VIHĀRAS IN THE KUSHAN EMPIRE

Christopher I. Beckwith

The traditional image of “Early Buddhism” is characterized by a striking individuality and the lack of
a social consciousness. Among the earliest attested practitioners of the late fourth century BC, the “forest
śramaṇas” were solitary, but the “town śramaṇas” were involved in helping people, and stayed overnight
in patrons’ houses. Nevertheless, there was no Buddhist community like the later Saṃgha, and śramaṇas
were not monks. It was still un-organized non-monastic Buddhism; monasteries and monks (bhikṣus) did
not yet exist (Beckwith 2015).
Then, in the first to second centuries CE, the developed Buddhist monastery appears in Kushan north-
western India as well as in North China, a culture just outside the sphere of influence of the Kushan
Empire, accompanied by written texts and historical references. It is “Normative Buddhism” – fully
developed organized later Buddhism, with the triratna, monks, devotion to Buddhas and Bodhisattvas,
karma, rebirth, etc.: more or less the entire tradition known to us, except for the Vinayas, all of which
have been shown to date to the fifth century CE (Schopen 2004:94).
The central institution of Normative Buddhist culture is the Buddhist monastery, or vihāra. The
importance of the monastery for a society dominated by Buddhism is so crucial that we must spend a few
minutes on what exactly the early monastery was, and what it was called. The early Buddhist monastery
first appears as a completely developed complex thing, with many special features:
¨ Its most important building, the raison d’être of the monastery as a whole, has a highly distinctive
architectual design known as the “cloister plan”. In Buddhist archaeological contexts it is known as the
“vihāra plan”. It was designed to house a medium-sized group of regimented monks in identical, tiny
cells – the opposite of Early Buddhist individuality. The standard plan consists of a rectangular structure
with a heavy exterior wall and one entrance. From the outside it is easily mistaken for a fort. The interior
is dominated by an open rectangular courtyard (in some cases occupied largely by a rectangular pool)
surrounded by a colonnaded or arcaded walkway or “veranda”. The entire structure between the veranda
and the exterior wall, all the way around except for the lone entryway, consists of more or less identical
cells, each housing one monk. A kitchen and other functional rooms are also attached to the cloister.
Evidently because of the central importance of the cloister building, the vihāra proper, the monastery
complex as a whole early came to be called a vihāra in many languages.
¨ It was a particular kind of legal entity – an endowed, tax-free institution with inheritable
ownership (Schopen 2004: 219-259; Van Bladel 2010).
¨ It was the locus for religious veneration of Buddhist relics in an associated stūpa (a Buddhist
caitya or ‘reliquary’) that sometimes occupied part of the vihāra courtyard.
¨ The vihāra was also the first college, the direct ancestor – via the medieval Islamic madrasa1 – of
our modern university, as recently shown in detail.2
The monastery was thus not a “generic” institution or architectural form. The residential quarter of all
early Buddhist monasteries found so far follows the same unique plan, the “cloister plan” or vihāra plan,
which is highly distinctive and can hardly be confused with anything else. Its later Central Asian form,
the “four-īwān” plan, dating from about the fifth or sixth century on, is best known from the site of

1 Already Barthold (1964: 30, originally published in 1918) noted that the madrasa was simply an Islamicized vihāra.
Archaeology done since his time has proven him to be right; cf. the following note.
2 The early monastery/college was the locus of the Buddhist invention of the early literary “scientific method”, the recursive
argument method sometimes referred to by its Medieval Latin name quaestiones disputatae ‘disputed questions’, which like the
college was introduced to Western Europe from the Islamic world in the second half of the 12th century (Beckwith 2012).
Adzhina-Tepa in what is now southwestern Tajikistan (Litvinsky and Zeimal’ (1971: 27-28).
Both halves of the Adzhina Тера ensemble were built according to the same plan, with allowances madе for the purpose
of еасh. The monastery part was arranged on the four sides of а square court (19х 19 metres), with four deep īwān
vestibules of identical size in the centre of each side. The īwāns wеге connected bу elbow-shaped corridors which had
exits leading into the соurt (Litvinsky and Zeimal’ (1971: 221).

Further detailed notes on the monastery plan, discussion connecting details of the vihāra plan
precisely with those of the madrasa plan, and its history, are given by Litvinsky and Zeimal’ (1971: 224-
225). In studies of Islamic architecture the design is known as the madrasa plan, as in the Madrasa of
Ulugh Beg in Samarkand, the Madrasa Mādar-i Shāh in Isfahan, and many others. The Western European
variant, the cloister plan, for example at Oxford and in many late medieval and Renaissance-period
monasteries, is a “no-īwān” version, telling us that it derives ultimately from the early vihāra plan, and
that this “monastic” version undoubtedly derives directly from Near Eastern monasteries.
The word vihāra is not the earliest attested word for the ‘monastery-college’. Early Chinese texts
seem to be unaware of the term vihāra, and the early Gāndhārī manuscripts do not mention it either.
Aramaic dērā ‘monastery’ is attested in several Late Old Chinese transcriptions, one of which is 寺 (in
Modern Mandarin pronunciation sì), the earliest attested word for ‘Buddhist monastery’ (Zürcher 2007:22,
38-39), all representing a borrowed word *dēRa (Beckwith 2014; 2017a; 2017b; Beckwith and Kiyose
2018). It was transmitted to East Asia by foreign Buddhist teachers and translators in China, of whom the
earliest known historically are the 2nd century CE Parthians An Shih-kao 安世高 (‘the Arsacid Shih-kao’)
and An Hsüan 安玄 (‘the Arsacid Hsüan’).3 The administrative language of the Parthian Empire was
heavily Iranicized late Imperial Aramaic written in Aramaic script.4 Aramaic dērā is well attested in two
different types of transcription in Chinese, as well as in Old Japanese (tera), Jurchen (taira), and Middle
Korean (tyǝr), which early borrowed the word via Chinese (Beckwith 2017a). Because the term vihāra is
used by archaeologists and art historians to refer specifically to the cloister plan of the core building, that
usage is followed here. The institution as a whole, including all of its structures, is referred to henceforth
as the “monastery”.
No one has yet identified who might have invented the monastery, but several remarkable historical
correlations do tell us approximately when and where it was invented. As noted, the Buddhist monastery
first appears in the second century CE at Taxila in Gandhāra and in various places in old “North China”,
including Loyang. The territory of Central Eurasia lying in between these two places was largely
occupied by the Kushan Empire and lands dominated, at least culturally, by the Kushans. The close
chronological correlation of the appearance of the monastery and monasticism in these two locations
suggests that it had been invented shortly beforehand. That would still put it in either the Kushan Empire
or the Parthian Empire, or both. The Aramaic word for monastery was evidently introduced to East Asia
by Parthians, the Parthian Empire, or perhaps the Indo-Parthian Kingdom, so they seem to be the locus of
invention. However, archaeological work has not revealed any definite vihāra sites in the former Parthian
Empire proper, leaving us with the territory of the Kushan Empire or the Indo-Parthian Kingdom. Since
the excavated mound of Sirkap, the Saka-Parthian and Kushan period city of Taxila, has no vihāras at all,
despite the number of Buddhist stūpas there,5 it seems that the monastery was invented in the late first
century CE, give or take a few decades, and then spread around the Kushan Empire and into neighboring

3 They arrived in northern China in the mid- to late-second century CE, and produced a number of Chinese translations of
Buddhist texts that are still preserved (Zürcher 2007).
4 The Parthian Empire lasted from ca. 141 BCE to 224 CE. Iranic words are already frequent in late Achaemenid texts written in
Imperial Aramaic (Naveh and Shaked 2012).
5 Schopen (2004: 77) notes, “Even considerably after Aśoka ... there are no references to vihāras. In none of the hundreds of
donative records from Bhārhut, Sāñcī, and Pauni does the term occur.”

2
regions (Beckwith 2015: 73; Schopen 2004: 79; Marshall 1951, I: 233, 320, 324).
The Kushan Empire, like the Parthian Empire, was founded by a steppe people who settled in heavily
agricultural, mostly sedentary Iran and Central Asia, where the majority of the adult population consisted
of land-bound farmers untrained for war. The small population of intrusive animal-herding Parthian
warrior rulers no doubt needed much additional local help to rule their vast sedentary realm, and the same
was surely true of the nomadic, herding warrior rulers of the Tokharoi-Saka-*Aśvin confederation
(Beckwith 2009: 380-383), who eventually became the Kushans, as noted in several other papers
presented at the conference.
The territory of the Kushan Empire, including even some of its more mountainous regions, seems to
have been blanketed with monasteries, as noted by Chinese travelers in Bactria in Late Antiquity and the
Early Middle Ages. In particular, Hui Ch’ao (fl. ca. 726), says that there were hundreds of vihāras in rural
eastern Bactria, in what is now Tajikistan, before Islamization (Fuchs 1938: 452; cf. Litvinskij 1985). In
1025-1026, when the Islamization of Central Asia was nearly complete, Arabic sources report that in the
very same region there were already twenty fully-endowed madrasas – that is, Islamicized vihāras – with
the same distinctive set of architectural forms, functions, and endowment (Beckwith 2012: 42). In modern
times, the remains of very many vihāras from Late Antiquity have been identified in Central Asia and
northwestern India, now Pakistan, many in rural areas (Gordon et al. 2009). These are the vihāras found
so far despite the difficult conditions still prevailing in those regions, which were the main centers of
Buddhism in the West from Antiquity into the Early Middle Ages. There must be many more sites. The
best-known ones archaeologically remain those excavated half a century or more ago in former Soviet
Central Asia and in the northwestern part of former British India, now Pakistan.
Next is the question of siting – the choice of locations where most vihāras were actually built.
Schopen (2006) has remarked that the Vinayas talk about the best siting for a monastery being an elevated
location with a good view of the surrounding region. And in Taxila, that is exactly where the excavated
vihāras are found, including some of the earliest ones. Taxila was an important Kushan city. Surely we
would expect at least one of its vihāras to be located in or very near the city proper, but all of them seem
to be located some distance away from it on one or another high hill overlooking the urbanized area. No
doubt the monks would have enjoyed the nice view, as Schopen has suggested. Yet consider the
architectural reality of the vihāra building. The oldest vihāras known, located around Taxila, and those
built as late as the eighth century in Central Tibet, such as the Kvacu vihāra and Haspo Rgyab vihāra,
both near Samye, are shown in old photographs to be small and heavily fortified (Snellgrove and
Richardson 1968: 36, 89). They are also on flat land, like Adzhina Tepa. Vihāras like these did not have
any windows or other openings to the outside other than the lone narrow doorway, which usually opens
onto the courtyard of another walled structure. This seems to be true of the remains of most of the early
vihāras in Gandhara that are located on high elevations, so it does not seem likely that the monks could
actually enjoy the view, or even see anything at all, either from the cells or from the courtyard, which was
not located on an outside wall of the vihāra at the brink of an overlooking height. The courtyard was
located in the center of the vihāra, surrounded by the cells, and the lone entryway was built, of necessity,
on the side facing the opposite direction, away from the view. In a few cases it seems to have been
possible to see out from some of the cells, though not very well. For example, at Takht-i Bhāi, an early
Gandhāran vihāra that does have openings in the wall built at the brink of the hillside looking out over the
settled region below, though not the other walls. They are small holes, perhaps not windows pe se, but
rather arrow-slits for archers to shoot from. Medieval historical works suggest that externally vihāras
were indistinguishable from forts. According to Jūzjānī, the Afghan raider Muḥammad b. Bakhtiyār Khiljī
[d. 1206] captured a fort in what is now Bihār Province in ca. 1200, killing the defenders; only afterward
did he discover that it was actually a “madrasa”, which the local people called a bihār, i.e., vihāra

3
(Raverty 1881: 552). The story may well be only a legend, but the point is that when Jūzjānī wrote in the
thirteenth century, everyone knew that a vihāra, from the outside, looked exactly like a fort.
It would not seem very wise for those in a fort not to be able to see what is outside the walls, and in
fact, a standard feature of many early Buddhist monasteries is a large stepped stūpa associated with the
vihāra, or even built inside its courtyard. It was often not so much large as it was saliently tall. In the case
of the two tallest ones noted by early medieval Chinese Buddhist writers, namely the Yung-ning Szu
永寧寺 or ‘Dērā (“monastery”) of Eternal Tranquility’ in Loyang, and the Nawbahār or ‘New Vihāra’ in
Balkh, described in the sixth and seventh centuries AD respectively, the great stūpa was also topped by a
very tall flagpole flying large banners that could be seen from many miles away. The stūpas’ height might
explain how they could see, but why did they spend so much effort and money on the flagpoles and their
banners if they were merely decorative? Both of these monasteries were large and were built in or
immediately outside the walls of large cities located on open flat land, where building on a hill or
mountainside was not an option. Did the low elevation necessitate an extra-tall stūpa with an extra-tall
flagpole? Or did the stūpa and flagpole have some other function? There are small monasteries in flat
rural areas too, such as the early Tibetan ones noted above, but none have tall stupas or flagpoles, so they
are a puzzling exception.
Early Buddhist monasteries thus have the following standard or ideal physical features:
¨ A cloister or vihāra – a fortified rectangular structure housing a medium-sized group of
regimented men in individual cells.
¨ When the monastery is situated outside a neighboring populated area, the vihāra is built ideally on
a hilltop or other promontory high above it.
¨ When the monastery is situated in an urban area, the vihāra is accompanied by a tall tower-like
stūpa structure topped by a tall flagpole flying large flags.
These features of early Buddhist monasteries, and many even today in predominantly Buddhist
countries, are all characteristics of military fortresses.
The center of a fortress in a relatively flat urbanized region is typically dominated by a donjon, as in
the 14th century Château de Vincennes outside Paris. The donjon is a tall, fortified tower that allows the
commander to see over the walls and provides a backup defense if the walls are breached. On the other
hand, many a famous fortified palace-capital is built on a high promontory and literally dominates the
area around it. For example, the Acropolis in Athens, the Potala in Lhasa, and the Alhambra in Granada,
all started out as hill-fortresses. All three lack anything like a donjon, and rely mainly on their strategic
height for protection, though clearly none of them was capable of holding out when faced with attack by a
serious military foe. Similarly, the fort-like vihāras were also not capable of holding out against a serious
military threat, as shown by the above-noted anecdote about Muḥammad b. Bakhtiyār Khiljī’s capture of
one despite its denizens’ stout defense. It has been pointed out that even the great walls of Balkh were
insufficient to defend that city (Bracey 2016), and surely the same is true of the very many cities taken by
determined foes in countless wars and insurrections over the centuries from East to West. The question
then must be, why spend all the resources to build city or monastery fortifications if they were useless for
keeping enemies out anyway? We must determine the real reason the fortifications were built.
Let us consider the wall par excellence, the massively monumental Great Wall of China (which is
famously visible from outer space) and its many Chinese predecessors and imitators, as well as Hadrian’s
Wall and other similar walls built by the Romans, Greeks, and others. If their purpose really was to keep
enemies out, all of them failed miserably. But that is not why they were built. The Chinese sources of the
time, which discuss the walls in frontier areas and their uses, plainly state that the walls were built mainly
to keep the people inside the walls from leaving. One might wonder why they would want to leave, and

4
where they would go, if the people outside the walls were terrible, cruel barbarians whose main interest
was in maiming and killing. But the primary historical sources written in Chinese, Greek, Latin, and other
languages actually tell us the opposite: the people inside wanted desperately to escape and join the people
outside. The Hsiung-nu,6 the later Huns, Mongols, etc., were better fed, worked less, and lived much
longer than the majority of people in the Chinese and Roman empires. It is well known that the ancient
Roman Empire had a declining population, but massive numbers of slaves. The ancient Chinese states
that built the early walls also had a population shortage, but again, many slaves. Clearly people were
considered to be a valuable resource. Yet when individual Romans and Chinese discovered how much
better off they would live as Huns or Hsiung-nu, and actually escaped, they encouraged their compatriots
to also run off and join the Huns or Hsiung-nu or whoever was out there, as the sources explicitly tell us.7
In short, the Great Wall, Hadrian’s Wall, and all the other walls were not built for defense.
Like walls, forts are not defensive constructions, they are offensive – designed to mark and hold
conquered territory and keep any people who had been captured, along with their territory, inside the new
borders that now included them. The Chinese and Roman empires expanded into non-Chinese and non-
Roman territory occupied by Central Eurasian peoples. These empires marked and held their conquered
territory with lines of forts and walls manned by soldiers and, sometimes, convicts.
What then about Buddhist monks? It is hard to say what the young men who took vows to become
monks were thinking when they joined the Saṃgha, but it is easy to imagine what many of them thought
about after they were cooped up for many years inside a vihāra, with their heads shaven, their bodies
obligatorily covered with a simple robe, eating barracks food, following a regimented routine day after
day. They thought of escape. Unfortunately, there was only one, guarded entryway, and the walls of the
vihāra were very thick and hard. Without serious tools it was practically impossible to escape. Yet if the
young men truly did freely join the Saṃgha, as we are told they did, why did they have to be kept in
vihāras most of the time?
There seem to have been many hundreds of monasteries in Kushan territory. When excavated they
are immediately recognizable because they are all but identical. Lack of evolutionary change suggests
rapid spread of a new design, which became a standard design and then lasted very long, surviving not
only in Buddhist cultures but later in Christian and Muslim cultures too.8 Nevertheless, this still does not
explain why the Saṃgha of monks had to be housed in regimented cells surrounded by thick, mostly
windowless walls, and fed and clothed by the neighboring villages.
In view of the typical vihāra’s fortified design and siting, whether on a promontory high above a city
or other important place (such as at Takht-i Bhāi in Gandhāra), or in a lowland city, with a high stupa or
tower of some kind (such as in Balkh and Loyang), it seems that the Kushan Empire covered the land
with vihāras to help hold conquered territory with garrison-like installations staffed by healthy, young,
male, sedentary citizens, Normative Buddhist monks, whose expenses were paid by the people of the

6 Despite many attempts to link the two peoples and names, Hsiung-nu 匈奴 could not have been pronounced anything like Hun
[hu:n] ~ [χuːn] when the Chinese first transcribed the name in the third century BC. The characters used were pronounced *suɣlâ
in the transcriptional Old Chinese dialect, becoming *suŋlâ, which became modern Hsiung-nu; *suɣlâ is obviously a relative of
attested Suguda ~ Suɣda ~ Suɣla ‘Sogdian’ (Beckwith 2018), from *skuδa ‘shooter; archer’, the source of the Greek form Skutha
‘Scythian’, which was pronounced Skula by the time of Herodotus, as shown by Szemerényi (1980), cf. Beckwith (2009:377-
380); not with Schmitt (2003). The traditional (non-linguistic) reconstruction (Pulleyblank 1991: 346 ✩xuawŋ + 227 ✩nɔ) gives
Old Chinese *swâŋlâ in the Central dialect, the ancestor of a Middle Chinese *χwɔŋlɔ, but see Beckwith (2018). Hsiung-nu was
Iranic (Beckwith 2018), whereas the Hun language was Archaic Turkic (Shimunek et al., 2015; Beckwith 2017c).
7 See Blockley (1983) for the first-person account of Priscus to the court of Attila the Hun, including Priscus’ account of a
Roman he met who had joined the Huns, and Beckwith (2009:116, 332) for the views of Chinese who had joined the Hsiung-nu.
8 Nālandā, the famous Mahāvihāra in Bihār, though a late site, consists of many standard vihāras that are virtually identical to
those built as much as a millennium earlier in Taxila. In the Seljuk Empire, the government built many standardized madrasas all
across its territory to spread the particular form of Islam favored by the Seljuks, but no doubt for more practical reasons too.

5
occupied territory. Monks were not allowed to work to support themselves. They also seem rarely to have
fought in actual wars,9 and no caches of weapons have been found at excavated vihāra sites. Yet the
distribution, completely regimented organization and lifestyle (down to the shaven heads and uniform-
like robes), fortified dwellings, and everything else involving the vihāras, all exactly followed the model
of a military movement into enemy territory – an invasion. It has often been noted by Indologists that
Buddhism seems to have been a strikingly alien religion in India; surely it was alien in Central Asia, too,
just as it certainly was in China.10
The sources are quite clear about how a vihāra was built and supported. Hui Ch’ao, who journeyed
through Central Asia to India and back in the early eighth century, remarks,
Whenever a vihāra is built, a village and its inhabitants are immediately donated as an offering to the Three
Jewels [i.e., to organized Buddhism]. Building a vihāra without making any donation of a village and its folk
is not done… (Van Bladel 2010: 67).
If we were to rewrite this account of religious expansion as an account of military expansion, it would say:
whenever a village and its inhabitants are subjugated, the rulers place them in servitude to the fortress
which is built to keep them in servitude to the conquerors. Hui Ch’ao tells us that this is exactly what
happened (though he makes it sound like a virtuous act). The similarity is too close to be coincidental.
Building a monastery, with its barracks-like cloister and its tower-like stupas, and staffing it with
regimented young men, was modeled directly on a military invasion, with its garrisoned forts and lookout
towers, and in both cases the conquest was accompanied by subjugation of the local people to support the
garrisoned invaders.
In many predominantly Buddhist cultures it is still customary for monks to go out once a day to beg
for food from the local people, and to accept used robes from them. Why did the early vihāras in Central
Asia and northwestern India switch to economic dependence upon the subjugated (“donated”) local agri-
cultural villages for their food, clothing, and other necessities? Why could the monks not beg for them, as
in earlier times among the “urban” śramaṇas (Beckwith 2015), and in other places down to the present?
It is significant that as far as we know, the Kushan monasteries were the first monasteries anywhere,
in any religion. There was no precedent for them. Adopting a specifically military model for something
new, but specifically religious, must have been a deliberate choice consciously made by someone.
Hagiographies, and even texts presented as histories, which were written during the expansionistic
period of Normative Buddhism, for example in early medieval Tibet and Japan, often describe the fierce
opponents of the heroic new Buddhist teachers attempting to introduce the religion to a land where other
beliefs were followed. The stories feature human opponents who supported various gods, as well as fierce
humaniform beings who are typically portrayed as warrior-demons that can only be overcome by the
magical power of fearless teachers (Shakabpa 1967: 36, 45). Once converted, the demons become fearless
Buddhist “protector” beings who are portrayed as warriors, whose former anti-Buddhist stance is shown
by their fierce demeanor, brandishing weapons. The most famous such “protectors” are the “four kings”,
still well known in Central Eurasia and East Asia. Their large, fearsome statues protect the gates of
medieval monasteries as far as Nara and Kyoto. The most famous of them is Vaiśravaṇa, protector of the
north, whose weapons are the umbrella (the flags on top of stupas) and the stupa itself. In some cases,
enemy invasions are described as diverted, avoided, or destroyed by such protectors, who are summoned
via the prayers or magic of the teachers (Walter 2009: 175, 200). These Buddhist tales, however
ahistorical, exactly mirror an actual military invasion.
The rapid spread of the newly invented, standardized Buddhist monasteries, with regimented soldier-

9 It is well known that in the later Middle Ages, in Europe, Tibet, and East Asia, many monasteries did send monks out to fight.
10 This once again raises the question of the actual homeland of the Buddha and Early Buddhism (Beckwith 2015).

6
like monks, was clearly intended to spread the new Normative Buddhism, which promises that virtuous
people, especially good monks, will be reborn in one of the new Buddhist paradises. Bearing in mind the
non-hierarchical nature of Buddhism, even Normative Buddhism, with no centralized organization to do
all this, one wonders how Buddhists found the money to build and staff the monasteries. How did they
carry out their peaceful “religious conquest” of foreign lands defended by warriors hostile to Buddhism,
as their folklore claims they did?
Buddhist tradition is remarkably consistent on this point. In every Asian country where Buddhism
successfully spread, the monks first convinced the rulers and other people at the top of society to support
them (Walter 2009). Historical accounts and legends both suggest this is exactly what did happen. For
example, the historical account of the formal establishment of Buddhism in Japan in 552 CE (Nihon Shoki
XIX.36) focuses on an aristocrat’s opening of the first vihāra there, the opposition of other aristocrats,
omenous events credited to unhappy native gods, and the final victory of the Buddhists. In the Tibetan
Empire (ca. 618 – ca. 866 CE), according to an “ex eventu” prophecy, Emperor Khri Lde Gtsug Brtsan
(“Mes Ag-tshoms”, d. 755 CE) and his Chinese consort Princess Chin-ch’eng invited in and supported a
number of refugee monks from Khotan and other Central Asian countries. Not long afterward, a plague
struck Tibet (recorded in the Old Tibetan Annals), killing the princess and the Tibetan crown prince. The
story continues that the monks were expelled and then went to Gandhāra, where they were at first
supported by the local king (Nattier 1991: 196). These are stories, even if based on actual history. The
Japanese and Tibetan rulers long supported their traditional non-Buddhist beliefs. Yet it is a historical fact
that at the same time they did support Buddhism and helped it spread throughout their realms.
The Kushans themselves seem to have been comfortable with the traditional Bactrian gods, as
suggested by the dominant images on their coins, but as in Japan and Tibet, Buddhism could not have
spread in the Kushan Empire without the Kushan rulers’ approval, at least. The above discussion suggests
that the new Kushan Empire and the new Normative Buddhist saṃgha did agree to work together, with
the Kushan dynasty handing over villages to be the feudatory supporters of the monasteries, which eager
young Buddhists staffed as monks. Perhaps as part of the agreement, monks in the Kushan Empire were
restricted to the standardized fort-like monasteries, which dominated the local villages that had to support
them. Even if the Kushans themselves did not really become enthusiastic Buddhists, we must assume that
they sanctioned the expansion of Normative, monastic Buddhism. Certainly they did profit from that
expansion. The end result of the arrangement was the rapid spread of Kushan Buddhism as well as
Kushan cultural and political influence throughout their own empire and far beyond it, across vast regions
of Central Asia apparently not ruled by them, all the way to China.
REFERENCES
Barthold, W. 1964. Сочинения, II.2. Moscow: Izdatel'stvo vostochnoj literatury.
Beckwith, Christopher I. 1993. The Tibetan Empire in Central Asia: History of the Struggle for Great
Power among Tibetans, Turks, Arabs, and Chinese during the Early Middle Ages. Princeton:
Princeton University Press. First edition 1987.
—— 2009. Empires of the Silk Road: A History of Central Eurasia from the Bronze Age to the Present.
Princeton: Princeton University Press.
—— 2012. Warriors of the Cloisters: The Central Asian Origins of Science in the Medieval World.
Princeton: Princeton University Press.
—— 2014. The Aramaic Source of the East Asian Word for ‘Buddhist Monastery’: On the Spread of
Central Asian Buddhism in the Kushan Period. Journal Asiatique 302.1: 111-138.
—— 2015. Greek Buddha: Pyrrho’s Encounter with Early Buddhism in Central Asia. Princeton:
Princeton University Press.

7
—— 2017a. Once again on the Aramaic Word for ‘Monastery’ in East Asia. Journal Asiatique 305.2:
211-224.
—— 2017b. How did the Chinese actually first hear and transcribe Buddhist terms? The two kinds of
Late Old Chinese transcriptions and their significance. Paper presented at the International Association
of Buddhist Studies conference, Toronto, August 25, 2017.
—— 2017c. The Archaic Turkic Prophecy and the sons of Attila: analysis of the earliest Turkic language
data. Paper presented at the Symposium on Turkic Languages and Linguistics, University of
Washington, Seattle, Oct. 5, 2017.
—— 2018. On the Ethnolinguistic Identity of the Hsiung-nu. In Zs. Gulacsi, ed. Language, Government,
and Religion in the World of the Turks: Festschrift for Larry Clark at Seventy-Five. Silk Road Studies
19. Brepols: Turnhout, pp. 33-55.
Beckwith, Christopher I. and Gisaburo N. Kiyose 2018. Apocope of Late Old Chinese Short *ă: Early
Central Asian loanword and Old Japanese Evidence for Old Chinese Disyllabic Morphemes. Acta
Orientalia Academiae Scientiarum Hungaricae 71.2: 145-160.
Bracey, Robert 2016. The Limits of Kushan Power and the Limits of Evidence. Paper presented at the
conference ‘The Limits of Empire in Ancient Afghanistan’, Chicago, October 6, 2016.
Fuchs, Walter. 1938. Hui-ch’ao’s Pilgerreise durch Nordwest-Indien und Zentral-Asien um 726.
Sitzungsberichte der Preussischen Akademie der Wissenschaften (philosophisch-historische Klasse)
30: 426–469.

Gordon, Stewart, T. Matthew Ciolek, Lizabeth H. Piel, and Gita Gunatilleke 2009-present. Electronic
Atlas of Buddhist Monasteries of South Asia, SE Asia, Central Asia, and East Asia between approx.
200 CE and approx. 1200 CE. <http://www.ciolek.com/geo-monastic/geo-monasteries-home.html>
and <http://monastic-asia.wikidot.com/>
Litvinskij, Boris A. 1985. Ajina Tepe. In: Ehsan Yarshater, ed., Encyclopaedia Iranica, I:703-705.
London: Routledge & Kegan Paul.
Litvinskij, Boris A. and Tamara I. Zeimal’ 1971. Аджина-Тепа. Moscow: Iskusstvo.
Marshall, John Hubert. 1951. Taxila: An Illustrated Account of Archaeological Excavations Carried out
at Taxila under the Orders of the Government of India between the Years 1913 and 1934. 3 vols.
Cambridge: Cambridge University Press.
Nattier, Jan 1991. Once upon a Future Time: Studies in a Buddhist Prophecy of Decline. Berkeley: Asian
Humanities Press.
Naveh, Joseph and Shaul Shaked 2012. Aramaic Documents from Ancient Bactria (Fourth Century BCE.)
from the Khalili Collections. London: The Khalili Family Trust.
Pulleyblank, E.G. 1991. Lexicon of Reconstructed Pronunciation in Early Middle Chinese, Late Middle
Chinese, and Early Mandarin. Vancouver: UBC Press 1991.
Raverty, Henry George, trans. 1881. T̤ abakāt-i-Nāṣirī : A General History of the Muhammadan Dynasties
of Asia, including Hindustan, from A.H. 194 (810 A.D.) to A.H. 658 (1260 A.D.), and the Irruption of
the Infidel Mughals into Islam. 2 vols. London: Gilbert & Rivington.
Schopen, Gregory 2004. Buddhist Monks and Business Matters: Still More Papers on Monastic
Buddhism in India. Honolulu: University of Hawaii Press.
—— 2006. The Buddhist “Monastery” and the Indian Garden: Aesthetics, Assimilations, and the Siting
of Monastic Establishments. Journal of the American Oriental Society 126.4: 487-505.
Schmitt, Rüdiger 2003. Die skythischen Personennamen bei Herodot. Annali dell’Università degli studi di
Napoli “L’Orientale” 63.1-4: 1-31.
Shakabpa, W.D. 1967. Tibet: A Political History. New Haven: Yale University Press.
Snellgrove, David and Hugh Richardson 1968. A Cultural History of Tibet. New York: F.A. Praeger.

8
Szemerényi, Oswald J.L. 1980. Four Old Iranian Ethnic Names: Scythian – Skudra – Sogdian – Saka.
Österreichischen Akademie der Wissenschaften, Philosophisch-Historische Klasse, Sitzungsberichte,
371 Band. Vienna: Verlag der Österreichischen Akademie der Wissenschaften.
Van Bladel, Kevin. 2010. The Bactrian Background of the Barmakids. In A. Akasoy, C. Burnett, and R.
Yoeli-Tlalim, eds., Islam and Tibet: Interactions along the Musk Routes, 43-88. Farnham: Ashgate.
Walter, Michael L. 2009. Buddhism and Empire: The Political and Religious Culture of Early Tibet.
Leiden: Brill.
Zürcher, E. 2007. The Buddhist Conquest of China: The Spread and Adaptation of Buddhism in Early
Medieval China. 3rd ed. Leiden: Brill.