You are on page 1of 22

"ALL ROADS LEAD TO ROME": CHINESE KNOWLEDGE OF THE ROMAN EMPIRE

Author(s): D. D. LESLIE and K. J. H. GARDINER


Source: Journal of Asian History, Vol. 29, No. 1 (1995), pp. 61-81
Published by: Harrassowitz Verlag
Stable URL: http://www.jstor.org/stable/41930974
Accessed: 28-12-2015 14:13 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.

Harrassowitz Verlag is collaborating with JSTOR to digitize, preserve and extend access to Journal of Asian History.

http://www.jstor.org

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
D. D. LESLIE AND K. J. H. GARDINER
(Canberra)

"ALL ROADS LEAD TO ROME":


CHINESE KNOWLEDGE OF THE ROMAN EMPIRE

Introduction1

The Roman Empire is not mentioned in the Shih-chi or Han-shu,


and was certainly not known of in China before the 1st century
C. E. However, from the Hou Han period (the Later Han, 23-220)
until the Sui dynasty (589-618), a great state in the west called Ta-
Ch'in is described in some detail, in particular in the Hou-Han-shu,
Hou-Han-chi, Wei-liieh (in the San-kuo-chih), Chin-shu and Wei-
shu (and Pei-shih).2
From the Sui, and especially from the T'ang (618-907), a new
name Fu-lin appears, looked on as the same state as Ta-Ch'in. These
sources also give another name Li-kan (or Li-hsien,etc.) forTa-Ch'in.
In fact, Li-kan (also called Li-kan/T'iao-chihand T'iao-chih in the
Shih-chi and Han-shu)3 is the Seleucid Empire, based on Antioch
in Syria and Seleucia in Mesopotamia, defeated by the Parthians
(An-hsi in Chinese sources) around 150-129 B.C.E., and by Pompey

1 Wehopesoonto publishourbook "TheRomanEmpirein ChineseSources",


whichwillincludenew translations of all thetexts,and a detailedanalysis
oftheplace-names, routesand productsmentioned, and also a fulldiscus-
sionofthehistorical andhistoriographical problems.Wewouldliketo thank
Dr.Rafede Crespigny forsomevaluablesuggestions abouttheroutes.
2 HHS 88 (lieh-chuan78); HHC 15; WLin SKC 30; CS 97 (ļ-c67); WS 102(ļ-c
90); PS 97 (l-c85). Muchofthismaterialhas beencopiedin theencyclopae-
dias,inparticular theTung-tien193,T'ai-p'ingyü-lan792,andKu-chinťu-
shu chi-ch'eng , 60 in ts'e214.Mostofthismaterialis in Hirth1885,in Chang
Hsing-lang, 1930,and elsewhere.
3 SC 123(l-c63); HS 96A(ļ-c 66A).For theproofthatLi-kan/T'iao-chih is the
SeleucidEmpire,see Leslie/Gardiner.
JAH
29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
62 D. D. LESLIEAND K.J.H. GARDINER

in 64 B.C.E. Fu-lin is generally accepted as Byzantium.4That leaves


Ta-Ch'in.
There can be no doubt that the earlier scholars who took Ta-Ch'in
"the greatest state in the west" as the whole Roman Empire based
on Rome were correct. Li-kan is replaced by Ta-Ch'in, Ta-Ch'in by
Fu-lin. The Seleucid Empire (in Syria) is replaced by Rome, Rome
in the east by Byzantium. The Chinese historians erred in thinking
these three were the same country,due no doubt to the continuity
of their own culture, but were accurate about the sequence and
transferof power in the geographical area west of Parthia and Per-
sia.
Unfortunately,more sophisticated scholars, Yule (and Cordier),
Hirth, Chavannes, Pelliot and Shiratori, followed by many others,5
insisted that only the eastern part of the Roman Empire was meant
by Ta-Ch'in, Hirth and Chavannes taking its capital as Antioch in
Syria, with Pelliot and Shiratori,slightlymore plausibly, taking it as
Alexandria in Egypt.
There are two overwhelming arguments against this revisionist
view.
Firstly,we suggest, it is axiomatic, in other words obvious and
not really in need of proof, that Ta-Ch'in "the greatest state in the
west" is the whole Roman Empire with capital Rome. The eastern
Roman Empire was not a separate state at this time, and neither
Antioch nor Alexandria was ever a capital cityof the Roman Empire.
Secondly, the question as to whether the Chinese imagined Ta-
Ch'in to be centred on Antioch, Alexandria or Rome is one which
can be positively settled by an analysis of the routes given in the
Chinese sources (in particular in the Hou-Han-shu and Wei-lüeh).
As we shall see, "All Roads Lead to Rome".

4 Hirthhas includeddescriptionsof Fu-lin(Byzantium)fromtheT'angHisto-


ries and elsewhere.These are misleadingratherthanhelpfulforTa-Ch'in
(theRomanEmpirebased on Rome).
5 IncludingB. Laufer,O. Franke,H. Maspero,J.Needham,J.Gernet.Among
thosewhohavetakenTa-Ch'inas thewholeRomanEmpireare WangHsien-
ch'ien,P.Demiéville,Tazaka Kôdô, E. Schafer,H. Bielenstein.See "Refer-
ences".
JAH
29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
CHINESEKNOWLEDGE
OF THE ROMANEMPIRE 63

The Reliability of the Texts

Before giving the evidence for this, we would like to dispose of


two other influentialand erroneous views.
Firstly,the view expressed by the classical scholar Raschke.6 In
his very detailed study, Raschke suggests that the Chinese sources
are not reliable, and moreover that such a work as the Hou-Han-
shu which purports to describe events of the 1st and 2nd centuries
is better looked on as referringto the 5th century,its date of compi-
lation. Raschke has, in fact, completely misunderstood the nature of
Chinese historiography.Pan Ku, author of the Han-shu, and most
later historians, used what was basically a "scissors and paste" me-
thod, scrupulously avoiding introducing their later knowledge into
archival sources collected during the previous dynasty. What we
have to watch out for is not the backward anachronisms that
Raschke suggests, but forward ones. As Hirthputs it:
In accordance with the Chinese method observed up to the pres-
ent day ... all that was recognized as true hundreds of years ago
must be true for ever, and thus may be quoted without further
scrutiny.7
The actual reliabilityof the Hou-Han-shu compiled by Fan Ye is
well expressed by Bielenstein:8
In rearrangingthe material, Fan Ye did not rewrite it. He copied
his sources so closely as actually sometimes to make the HHS
inconsistent.
Daffinà and Petech have also discussed this phenomenon.9 Petech
writes:

Geographical terminology is most long-lived with the Chinese;


they could not discard the famous name, and sought thereafterto
connect it with other countries.

6 Raschke's workis veryvaluableforthewesternend,butis misleading


for
theChinesesources.
7 Hirth,1885,p. 17.
8 Bielenstein,
1954,pp. 11,15.
9 Petech,1950,pp. 78-79; Daffinà,1972.
JAH
29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
64 D. D. LESLIEAND K.J.H. GARDINER

"and thus the spook of Chi-pin haunted traditional Chinese


geography down to its end in the 19th Century."
This observation for Chi-pin(Gandhâra in the early texts, Kashmir
in later ones) can also be applied to An-hsi (Parthia) which survives
in the texts well afterParthia had been conquered by Sassanian Per-
sia (Po-ssu); and also to T'iao-chih (originally the Seleucid Empire,
later the rump state of Characene). It is sometimes impossible to
decide whether the name is merely an echo or spook of the previous
identity.
Linked to this to some extent is the belief of Shiratori that the
Hou-Han-shu has used the earlier Wei-lüeh and has summarized it.
We cannot go into any details here, but just wish to state our firm
view that the shorterHou-Han-shu description antedates the longer
Wei-lüeh one, which refers to the period 221-265. It is likely that
both sources, and the Hou-Han-chi, have used the 2nd century
Tung-kuan Han-chi, only parts of which have been preserved.10
A second misleading attitude is that of Shiratori who looked on
much of the description of Ta-Ch'in as Utopian. It is true that later
sources, including some Taoist ones,11 have mixed in information
that is best looked on as legendary or mythological,possibly even
Utopian,but this is not true for our basic sources, the Hou-Han-shu
and Wei-lüeh.There are no doubt errors and discrepancies, but they
are describing a real country,relyingon the informationthey have
received from Kan Ying who reached Mesopotamia in 97 C.E., and
the An-tun (Antonius) 'envoys' (merchants?) of 166 who reached
China. The Wei-lüeh undoubtedly had a furthersource which unfor-
tunatelycannot be traced.
Wolters believes that
"the Chinese were capable of writingmatter-of-fact
accounts of
foreignparts."12
Miyazaki actually writes:
"The Han people was more conversant with Roman affairsthan
the Romans were with China."13

10 The extantTKHChas onlyone sentenceaboutTa-Ch'in.


11 See, in particular,
Maspero,1950.
12 Wolters, 1967,p. 33.
13 Miyazaki, 1959.
JAH
29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
OF THE ROMANEMPIRE
CHINESEKNOWLEDGE 65

We would add to this (somewhat unfairly) that no Chinese de-


scription of the Romans is more absurd than Pliny's description of
the Seres ("the silk people") as "red-hairedwith blue eyes", were it
not for the fact that Janvierhas recentlyargued stronglyagainst the
commonly accepted identification of the Seres ("the silk people")
with the Chinese.14
We would like to add here two remarkable pieces of evidence for
the authenticityand reliabilityof the Chinese texts.

Firstlythe Wei-liiehhas the following passage:


They (the people of Ta-Ch'in) also commonly profitby obtaining
Chinese silk and unravelling it to make hu (Iranian or Central
Asian) damask.
This unravelling of the heavy Chinese silk by the Romans to pro-
duce a more delicate fabric has been pointed out by several scholars,
in particular Hirth (who criticizes Yule for doubting it), Wieger,Col-
less, Ferguson and others.15 Several scholars, including the above
ones, have pointed out two references in Pliny and one in Lucan
which mention this identical fact.16
Secondly, the Hou-Han-shu states that in 100 C.E. an embassy
came to China from Meng-ch'i Tou-le, two states far in the west.
There is a definite possibility that this refers to the famous trip
undertaken by the merchant envoys sent by Maes Titanios the Mace-
donian around 100-120 C.E. as recorded by Ptolemy.17If so, the
Chinese mistake in taking this as two states Meng-ch'i and Tou-le is

14 Thoughthisdescription ofsomeintermediaries
ofPliny'sis almostcertainly
in thesilktraderatherthanof the Chinesethemselves, we are inclinedto
stickwiththeconsensusviewthattheSeres,howevervaguelydescribedin
Greekand Romansources,are meantto be the originalproducersof silk.
However,see thecriticism ofJanvier.
15 Hirth,p. 257ff.(Yule/Cordier,pp. 197-200),Wieger,I, p. 885,Colless,p. 60,
Ferguson, pp.590-592.Hudson,pp.91-92, 97,Needham,I, p. 193,missthe
originalWLsource,onlycitingmuchlaterversions.
16 Pliny,NaturalHistory , VI, 17.,xx(54), Rackham,pp.378-379,Coedès,pp.
xiv,13; Lucan,Pharisalia X, pp. 141-143,translation Duff,Loeb Classical
Library(London: 1928), pp.600-601. Schoff,pp.264-265, and Janvier,
p. 279,mentionbothPlinyand Lucan.Raschke,p. 623 and note284,p. 723,
notestheLucanreference, butmostsurprisinglyhas overlooked theChinese
passagewhichmentions it.
17 Ptolemy, I, 11,1-7, McCrindle,p. 9ff.
JAH
29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
66 D. D. LESLIEAND K.J.H. GARDINER

excusable. If correct, this would be the one and only meeting be-
tween early China and Rome mentioned in both.18

The Transmission of Information

Where does the Chinese information about the Roman Empire


come from?
The great explorer Chang Ch'ien visited Central Asia around 138-
123 B.C.E., and he or his close followers described An-hsi (Parthia),
Li-kan/T'iao-chih(the Seleucid Empire), but not Wu-i-shan-li(Ara-
chosia) or Chi-pin (Gandhâra). Chang Ch'ien himself did not visit
any of these four places. The descriptions of Arachosia and Gand-
hâra are, in fact, not found in the Shih-chi of Ssu-ma Ch'ien, but
only in the (later) Han-shu of Pan Ku. This suggests that a new
route, to Gandhâra and Arachosia (and thence to Characene) was
added to the earlier route to Parthia.
When we turnto the Hou-Han-shu, we findthis confirmedforthe
route of Kan Ying, sent in 97 C.E. by the great general Pan Ch'ao on
an embassy to Ta-Ch'in. He went via Gandhâra and Arachosia to
Characene, almost certainly across the Kerman desert. He did not
reach furtherthan Mesopotamia, possibly crossing over, on the way
back to Pan Ch'ao's headquarters in the Western Regions subordi-
nate to China, fromCharax (in Characene) to Ctesiphon (Seleucia),
but not gettingto Antioch. His description of Ta-Ch'in is the first.
The next major source of informationcame in 166 C.E. with a
visit to China of envoys fromAn-tun(Antoninus, whetherAntoninus
Pius, ruled 138-161, or Marcus Aurelius Antoninus, ruled 161-180,
is not certain). These were probably merchants claiming to be en-
voys, for Roman records do not mention any attempt to send an
embassy to the Seres. We can accept the date 166, though it may be
that these merchants had been in South-east Asia for several years
and may not have known of the death of Antoninus Pius.
The Hou-Han-shu account is supposed to be based on Kan Ying's
report of 97 and a follow-up report by Pan Yung, son of Pan Ch'ao,
in c. 125. However, the 166 An-tunmission is recorded, and we sug-

18 The identification
as Macedoniawas givenby Herrmann,
pp.256-257,and
K. Chang,1955,p. 80.
29/1
JAH (1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
CHINESEKNOWLEDGE
OF THE ROMANEMPIRE 67

gest that some other portions of the description must have been
from 166 rather than earlier. One example of this is the claim that
the Parthians prevented communication between Rome and China.
Several texts write this explicitly.
"The king of this kingdom (Ta-Ch'in) always wanted to enter into
diplomatic relations with the Han. But the Parthians wanted the
trade with them in Chinese silk and so put obstacles in their way,
so that they could never have direct relations (with China)."
This could hardly have come from Kan Ying, whose main source
of informationwas Parthian.
We will discuss later this distinction between information ob-
tained overland and that obtained by the sea route.

The Cities of Ta-Ch'in

The Hou-Han-shu does not name a single city of Ta-Ch'in. The


Wei-lüeh names two or three, and also has a most significantstate-
ment:
"There are altogether three large capital cities."
These three capital cities of the "greatest state in the west" are
identifiedas follows:19
Hirth:Delta, Heptanomis, Thebai's (in Egypt);
Forke: Sestos, Abydos and Lampsakos (on the Hellespont);
Ting Ch'ien: Milan, Marseilles and Taragon;
Herrmann:All in Arabia;
Shiratori:All in Egypt.
It is hardly surprisingthat a classical scholar like Raschke is a bit
skeptical about Chinese sources!
These ludicrous identifications are all based on the individual
scholar's personal idée fixe. For Hirthand Shiratori,only the eastern
part of the Roman Empire is described; for Herrmann,Ta-Ch'in, the
greatest power in the west, is Arabia Felix; for Forke, the main area
described is Asia Minor, though he does take Ta-Ch'in as Rome in
Italy.

pp. 124-
p. 182;Forke,p. 55; TingCh'ien(in Lu Pi, p. 63a); Shiratori,
19 Hirth,
125.
JAH29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
68 D. D. LESLIEAND K J.H. GARDINER

Our own idée fixe is to adopt a naive commonsense attitude.


Thus "the greatest state (country) in the west" can only be the
Roman Empire, "the three large capital cities of Ta-Ch'in" can
only be Rome, Alexandria and Antioch.20
One should stress here that the text does not name these three
cities. We are inclined to look on the statement as a general one. The
Chinese knew that the Roman Empire had three verylarge cities, but
did not necessarily know their names. We have, in fact, great diffi-
culty in identifyingthe cities actually named.
These are An-ku, Ch'ih-san (Wu-ch'ih-san is almost certainly the
same city), and Wu-tan.A fourthname An-tu,called the capital of
Ta-Ch'in in the Wei-shu, but not earlier, refers to the Wei period
(386-534). It may be Antioch, as Hirth believes, but might also be
Constantinople. It is not connected to the capital of Ta-Ch'in in the
earlier period.
Ch'ih-san mightbe Alexandria, Wu-tanmightbe Athens, but these
identificationsare unproven, speculative in fact.
However, An-ku, we suggest, is identifiable. Most of the routes
given proceed fromhere to the capital city of Ta-Ch'in. Forke ingeni-
ously identifiedit as Akko (Acre), located on the Phoenician coast.
It is, in fact, Antioch (as we hope to demonstrate below). When we
look at Karlgren's archaic and ancient pronunciations, an-kuk and
an-kuk, with a variant an-giuk and an-iwok, we see an excellent fit.
Taking An-ku as Antioch is one of the rocks upon which all our
analysis is founded. However, we have one apparently inescapable
objection to this identification:An-ku is stated to be on the frontier
of An-hsi (Parthia), it is not in Ta-Ch'in (the Roman Empire).
We have two plausible arguments to counter this. Firstly,there is
no evidence that the Chinese knew of the Roman conquest of Syria
by Pompey in 64 B.C.E. The Shih-chi and Han-shu do not mention
the Roman Empire, only the Seleucids, most of their information
stemming from missions to and from Parthia in c. 100 B.C.E. New
informationonly came much later with Parthian embassies to China
of 87, 94 and 101 C.E., and especially with Kan Ying's abortive mis-
sion of 97.

20 A fourth
largecitywas Carthage.
Butthisseemsless likelyto haveimpinged
on theChinese.
JAH
29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
CHINESEKNOWLEDGE
OF THE ROMANEMPIRE 69

Since Kan Ying never got beyond Characene and Parthia, his infor-
mation did not come fromthe Romans. Ta-Ch'in,the Roman Empire,
is "west of the sea" (i. e. Rome and/orAlexandria, west of the Indian
Ocean or of the Mediterranean). There is no mention of Antioch; he
did not reach Antioch. He, in tact, believed that Ta-Ch'in was re-
ached only after a long and hazardous sea-journey across the sea
(to Alexandria and/or Rome). When the Chinese did get furtherin-
formation about the Roman Empire in 166 C.E. it came from
seafarers who presumably came from Alexandria and would have
no reason to mention the rival Metropolis Antioch. Antioch is not
mentioned in Chinese sources before the 3rd century.
How can the Wei-lüeh,which does mention Antioch as An-ku,put
it on the Parthian frontier?Here is our second argument.We suggest
that the Chinese looked on the whole area east of the Mediterranean
as part of the Parthian/Persianempire.
Wolters writes:
Until 455 it was probably assumed that An-hsi was still the appro-
priate name for the Iranian Empire. A screen of middle men on
whom the northern Chinese depended for political intelligence
would have kept them in ignorance.21
We have a problem here. When the Chinese write kuo "state,
countryor kingdom",we are not always sure that they are referring
to a political entityrather than a geographical area. The Seleucid
Empire, Parthian Empire and even Roman Empire had a weak hold
on some of the outlying parts. Mesopotamia, in particular, was
subject to various invasions and temporary conquests. Several
scholars, including Raschke,22 have pointed out that culturallyand
economically Syria and Mesopotamia were linked.
The name An-hsi is found for Iran even after centuries of Sassan-
ian rule. Perhaps the Chinese never realized that Syria (and Antioch)
was no longer part of the Seleucid Empire, replaced by the Parthian
Empire, replaced by the Sassanian Persian Empire.

p. 82.
21 Wolters,
22 Raschke,p. 642.
JAH29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
70 D. D. LESLIEAND K.J.H. GARDINER

Vassal States of Ta-Ch'in

The Hou-Han-shu does not name any of the "several tens of small
kingdoms which are subject to Ta-Ch'in". The Wei-lüeh picks out
six: Ts'e-san, Lü-fen,Ch'ieh-lan, Hsien-tu, Ssu-fu,Yü-lo.
We are confidentof the location and identityof the firsttwo. Here
are their descriptions.
The king of Ts'e-san is subject to Ta-Ch'in. His (seat of) govern-
ment lies rightin the middle of the sea. In the north,one reaches
Lii-fen,going by water for over a year, (but) with a favorable wind
one arrives after one month. It is the nearest to the city of An-ku
in Parthia. To the south-west,one reaches the capital of Ta-Ch'in,
we do not know how many li distant.
The king of Lü-fenis subject to Ta-Ch'in. His (seat of) government
is 2000 li (i. e. 500 miles) from the capital of Ta-Ch'in. From the
cityof Lü-fen you go west to reach (the capital of) Ta-Ch'in, cross-
ing a sea-flyingbridge, 230 li (i. e. 57 miles!) in length. The route
crossing the sea goes south-west; if you go round the sea, it is due
west.
We have no hesitation, with Ting Ch'ien,23in taking Ts'e-san as an
island kingdom in the Mediterranean, almost certainlyCyprus, near
to Antioch, and south of Asia Minor,with Lü-fenin Asia Minor,prob-
ably Cilicia.24
We also favour the suggestion of Bretschneider, Wang Hsien-
ch'ien, Forke, FitzGerald, Lu Pi and others that the sea-flyingbridge
must refer to the legendary bridge over the Hellespont built by
Xerxes.25

23 TingCh'ien,in Lu Pi,p. 66b.


24 Wehavenotemphasizedphoneticidentifications. Atbest,theycan onlybe
consideredas confirmatory evidenceforgeographical andhistoricalidentifi-
cations.For Ts'e-san,we have no phoneticsuggestion whatsoever. For Lü-
fen,we offerthepossibility thatLü-fenshouldbe Lü-hsi,alreadysuggested
byShiratori,p. 105,thoughhisidentificationwas notCilicia.
25 WangHsien-ch'ien et al., p. 3248,Forke,p. 54, FitzGerald,p. 301, Lu Pi,
pp.62b,67a. Severalscholarshavearguedthatthisflying bridgemustbe at
Zeugma.However,theHHS andHHC bothwriteof a sea-flying bridge,and
therouteinvolvedis fromthecapitalofTa-Ch'in(i.e. Rome)to Hai-pei'north
ofthesea', notfromor to Antioch.
JAH
29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
CHINESEKNOWLEDGE
OF THE ROMANEMPIRE 71

We have a reference to this sea-flyingbridge in the Hou-Han-shu


and Hou-Han-chi, going from Ta-Ch'in to the various countries of
Hai-pei "northof the sea", which we take to be Asia Minor,north of
the Mediterranean. Hai-hsi, "west of the sea", another name for Ta-
Ch'in, is Rome or Italy,to the west of the Mediterranean, Hai-tung,
"east of the sea" is probably Syria and Mesopotamia, east of the
Mediterranean.
Identificationof the other four vassal states is complicated by the
fact that they are not linked geographically with any route to Ta-
Ch'in. They are, however, linked to one another, and to two places,
Ssu-t'ao, which seems to be in Parthia, and Ssu-lo, stated to be in
Parthia.26
They are not linked in any way to An-ku,Lü-fenor Ts'e-san. This
implies, we suggest, two separate routes. The firstto the north,from
Antioch (An-ku) via Asia Minor (Hai-pei) to Rome (Hai-hsi), the se-
cond, to the south, from Seleucia, via Palmyra and Damascus. Now
the routes to the vassal states start at An-ku (Antioch) and Ssu-t'ao.
We suggest that Ssu-t'ao, taken by Hirth and other to be Sittake, a
small city near Seleucia, is an error for Ssu-lung (the characters ťao
and lung are easily confused), and take Ssu-t'ao as Seleucia.
For what it is worth, we tentativelyput Ch'ieh-lan (corrected to
Tan-Ian by Shiratori) as Palmyra (Tadmor);27 Ssu-fu as Damascus;
Hsien-tu (the vassal furthestsouth) as perhaps Jerusalem; and Yti-
lo as Dura Europos.28
Almost certainly,these four vassal states are in the area between
the Roman Empire proper and Parthia, perhaps even in Hai-tung
"east of the sea", contrasted with Hai-hsi "west of the sea", Italy or
Rome, though this is nowhere stated in the text.

26 A speculationis thatSsu-lois Osrhoene,semi-independent of Parthiauntil


the3rdcentury whenitbecamesubjectto Rome.Another is thatitrefersto
Syria.
pp.28, 108.As pointedout by Pelliot1921,p. 143,an important
27 Shiratori,
argument forsuchcorrections is thatcertaincharacters are commonly used
othersnot.
fortransliterations,
28 Yü-lois perhapsthemostcontroversial ofall. It aloneis namedin theHou-
Han-shu, butitspositiontherecannoteasilybe reconciledwiththevassal
stateintheWei-lüeh. Weplace theHHS Yü-lofurther north, perhapsinAsia
Minoror evenin Greece.Butsurelytheymustbe thesameplace?
JAH
29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
72 D. D. LESLIEAND K.J.H. GARDINER

There are routes given to the capital of Ta-Ch'in (Rome) fromAn-


ku (Antioch), from Ts'e-san (Cyprus), and from Lü-fen (Cilicia in
Asia Minor), but none fromSsu-t'ao. Unfortunately,the Chinese text
just takes for grantedthe two places An-ku and Ssu-t'ao (surely they
were prettylarge), it gives no route to either fromthe east.

The Routes fromAn-ku

From the city of An-ku going due north by land one gets to Hai-
pei "northof the sea". Continuingon due west one gets to Hai-hsi
"west of the sea". Continuingon due south and crossing one gets
to the city of Wu-ch'ih-san.Going all the way round the sea, one
must still always cross over a great sea (the Adriatic?), and only
after six days does one arrive at this country (Ta-Ch'in or its
capital). ( Wei-lüeh)
From the city of An-ku on the border of An-hsi (Parthia) one trav-
els by boat directlyacross to Hai-hsi "west of the sea". ( Wei-lüeh)

We must stress here how vital it is to produce exact literal trans-


lations. Chavannes did not translate the Wei-lüeh section on Ta-
Ch'in, relyingon Hirth'sinferiorversion, which at times deliberately
gives a mistranslationto bolster his eroneous idée fixe. Shiratori's
English (and Japanese) version is far closer to the original.
If our translation and interpretationis correct, we have a route to
Rome from Antioch via Asia Minor overland (presumably that via
Lü-fen),which thereaftercontinues to Wu-ch'ih-san;a second route
directlyacross the sea fromAntioch (which may be that across the
sea fromCyprus). A third(or is it the first?)seems to go across the
Adriatic to Italy. These two or three routes all go from Antioch to
Rome.
We are inclined to follow Hirth and Hudson when they take the
frightening journey about which Kan Ying was warned by the Parth-
ian boatmen in Charax as that around the Arabian coast and through
the Red Sea, thence by land to Alexandria, This route from Charax
(Characene, T'iao-chih) is not mentioned in the Wei-lüeh,which con-
centrates on the routes fromAn-ku,but it is supported by the state-
ment in several texts that Ta-Ch'in trades by sea with India and Par-
thia.
JAH29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
CHINESEKNOWLEDGE
OF THE ROMANEMPIRE 73

However, the attempt by both Hirth and Hudson29 to take the


route via Hai-pei "northof the sea" as this same southern sea route
is totally unacceptable. That it refers to a northern land route is
clear. We find, in fact, this same route north of the Mediterranean
mentioned in the Hou-Han-shu, Hou-Han-chi, a furtherpassage in
the Wei-lüeh,and probably in the Wei-shu.

Routes to China

Prior to the 1st century C.E., Chinese travellers had gotten no


furtherthan Wu-i-shan-li(Arachosia in Afghanistan).
However, the route that Kan Ying took in 97 C.E. to get to T'iao-
chih (Characene) was via Chi-pin (Gandhâra) and Wu-i-shan-li(Ara-
chosia), then almost certainly across the Kerman desert direct to
Charax.30
The Hou-Han-shu also gives a route via A-man and Ssu-pin to Yü-
lo, and thence by sea to the south one gains communication with
Ta-Ch'in.
There is no consensus as to the identificationof these places. We
favour Armenia to Sophene to somewhere north of the Mediterran-
ean, thence to Rome by sea.31 In any case, this route seems to be an
overland route, not via the Indian Ocean. The Chinese travellers,

29 Hirth, pp.43, 185,Hudson,p. 84.


30 Thisroutewas emphasizedalso byKungChün,1942,pp. 14-15. The actual
routethatKanYingtookis onlygivenin theHHC. However,it is supported
by severaltexts,especiallytheHS, whichgivetheidenticalsequenceas a
generalgeographical guide.OnepassageintheHHS has KanYinggoingfrom
T'iao-chih(Characene)acrossAn-hsi(Parthia)and thenlookingout to Ta-
Ch'in.Thisis unlikely.The suggestionofTingCh'ienand severalothersthat
T'iao-chihis Antioch, and thatKan Yingwas lookingout (fromthere)over
theMediterranean, is notconvincing,forKan Yingcould nothave been in
AntiochwithParthiansailors,norcouldhe havedescribedAntioch(An-ku)
as thefrontier ofParthiaifhe had actuallyreachedthere.
31 Thisis supported byTasakaKôdô.The alternative viewofHirthand several
othersis thatA-manis Ecbatana,andSsu-pinis Ctesiphon. Ifso, Yü-lomight
feasibly be DuraEuropos.Wemustreject,in anycase, Hirth's viewthatYü-
lo is in theEuphrates/Tigris to theWLtext,which
delta,quitecontradictory
clearlyputsit further north.
JAH
29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
74 D. D. LESLIEAND K.J.H. GARDINER

Chang Ch'ien and Kan Ying,went overland to Parthia and Mesopota-


mia, not by sea.
The Hou-Han-shu mentions trade of the Roman Empire by sea
with Parthia and India, and also has envoys, or more likely mer-
chants, from the king of Ta-Ch'in, An-tun, in 166 C.E. coming to
China via Vietnam, at approximately the same time as the firstIn-
dian embassies that came by sea. It also has Ta-Ch'in coi'jurors/
magicians sent to China from Shan, a state in the Burma region, in
120 C.E.
We can conclude fromthis that Chinese did not use the sea route
fromChina, but travelled only overland, whereas Romans did get to
China by sea from the 2nd century C.E. They also used the land
route as Ptolemy stated.
Turningnow to the Wei-lüeh,we have the following strikingpass-
age.
Not only is there a route to Ta-Ch'in communicating fromHai-pei
"north of the sea", by land. But there is also one coming south
following the sea, which communicates with the north of the Yi
tribes outside the seven commanderies of Chiao-chih (Tongking).
There is also a water (sea) route communicatingwith Yung-ch'ang
in I-chou, that is why Yung-ch'angproduces exotica. Former gen-
erations only mentioned a sea route, they did not know of a land
route.
There seems no doubt but that this passage is referringto routes
between the Roman Empire and China. The route to Vietnam is
probably linked to the 166 An-tun mission, that to Yung-ch'ang in
Ytinnanto the 120 Shan tribute of Ta-Ch'in coi'jurors.
We do have some problem with the last part of the passage. One
explanation of the apparent contradictionis that Kan Ying only knew
of the hazardous sea route via the Red Sea; he did not know that
you could get to Rome via Asia Minor.
The Chinese sources give less informationabout the land route
from China in the far north, their description of the routes stop-
ping in Shu-lo (Kashgar). However, the Wei-lüeh does describe the
country of the Aorsi, taken over by the Alans, and most signifi-
cantly states that this Aorsi/Alan country is contiguous with Ta-
Ch'in. It does not give a northernroute to Rome, but we can say
that Ta-Ch'in extended much furthernorth than Syria or even Asia
JAH29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
CHINESEKNOWLEDGE
OF THE ROMANEMPIRE 75

Minor. Ferguson states that rhubarb came to the west from China
via the Aorsi.32
What we do know fromthe Chinese sources is that the sea route
became importantonly afterthe land route got blocked around 150
C.E. Embassies fromT'ien-chu (earlier called Shen-tu,the Indus area
in north-westernIndia) had got to China by land, presumably via
Arachosia, around 100 C.E., but in 159 and 161 we have embassies
coming by sea, at almost the same date as the An-tunenvoys. There-
after, Chinese sources record merchants and envoys coming to
China from the Romans by sea from the south in 226 (or perhaps
23333) and in 285.34
There is also a key statement:
Merchants of Ta-Ch'in oftenvisit Fu-nan (Thailand or Cambodia?),
Jih-nan (Annam) and Chiao-chih (Tonking), but few from these
countries have been to Ta-Ch'in.35
Westernsources, Pliny,the Periplus and Ptolemy,are all aware of
both sea and land connections with China, but are equally vague
when they come to the details.36 The routes in the Wei-lüeh leading
to Rome are, we suggest, more convincing than those given for the
China area in our western classics.

32 Ferguson, p. 590.See also Hudson,pp.94,96.


33 This is recordedonlyin theLiang-shu54 (l-c 48). Thereis no reasonto
doubtitsactuality, except thatthedate226 is suspect.
34 This embassyis mentionedin the Chin-shu3, 97, and othersources,in-
cludingtheNan-fangts'ao-mu-chuang to the3rdcentury,
, attributed which
also mentions plantsand spicesreachingCantonfromTa-Ch'in.
35 LS 54 (l-c 48). We have otherreferences to Ta-Ch'inin passagespreserved
intheencyclopaedias fromno longerextantworkslinkedto South-east Asia
byK'angT'ai and others.One exampleis themention ofa race ofsmallmen
southof Ta-Ch'inwho,menacedby cranes,seek aid fromTa-Ch'in.Thisis
surelybased on thelegendofthecranesand pygmiesfoundin Homerand
Aristotle,and severallaterclassicalauthors.
36 Mostvaluableforthe Greekand Latinsourcesare Coedès, Schoff,Yule/
Cordier, Raschkeand Janvier.
JAH
29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
76 D. D. LESLIEAND K.J.H. GARDINER

C ' A
w ' X/
I> ^ ' X O§^ /2 '
// c
VL s
/

.3 ' -

f
y#

£ «
._2 /y/ -c
"
"II 4jf i
IC] I

Ï
-S
!§C]

I
©
W "O
g
C
8
3
i
1
Ì

I sì
-
IP
í-£
Si
.jé
i

JAH
29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
CHINESEKNOWLEDGE
OF THE ROMANEMPIRE 77

cs* nn
E co nn
.c 5
w8 co 3 OSH ^ ? > ¿®
¿8 LU ULI

I jo
je
I 0
II 3

^ IP
Ii s «"
£j -
^ f' i
ii 5
£ 11 ^? s fa ' f' c ?s
il
£ ¿I
> II .*-// •= *:O)
2 <g
« -ļ/ož 3
œ" £<2 fr S.^
£T-- § ,f - Xo
i H? - X ^
«5 2* I I X - Is^l
lili
- o <*-z> Ss5
o>- fr
/o/I £ ^ I o I TÇH a.
® E®
«0«J «O 1I «g
- "Č
Un -(0
"V-* ?(O5
5 .2
OT
ļ£
ü3 c
©
¿SiP ¿I
8-£82
=-Sč á I I í " -fi8«
"■
.§g
>£ ¿SS8
Slil
SoSz ¡ ž*lf
ř
eviCOT* I S ,ļ ,^2J
!* r õtít

VÍ 5 I
> -o 5
o>S
c<o35 o>
»- ^
&
u £ -§
8
l__, < _

' J*
ti
/
Jí <ïo> v
/ -
' -
II ITjļ^i L Uī
-<0 § E-c
sì -Í2g,£ r si
^^ t:
« ¿ īsi
^£ <0* S *!
I 1±w
-J 08E
_Q

JAH
29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
78 D. D. LESLIEANDK. J.H. GARDINER

c. Hou-Han-shu
- 3600li-
Ssu-pin - 3400li- An-hsi
A-man

(river)
960li
Yü-Io
(bysea)
Ta-Ch'in

d. Wei-lüeh
1
(toTa-Ch'in) - Lü-fen
230//bridge Hai-pei

2000li
SW Ts'e-san An-ku
' in
(island (inAn-hsi)
sea)

'SW
//
Hai-hsi Ta-Ch'in

Wu-Ch'ih-san

e. Weih-lüeh
2
Ssu-lo Ssu-pin A-man
/
Yü-Io Ssu-t'ao
/
(river)
340li (river)
SW >
/ /
Ssu-fu 600li Ch'ieh-lan 3000li
/
600li
SW
/
Hsien-tu

Chi-shih
I
Sea (Chang-hai) JAH
29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
CHINESEKNOWLEDGE
OF THE ROMANEMPIRE 79

References

ListofAbbreviations
BMFEA BulletinoftheMuseumofFar EasternAntiquities
CS Chin-shu
E&W East and West
HHC Hou-Han-chi
HHS Hou-Han-shu
HS Han-shu
JA JournalAsiatique
1-c lieh-chuan
LS Liang-shu
MS Monumenta Serica
PS Pei-shu
SC Shih-chi
SKC San-kuo-chih
TKHC Tung-kuan han-chi
TP T'oungPao
WL Wei-lüeh
WS Wei-shu

Bielenstein,H., "TheRestoration oftheHan Dynasty; withProgelomena on the


Historiography oftheHouHanShu",BMFEA26 (1954),esp.pp.9-81.
Bretschneider, E. V.,On theKnowledge PossessedbytheAncientChineseofthe
Arabs, (London:1871).
ChangHsing-lang, Chung-hsichiao-ťungshih-liaohui-p'ien,(Peking:1930),
6 vols.,esp.vol. 1.
Chang,K.(Kuei-sheng), "ChineseGreatExplorers: TheirEffect uponChineseGeo-
graphicKnowledge priorto 1600",Ph.D. thesis,(Michigan:1955).
Chavannes,E., "Les paysd'Occidentd'aprèsle WeiLio",TP 6 (1905),pp. 519-
571.
Chavannes,E., "Les pays d'Occidentd'aprèsle Heou Han Chou",TP 8 (1907),
pp. 149-234.
Coedès,G., Textesd'auteursgrecset latinsrelatifsà l'ExtrêmeOrient , (Paris:
1910),(Reprinted N.Y.,1977).
Colless,B. E., "HanandTa-Ch'in:China'sAncientRelationswiththeWest", Pro-
ceedingsof theFirst New Zealand InternationalConference on Chinese
Studies, 1972,pp. 56-66.
Daffinà,P. "TheReturnoftheDead",E & W22 (1972),pp.87-92.
Demiéville, P.,ReviewofR. A. Stein,Le Lin-yi.. ., in TP 40 (1950/1951),
p. 351.
Duff,J.D., (trans.),Lucan: TheCivil War; 1928.
Ferguson,J. (MiltonKeynes),"Chinaand Rome",pp.581-603 in Aufstiegund
Niedergangder RömischenWeltII (Principát),vol.9.2, ed. H. Temporini,
(Berlin/N. Y.: 1976).
FitzGerald, C. P.,China, A ShortCulturalHistory , (London:1935).
JAH
29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
80 D. D. LESLIEAND K.J.H. GARDINER

Forke,A., "Ta-ts'indas römischeReich",OstasiatischeZeitschrift 14 (n.s. 4)


(1927),pp.48-60.
Franke,O., Geschichte des chinesischen Reiches, 5 vols.,(Berlin/Leipzig), esp.
vol. 1, (1930),vol.3, (1937).
Gernet, J.,Le Mondechinois, (Paris:1972)(English,Cambridge: 1982).
Herrmann, A., "Ta-Ch'inoder das Chinades FernenWestens:eine historisch-
geographische Untersuchung", MS 6 (1941),pp.212-272.
Hirth,F., China and the Roman Orient , (Shanghai/H. K: 1885). (Reprinted
Chicago,1975).
Hudson,G. F.,Europeand China, (Boston:1931).
Janvier, Y.,"Romeet l'Orientlointain:Le problèmedes Sères.Re-examen d'une
questionde géographie antique",KTEMA9 (1984),Strasbourg, pp. 261-303.
KungChiin,"Lian-Hanyü Luo-ma-ti ssu mao-ik'ao", Wen-shih tsa-chih2, 5/6
(June1942),pp. 13-26.
Laufer, B.,Sino-Iranica, (Chicago:1919),(Reprinted Taipei:1973).
Leslie,D. D., Gardiner, K.J.H., "ChineseKnowledgeofWestern Asia duringthe
Han",TP 68 (1982),pp.254-308.
Lu Pi,San-kuo-chih chi-chieĶ(Peking:1957),(Reprinted Shanghai:1962).
McCrindle, J.W.,(trans.),AncientIndia (as describedbyPtolemy}, (Faridabad:
1885).
Maspero,H.,"Untextetaoïstesurl'OrientRomain", pp.93-108 ofhisMélanges
Posthumes , vol.3, ÉtudesHistoriques , (Paris:1950)(Originally from1937).
MiyazakiIchisada,"Jôshito Taishinto Saikai",pp. 154-184ofhisAjia-shiken-
kyû, Kyoto,1957,vol. 1 (fromShirin24 (1939),pp. 55-86).
Miyazaki, 1., lhe ťour Penods on theHistoryof Intercourse betweentheEast
andtheWest",pp.97-98 inInternational Symposiumon HistoryofEastern
and Western CulturalContacts(.1957, Tokyo , Kyoto),(Tokyo:1959).
Needham, J.,Science and Civilizationin China, 7 vols.,(Cambridge: 1954)etc.,
esp. vol. 1.
Pelliot,P.,"Notesurles anciensitinéraires chinoisdansl'Orientromain", JAser.
11,no. 17 (1921),pp. 139-145.
Pelliot,P.,Noteson MarcoPolo, 3 vols.,(Paris:1959,1963,1973).
Petech,L.,Northern India According totheShui-ching-chu , (Rome:1950).
Rackham,H.,NaturalHistoryofPliny, (London),10vols.,esp. vol.II, book 6,
(1947).
Raschke,M. G.,"NewStudiesinRomanCommerce withtheEast",pp.604-1233
inAufstieg undNiedergang derRömischenWelt , II (Principát),ed. H. Tempo-
rini,vol.9.2,(Berlin/New York:1976).
Schafer, E. H., TheGoldenPeachesofSamarkand, (Berkeleyand Los Angeles:
1963).
Schoff,W H., The Periplus of the ErythraeanSea, (London,Bombayand
Calcutta:1912).
ShiratoriKurakichi, theseriesofEnglishtranslations inMemoirsoftheResearch
Department oftheToyoBunko15(1956),pp. 1-23,25- 72,73-163,165-329.
Tazaka(Tasaka) Kôdô,Chugokuni okeruKaikyôno denraitosonogutsû, (To-
kyo:1964),2 vols.,esp. pp. 71-80.
JAH
29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions
CHINESEKNOWLEDGE
OF THE ROMANEMPIRE 81

TingCh'ien,citedin Lu Pi.
WangHsien-ch'ien, et al., Hou-Han-shuchi-chieh , (Changsha:1915).
Wieger,L., TextesHistoriques , 2 vols.,(Sienhsien:1903),esp. vol. 1, pp.843-
846,883-886.
Wolters,O. W, Early IndonesianCommerce , (Ithaca and London:1967),(Re-
printed1974).
Yule,H., ed. H. Cordier,Cathayand theWayThither, ; 4 vols.,(London:1913-
1916),(Reprinted Taipei,1972).

List of Chinese characters

A-man [föjg* Nan-fangts'ao- ^


An-hsi £,g. mu-chuang
An-ku panCh'ao Jf[@
An-tu PanYung m
An-tun Po-ssu ®)Ut
ChangCh'ien shan ^
Chi-pin Jgf® Shen-tu MB
Chiao-chih 3cfct Shu-lo In.¥/J
Ch'ieh-lan JLK Ssu"fu
Ch'ih-san Äife Ssu-lo Iff
S
Fu-lin Ssu-lung SPÍ
Fu-nan Ssu-pin W*
Hai-hsi Ssu-ťao SP§
Hai-pei UX Ta-Ch'in
Hai-tung řSm Tsm-lan SS8
Hou-Han-chi -SšiiCŽĒ T'iao-chih
Hsien-tu fřlř Tien-chu
Hu S3 Tou"le ^
I-chou Ts'e-san
Jih-nan BM Tung-kuanhan-chiMHI9ÍIB
KanYing Wei-lüeh §tB§
K'angT'ai Wu-ch'ih-san ÄSISi
Li-kan etc. Wu-i-shan-li M lilSS
Láeh-ch'uan Wu-tan
Lü-fen ISíâ" ^
Lü-hsi WŠ Yung-ch'ang ^<1
Meng-ch'i IRtäf Yü-lo ŤS

JAH
29/1
(1995)

This content downloaded from 138.251.14.35 on Mon, 28 Dec 2015 14:13:07 UTC
All use subject to JSTOR Terms and Conditions

You might also like