύς and βασιλεύς βασιλέων. The sc ip used in hese legends
is o he mos pa he ypical a ian o G eek in use in Bac ian epig aphy since he la e
G aeco -Bac ian pe iod as indica ed by he moulds om he Oxus Temple o Takh -iSangin.24
This sc ip employs cu si e o ms o le e s such as Alpha, Kappa, Sigma, and Omega, and
an unusual o m o E a. Unde Kaniška Ihowe e , he e is a lee ing hough cu ious change.
Kaniška d ops he Gāndhā ī legends on he e e se in a ou o G eek legends naming he
dei y depic ed he e. Gāndhā ī legends ne e eappea on Kušān coins, despi e he ac ha
Gāndhā ī epig aphy lou ishes. In he e y la e pe iod, he Kušān in oduce B ahmī legends
on hei coins.
The sc ip used by Kaniška I o his G eek coin legends and al eady employed by Vima
Kadphises on some o his gold issues, is much clea e han he no mally used G eek a ian
and con ains apa icula ly in e es ing a ian o he le e Alpha wi h he ba poin ing down,
which does no occu in any ex an epig aphic ma e ial om Bac ia, bu was common in
wes e n Asia in he Seleukid pe iod (Fig. 1). I disappea ed unde he A sakids wi h asudden,
bu lee ing eoccu ence on coins o Go a zes II (40–51 CE) om he Seleukeia min (Sell-
wood ype 66, Fig. 3), who also e i ed he dynamic G eek legends on A sakid coins ha had
become s agnan and inc easingly degene a ed since ca. 80 BCE. Kaniška hen changed he
coin legends o Bac ian, d opping his a ian o he G eek sc ip and e u ning o amuch
mo e cu si e a ian o w i e Bac ian (Fig. 2). In some cases, he exis ing G eek dies we e
ecu o ea u e Bac ian legends,25 bu e en on hose he s yle o he sc ip is changed. The
names o he dei ies we e ansla ed oo, om G eek o Bac ian, al hough he selec ion o
20 So no ed in Jongewa d– C ibb 2015, 23.
21 La e , his pai was used in o acompound lʾd š as ound in he Manichaean Bac ian agmen ,
c . Sims -Williams 2009, 257.
22 Henning 1956, c . on he G eek line also Rougemon 2012, 194–195, no 94.
23 Sims -Williams 1998, 88.
24 Fo ade ailed s udy including epig aphic issues c . I an chik 2009, esp. 125–130 and 134–137.
25 Göbl 1960, 94–95.
112 STUDIA HERCYNIA XXVIII/2
Fig. 1: Ob e se o acoin o he ea ly ‘G eek’
issue o Kaniška Iwi h G eek legend eading
ΒΑCΙΛΕΥC ΒΑCΙΛΕWΝ ΚΑΝΗÞΚΟΥ, poin ed-
-ba Alphas isible in he wo ds ΒΑCΙΛΕWΝ
and ΚΑΝΗÞΚΟΥ. Be lin, Münzkabine de
S aa lichen Museen, 18217594. Pho og aph
by Reinha d Saczewski.
Fig 2: Ob e se o acoin o Kaniška Iwi h
Bac ian legend eading ÞΑΟΝΑΝΟÞΑΟ
ΚΑΝΗÞΚΟ ΚΟÞΑΝΟ. Be lin, Münzkabine
de S aa lichen Museen, 18203703. Pho o-
g aph by Reinha d Saczewski.
Fig. 3: Re e se o acoin o he A sakid Go a zes II om Seleukeia, wi h se e al poin ed -ba Alphas.
Be lin, Münzkabine de S aa lichen Museen, 18202782. Pho og aph by Lu z -Jü gen Lübke
(Lübke und Wiedemann).
his s ill limi ed pan heon should be seen p ima ily as e lec ing impe ial Kušān ideas a he
han e e ing o any pa icula ac i e cul s.
A u he peculia i y is he use o he ‘monumen al’ Sigma (Σ) o he numbe 200 a he han
he ‘cu si e’ o ‘luna ’ a ian (C) ound in all o he ins ances o he le e in Bac ian epig aphy.
This can be obse ed bo h in he Daš -iNawu insc ip ion and in la e Bac ian documen s.
26
Since
his occu s as ea ly as Daš -iNawu ,27 i demons a es ha he Kušān sc ibes had bo h a ian s
o he le e a hei disposal and made he choice o use he ‘cu si e’ a ian o w i e Bac ian.
26 Fo occu ences c . Sims -Williams 2007, 181.
27 P e iously una ailable pho os and new acings o he insc ip ion oge he wi h anew eading and
ansla ion can now be ound in Hal manne al. 2024.

113STEFAN HÄRTEL
A inal elemen ha shows he ac i e knowledge o G eek in Bac ia in his ime is he
use o he G eek sc ip i sel .
28
I s adap a ion was by no means mechanical, bu i used es-
ablished o hog aphic con en ions such as he spelling o anasal /ŋ/ using wo Gammas.
The cha ac e is ic ea u e o a inal Omik on can al eady be obse ed on ea ly Kušān coins,
whe e some imes, he G eek geni i e ending o Omik on -Ypsilon seems o be educed o an
Omik on.29 This is no o mean ha he Bac ian inal Omik on is de i ed om aG eek geni i e
ending, bu i may indica e a endency owa ds his con en ion om he ea lies s ages o
w i ing Bac ian. In any case, he use o he G eek sc ip o w i e he na i e Bac ian language
indica es ha i was pe cei ed by he Kušān as ana i e ‘Bac ian’ sc ip which was a ou ed
o e o he a ailable sc ip s such as Kha oṣṭhī o A amaic, despi e hese being be e sui ed
o w i e an I anian idiom.30
KUŠĀN BACTRIAN
Tu ning o Bac ian now, i is in e es ing ha he Raba ak Insc ip ion uses he e m αριαο,
‘A yan’, o name i . This ecalls he use o his name by he Achaemenids in e e ence o he
Pe sian language in he Biso un Insc ip ion (§ 70).31 In he case o Old Pe sian, his may be
explicable i i is accep ed ha his w i en language was no in ac Pe sian a all, bu an I a-
nian koiné assembled om Pe sian, Median, and o he I anian languages.32 I was hus gi en
he same o e a ching iden i y as he Pe sian king himsel , who de ined himsel i s as an
Achaemenid, hen aPe sian, and hen an A yan.33 This migh be asa is ac o y explana ion
o he many dialec ological con adic ions ound in Achaemenid Old Pe sian. Could his be
he case o Kušān Bac ian as well? The e a e s ong indica ions o as eamlining p ocess
o he language unde Kaniška Iand Hu iška. The g amma ical complexi y is educed. Mos
no ably, in he ime ame o me ely wo decades be ween Raba ak and Su kh Ko al, g am-
ma ical gende appea s o ha e been d opped in many occasions. This appea s o ha e been
aconscious e o , gi en he signi ican a ia ions in wo d endings be ween he disca ded
e sion B o Su kh Ko al and e sion M which was se up o gene al iewing,34 and which
may ha e con used sc ibes wo king on hese epig aphs.
I is mo e di icul o de e mine whe he he language is made up o a ious dialec ological
in luences, and his ques ion mus be pu aside he e. In any case, he language o he Kušān
insc ip ions is la e ound ex ensi ely in he co pus o he Bac ian documen s om no he n
A ghanis an, wi h no eason o doub ha he language was oo ed in his e i o y.35
28 This p ocess o adap a ion is he subjec o achap e in he disse a ion o F. Miche i.
29 Examples o such legends in Senio 2001/II, 220, no B7.1.
30 Impe ial A amaic has now been shown o be he basis o he ‘undeciphe ed sc ip ’ om he Daš -
-iNawu and Almosi Go ge insc ip ions. Falk 2023 sugges s ha he language w i en in his sc ip is
Bac ian, whe eas Bonmanne al. 2023 a gue ha i is ahi he o unknown Middle I anian language.
31 Schmi 2009, 87.
32 So i s sugges ed by Lecoq 1997, 50. Fo ecen li e a u e on his, c . Rossi 2021, 54.
33 So exp essed in he i s insc ip ion o Da ius in Naqš -e Ros am (DNa), § 2, c . Schmi 2009, 101.
34 The e sions can bes be compa ed in Göbl 1965, pl. IV–VI.
35 Henning 1960, 47 al eady de e mined ha , due o i s linguis ic p oximi y o Pash o, Yidgha -Munji,
Sogdian, Khwa ezmian, and Pa hian, i is ‘in i s na u al and igh ul place in Bac ia’, a e dic
con i med by Sims -Williams 1997, 22.
114 STUDIA HERCYNIA XXVIII/2
THE COMPOSITION OF KUŠĀN BACTRIAN ÉLITES
Looking a he uppe hie a chies o he Kušān Empi e speci ically in he ime un il in o he
eign o Hu iška, an in e es ing si ua ion occu s. O he high o icials men ioned in he Bac i-
an insc ip ions o he pe iod (Tab.1), only ew pe sonal names can eliably be e ymologised as
Bac ian. This o cou se p ima ily conce ns he names o he Kušān kings hemsel es, which
a e in an unknown, hough mos likely I anian, language.36 The su ix -šk commonly ound in
he names o Kušān empe o s also appea s in he name Kozgašk, an indi idual o unknown,
hough high, s a us in he insc ip ion o Su kh Ko al. Two o i e men known o us in he
ank o καραλραγγο, ahigh mili a y o ice appa en ly answe ing di ec ly o he empe o , ha e
non -Bac ian names: Piaš and Nukunzuk. The la e is he mos ema kable indi idual in he
Bac ian epig aphic co pus, as he appea s in h ee dis inc insc ip ions which allow us o ace
his ca ee and he scope o his powe o ace ain ex en . His name is unclea , bu dis inc ly
non -Bac ian due o he in e ocalic /k/, which in Bac ian would u n o a oiced and pe haps
ica i e /g/ o /γ/ sound. Ša a , ano he καραλραγγο appea ing in he same insc ip ion as
he a o emen ioned indi iduals, has agene ic I anian name ha may o may no be Bac ian.
Only a ou h man called Na kas appea ing in an insc ip ion om he ime o Vāsude a, can
be eliably in e p e ed as Bac ian.37 This same Vāsude a insc ip ion also names wo o he
indi iduals in high anks wi h Bac ian names: Ram and Humug -āgad, he la e sha ing
his name wi h aκαραλραγγο known om acoin o he ime o Hu iška.38 A u he en i ely
unclea name Aldšo (ΑΛΔÞΟ) appea s on aseal clea ly om he Kušān pe iod showing wha
may be aKušān p ince o oyal igu e iding aho se.39
One may wonde a his poin i he e is any e idence o Indian indi iduals in Bac ia a
his ime, and wha hei s a us was. Insc ip ional e idence is qui e abundan in he con ex s
o Buddhis si es such as Ka a epe, Fayaz epe, and in owns wi h Buddhis communi ies, in-
cluding Dal e zin epe and Dilbe djin, bu hese e e mos ly o Buddhis monks. Apa om
ha , he Raba ak Insc ip ion e e s o he Indian in e p e a ion o he dei ies S oša do and
Na asao as Maaseno and Bizago espec i ely.40 This is apa icula ly in e es ing occu ence,
because i is in an in e linea gloss ha seems o ha e been added a e he insc ip ion was
w i en, and hus may indica e ha Indian elemen s in he popula ion el unde - ep esen ed.
I may also be signi ican ha he wo gods S oša do and Na asao a e he only ones in he
Raba ak pan heon no ep esen ed on Kušān coins.41 Maaseno and Bizago howe e do appea
36 C . Sims -Williams 2003, 237–239; his may be he language o he ecen ly deciphe ed ‘unknown
sc ip ’ as discussed abo e. The indi idual pe sonal names a e discussed in Sims -Williams 2010.
37 Falk– Sims -Williams 2017, 133–134.
38 On all aspec s ela ed o his insc ip ion, c . Falk– Sims -Williams 2017.
39 He e, Fa idnejad 2018, 315 eads he name as ΑΘÞΟ, he name o agod ound on Kušān coins, bu
Icanno ollow his in e p e a ion, as he clea ly isible le e s ΛΔ con a an assumed Θ a e no
accoun ed o . On he seal i sel c . Callie i 1998, 200–201 and 310–311 (Ca U7.29).
40 Icanno ollow Falk 2019, 37 in his sugges ion ha hese wo in e p e a ions e e o Mii o (Mi h a).
Con a y o Falk, he gloss does no begin o e he Sigma o ΝΑΡΑCΑΟ, bu e y clea ly abo e he
Nu o he same name. Falk also igno es he ela ionship be ween S aoša and he Maaseno o Hu-
iška’scoins ia he cocke el a ibu e, on his c . now Hä el 2024a, 7–8.
41 Al hough his may also apply o Omma, i Falk 2019, 30 is co ec in eading he legends o acoin
(Göbl 1984, ype 310) usually ead as ΟΜΜΟ as ΟΗΜΟ = Vima (Kadphises). Howe e , his in e -
p e a ion o he Omma o he Raba ak Insc ip ion as aSemi ic loanwo d umma e e ing o he
ollowing g oup (communi y) o dei ies is di icul o ollow. ASemi ic loanwo d is mos likely o
ha e come o Bac ia in he Achaemenid pe iod, and Achaemenid sou ces a e de oid o bo h he
wo d umma and any e e ence o he idea o adi ine communi y.
115STEFAN HÄRTEL
on Kušān coins in ways ha closely associa e hem wi h each o he .42 The ac ha hese coins
we e issued unde Hu iška does no mean he Kušān did no o icially ecognise he cul o his
g oup o dei ies unde KaniškaI. Some o he specimens ha e he amgha o Kaniška I, sugges -
ing ha he ob e ses may ha e been concei ed sho ly be o e his dea h.43 On he o he hand,
he gloss may only ha e been added in he eign o Hu iška. In any case, he Indian Skanda
cul was impo an enough o he dynas ic eligion o he Kušān o wa an e e ence, and i
is no doub signi ican ha he co esponding dei ies we e iden i ied as Indian. Howe e , no
42 Maaseno appea s bo h alone (Göbl 1984, ypes 227, 298) and as pa o a iad consis ing o himsel ,
Bizago and Skando Koma o (Skanda Kuma a; Göbl 1984, ype 157). Bizago and Skando Koma o also
appea oge he on coins (Göbl 1984, ypes 156, 166), bu ne e alone.
43 So sugges ed in Göbl 1984, 25.
Tab.1: Names o o icials om Kušān Bac ian insc ip ions and seals. The insc ip ions a e e-
e ed o as ollows: Raba ak (Rab): Sims -Williams 2008; Nukunzuk Sil e Pla e (NSP): Sims-
-Williams 2019; Su kh Ko al (SK) 1: Sims -Williams 1973, 95, n. 1; SK 3: Henning 1956; Su kh
Ko al Monoli h (SK4M): Ge she i ch 1979; Ay am (Ay ): Tu guno – Li shi s– R eladze
1981; ‘Coin’ and Vāsude a Sil e Pyxis (VSP 1, 3, 4): Falk– Sims -Williams 2017; Seals acco ding
o Callie i 1998 and Bi a 1968. IPBN e e s o he en ies in Sims -Williams 2010. The names
om VSP a e discussed in Falk– Sims -Williams 2017, 133–37.
Name Rank Name O igin Sou ce IPNB
Ša a o καραλραγγο I anian, Bac ian e ymology possible Rab 7, 15, 16 554
Piašo καραλραγγο Unknown Rab 14 371
Nokonzoko αμβουκαο, αþτοοαλ-
γο, καραλραγγοUnknown, no Bac ian (in e ocalic -k-) Rab 16, NSP 1,
SK4M 6-7 289
]a indos? Sc ibe? I PN G eek? Rab 23 -
Iōlesagōgo Sc ibe? I PN I anian, Bac ian e ymology possible SK1 190
Palamēdēs Sc ibe G eek SK3 -
Kozgaško ? Unknown, su ix -šk also ound in Kušān oyal
names SK4M 21-22 217
Bo zomiy o As ilo-
ganseigo ? I anian, Bac ian e ymology possible SK4M 21 98
Miy amano Sc ibe I anian, e ymology unclea SK4M 24-25 262
Mii ozado Sc ibe I anian, Bac ian e ymology possible Ay 6 250
Šodba ? Unclea , bu possibly I anian Ay 557
Omoiago[ καραλραγγο I anian, Bac ian e ymology possible Coin -
Na kaso καραλραγγο Bac ian (compa e Sogd. cogna e n kc) VSP 1 -
Ramo ωστειγο I anian, Bac ian e ymology possible VSP 3 -
Omoiogoagado ? Bac ian (-αγαδο dis inc ly Bac ian s Sogd. ʾγ ) VSP 4 -
Iōlo lano ? I anian, pe haps Bac ian name o ma ion based
on non-Bac ian o m ορλα(γ)νο o Və ə h aγna Seal: Callie i
Ca U 7.4 195
Xa obalago ? Bac ian (common Bac ian i le kha , in e nal
oicing o -p- om παλαγο) Seal: Callie i
Ca U 7.11 516
Šooago ? I PN I anian Seal: Callie i
Ca U 7.22 559
F eixoadēo ? Bac ian (compa e Sogd. cogna e p yxw ʾw)Seal: Callie i
Ca U 7.23 501
F amano/
Vh amana ? Bac ian (compa e Sogd cogna e p mʾn)Seal: Bi a
Sig 70 499
Aldšo Royal Unknown Seal: Callie i
Ca U 7.29 16
116 STUDIA HERCYNIA XXVIII/2
impe ial o icials wi h Indian names a e known wi h ce ain y om Bac ia. The only possible
excep ion would be he name ÞΟΔΙΑ ead by he o iginal edi o s o he Ay am Insc ip ion.44
Howe e , Sims -Williams ead he name as ÞΟΔΒΑ and p oposed an I anian e ymology.45
ETHNIC IDENTITY AND IMPERIAL SELF ‑REPRESENTATION
G. Fussman a gued ha Kušān poli ics possessed an an i -G eek sen imen .46 Akey poin in his
a gumen is he sanc ua y o Su kh Ko al, which Fussman in e p e ed as adynas ic sh ine,47
and which acco ding o him is nei he ap oduc o G eek in luence, no aconscious e u n
o Achaemenid o ms.48 Ins ead, he Kušān a e iewed as p esen ing hemsel es as a‘dynas ie
indigène’, by which Fussman means I anian.49 The ac ha an o e a ching e m such as ‘A yan’
was used o he local linguis ic iden i y, a he han aspeci ic one analogous o swγδyʾw o
Sogdian,50 appea s o be con i ma ion o his hesis.
Howe e , as he e idence p esen ed abo e shows, he pic u e is a mo e complica ed han
asimple ‘I anian’ eplacemen o ‘G eek’ elemen s. Fi s , he disappea ance o G eek iden i y
is no o al. The Raba ak Insc ip ion i sel is e idence ha G eek was used as an adminis a i e
language a leas ea ly in he eign o Kaniška I, and i is ques ionable i he ansla ion o he
G eek edic o an ‘A yan’ one means ha he e was compe i ion o hese wo languages. No
u he e idence o his so is ound a e Raba ak, bu he p esence o G eek sc ibes, a leas
in he case o he Palamedes Insc ip ion, is indispu able. I is no ewo hy ha he Kušān did
much o dis inguish hese wo iden i ies, bu i canno be posi i ely obse ed ha he e was
adecla ed o delibe a e an i -G eek policy. The be e explana ion o he e en ual disappea -
ance o G eeks om ou sou ces in Bac ia is, in my iew, ha apopula ion de ining i sel as
G eek will al eady ha e been so small and ma ginalised by his ime ha hey should no be
expec ed o be p ominen ly ea u ed in he sca ce sou ces. The li le e e ence ha is ex an ,
howe e , demons a es ha hey we e s ill pa o he impe ial éli e in he ime o KaniškaI. O
cou se he swi ch o coin legends om G eek o Bac ian canno be a gued away. He e, i could
be obse ed ha his was an ac o e lec he eali y as i exis ed. Fussman d ew acau ious
compa ison o he pos -colonial a i udes in o me Eu opean colonies in he second hal o
he wen ie h cen u y, which a emp ed o ind a‘na ional iden i y’ and hus shed he G aeco-
-Bac ian pas .51 I ind his o be exagge a ed no only in ega d o he e iden pe sis ence o
G eek elemen s in éli e con ex s discussed abo e, bu also in some o he clea e e ences o
G eek eligion and eligious iconog aphy on impe ial Kušān coins. He e, we encoun e He -
aklēs, Sa apis, and Dionysos,52 and dei ies such as A doxšo, Šao eo o, and A hšo, who, despi e
44 Tu guno – Li shi s– R eladze 1981, 41, ollowed by Ha ma a 1986.
45 Sims -Williams 2010, 157 (§ 557).
46 Fussman 1977, 322.
47 Fussman 1977, 320.
48 Fussman 1977, 321; he la e a gumen in pa icula is a ejec ion o compa isons o Susa made by
D. Schlumbe ge , c . Fussman 1977, 319.
49 Fussman 1977, 322.
50 Sims -Williams– Du kin -Meis e e ns 2012, 179.
51 Fussman 1977, 322.
52 Dionysos (Bac ian Deineiso) was i s desc ibed in Jongewa d– C ibb 2015, 291. C . Jongewa d–
C ibb 2015, 259–301 o acomple e and illus a ed p esen a ion o all dei ies ound on Kušān coins
(now also Hä el 2024, 210–292).