The Cambridge History of China. Vol. 12: Republican China, 1912-1949, Part 1
Подождите немного. Документ загружается.
THE VIEW FROM
THE
PROVINCES
293
Warlord
cliques
The leading warlords commonly belonged
to
factions
or
cliques, groups
held together
by a
presumed identity
of
interests, much
as the
political
factions were held together
(see
chapter
5). But
cohesion among warlord
groups varied from loose affiliation
to
tightly structured unity.
The
loosest cliques were essentially alliances concluded
for the
advantage
envisaged
by
each participant.
But
bonds
of
personal association
and
obligation also played
a
role, particularly
in the
stronger cliques.
The
key relationship
was
that between each clique member
and the
clique
leader; personal, lateral ties among clique members could
be
weak
or
non-existent.
The
personal bonds between
the
members
and the
leader
were exactly those already discussed
as
supplementing
the
cohesion
of
warlord armies: family ties, teacher-student
and
patron-protege relations,
common provincial
or
local origins, friendship, academy
or
school ties.
Ch'i Hsi-sheng
has
ranked these personal associations from strongest
(father-son)
to
weakest (schoolmate),
and
with that ranking
in
mind
has
examined
the
Anhwei, Chihli
and
Fengtien cliques,
the
three leading
groups.'
7
The
Fengtien Clique
was the
simplest
and
strongest
in
internal
organization,
for
virtually every member
was
bound
to the
clique leader
by ties that
Ch'i
judged
to be
strong.
The
Chihli Clique,
on the
other
hand,
had an
extremely complex structure.
It
involved
a
larger number
of
militarists,
and a
larger number
of
ties,
but
most
of
them
of
the sort
Ch'i
ranked
as
weak. Although
Ch'i
suggests that
the
very multiplicity
of the
ties reinforced them,
and
made
for
strong integration,
the
clique appears
to have come much closer
to an
alliance
of
near equals than that
of
Feng-
tien.
The
Anhwei Clique
was
weaker than
the
other
two
because
the
commanders with most military strength were bound
to the
leader
by
the weakest ties.
The
Kwangsi Clique, which
Ch'i did not
analyse,
dif-
fered from
the
three already mentioned
in
that
it
was,
for
most
of
the time,
confined
to a
single province;
it was
essentially
a
single political-military
organization
in
Kwangsi, with three leaders informally assuming
dif-
ferent leadership roles,
and
surprisingly, maintaining
an
extraordinarily
16 The information about warlord revenues is culled from a wide range of sources: mono-
graphs on individual warlords, reports from diplomats and journalists, newspaper and
periodical accounts. Some of the most interesting and revealing material is in 'The reminis-
cences of Chang Fa-k'uei, as told to Julie Lien-ying How', and Tong and Li, The
memoirs
of Li
Tsung-jen.
A concise and knowledgeable treatment of warlord revenues is Ch'i Hsi-
sheng, Warlord
politics
in
China,
150-78. Jerome Ch'en discusses the subject in
The
military-
gentry
coalition:
China under
the
warlords, 130-45.
Wou,
Militarism
in
modern China, 55-80,
examines in detail the sources of Wu P'ei-fu's finances. Chang Yu-i,
Chung-kuo
chin-tai
nung-yeh shih tzu-liao
(Source materials on China's modern agricultural history) emphasizes
the range, diversity and extortionate nature of the warlords' quest for money.
17 Ch'i,
Warlord politics
in
China, 56-76.
Cambridge Histories Online © Cambridge University Press, 2008
294 THE
WARLORD
ERA, I 9 I 6-2 8
high degree of cohesion. A brief examination of the origins of the cliques
will explain how they came to show such differences.
The Chihli and Anhwei cliques were based on relationships established
among officers in the Peiyang Army under Yuan Shih-k'ai. Yuan culti-
vated various kinds of personal obligations among his subordinates to
assure the unity and loyalty of the army. His followers did much the same
with regard to their own subordinates. As long as Yuan lived, those
networks of associations within the Peiyang Army were subordinated to
the overall pattern of loyalty to Yuan. When he died, Peiyang officers
had to adjust to a totally new situation. That adjustment required several
years,
as each officer made decisions about whose lead he would follow,
what his geographic and military circumstances dictated, what his per-
sonal inclinations and aspirations were, how his personal interests would
best be served, and how he judged the national political situation.
We have already noted (see chapter 5, page 272) how Tuan Ch'i-jui
gradually built up the Anhwei Clique (Tuan was from Anhwei) and a
political club to promote his interests, the Anfu Club. Officers not among
Tuan's followers, and those who foresaw the day when they would be-
come targets of his unification policy, were naturally hostile to Tuan.
Moreover, Tuan's insistence on putting his own proteges in top positions
left those who were passed over resentful. From that background of fear
and resentment emerged an opposition group who looked for leadership
to the only other Peiyang officer of Tuan's stature who also held high
position in the government, Feng Kuo-chang.
Feng Kuo-chang had served in top-level Peiyang commands, as military
governor of Chihli and, later, of Kiangsu, and became vice-president of
the republic in 1916. In 1917 he became acting president, a post he would
not assume until he was assured that his own proteges would retain con-
trol of three important Yangtze provinces. It is not completely clear why
Feng should have come to head a group hostile to Tuan. Perhaps one
reason was that Sun Yat-sen had set up a separate regime in the south,
and Feng and Tuan disagreed about the proper policy for eliminating it
and reunifying the country. Feng advocated negotiation, Tuan wanted
to use force.
Feng retired from the presidency in 1918, but those who opposed Tuan
continued to look upon him as leader. Because Feng was from Chihli,
this developing group of militarists and politicians came to be known as
the Chihli clique. At the end of 1919 Feng died, and there was nobody in
a clear position to succeed him as head of the group. As Tuan Ch'i-jui
attempted more vigorously than ever to establish his military and polit-
ical superiority and so unify the nation, Ts'ao K'un, military governor of
Cambridge Histories Online © Cambridge University Press, 2008
THE VIEW FROM THE PROVINCES 295
the metropolitan province, increasingly felt threatened by the Anhwei
group. His relations with Tuan steadily cooled. Because of Ts'ao's leading
role in the 1920 war that ultimately forced Tuan out of power, Ts'ao K'un
finally emerged as the acknowledged leader of the Chihli Clique.'
8
The Fengtien Clique was created by Chang Tso-lin. Chang was one of
the many militarists who rose from low estate. Born into a Manchurian
peasant family, he started out as an enlisted soldier, and later organized a
local defence force that became part of
the
regular Manchu military forces.
As he went up the military ladder, he took the subordinates and associates
of his youth with him, all bound by close personal ties. Chang supported
the Ch'ing authorities during the 1911 Revolution, and when the dust
had settled held the second spot among the Fengtien military. When the
military governor retired in 1915, Chang used his strong local connections
and his military following to make life impossible for the officer appointed
by Peking. Finally, in April 1916, he was acknowledged as head of the
Fengtien government.
Once in firm control of Fengtien, Chang moved to establish similar
authority over the other two Manchurian provinces, using a combination
of military threat and political influence. In 1917, Peking dismissed the
Heilungkiang governor for supporting the Manchu restoration, and
Chang's ready military force assured that the dismissal was effective.
His influence in Peking obtained the appointment of his own choice as
head of the province, and from that time Heilungkiang was run by men
approved by Chang. The sequence of events in Kirin was similar, but
Chang did not have the province fully under his subordinates until 1919.
These three provinces were the core territory of the Fengtien Clique
because of Chang's solid control over them, and, of course, because of
their great wealth and geographic position. Later Chang extended his
authority into North China, and other warlords became part of the Feng-
tien clique, though never as strongly bound to Chang as the militarists
of Manchuria.
1
'
The Kwangsi Clique remained for most of the time within a single
province. Li Tsung-jen headed the clique, but it is more accurate to think
of him as first among equals. Pai Ch'ung-hsi and Huang Shao-hung were
the other two members of the clique at the outset, until Huang Hsu-ch'u
replaced Huang Shao-hung. The clique emerged from the loose alliance
18 Andrew J. Nathan,
Peking
politics 1918-192):
factionalism
and
the
failure of
constitutionalism,
128-75,
2
3
2
~9-
Odoric Wou, 'A Chinese "Warlord" faction: The Chihli Clique, 1918-
1924',
in Andrew Cordier, ed.
Columbia essays
in
international affairs
vol.
Ill,
the
Dean's papers,
ta6j, 249-74. See also Wou, Militarism in
modem
China.
19 Gavan McCormack,
Chang Tso-lin
in Northeast
China
1911-1028:
China,
Japan, and
the
Man-
churian
idea.
Cambridge Histories Online © Cambridge University Press, 2008
296 THE
WARLORD
ERA, I 9 I 6-2 8
established
by Li, Pai and
Huang
in the
early 1920s, aimed
at
unifying
the highly fragmented province.
By 1924
they
had
eliminated
so
many
petty Kwangsi warlords that they faced
the
choice
of
fighting
one
another
or making their cooperation permanent. Since they came from similar
backgrounds
and had
developed
a
strong mutual respect, they decided
to
work together.
20
Factions grew
up
within
the
warlord cliques.
The
Chihli Clique
was
split into
two
factions,
one
under
Wu
P'ei-fu,
one
under Ts'ao
K'un.
Ts'ao's faction,
in
turn,
was
divided
in
two. These factions wrangled over
who should hold what office
and
control what revenues. Factions also
existed
in the
Fengtien Clique, especially after Chang reorganized
his
army following
a
serious defeat
in 1922. He
advanced younger officers
with modern military training,
and yet
continued
to
look
for
support
to
his
old
associates. This produced
'new' and 'old'
factions
in
Fengtien.
21
Wars
There were literally hundreds of armed conflicts, short
and
long,
on
local,
regional
and
national scales. Many wars were waged
for
control
of an
administrative area, such
as a
province
or
county. Others were fought
over control
of
local
and
regional economic networks that transcended
administrative areas.
For
example, opium from Yunnan
and
Kweichow
was shipped over
a
well-established trade route
to
Western Hunan, where
it could
be
sent northward
to the
Yangtze Valley
or
southward
to the
Canton delta.
The
allegiance
of the
warlord controlling western Hunan
determined which route would
be
selected, enriching
the
warlords
of the
Yangtze Valley
or
those
in
Kwangtung. Western Hunan
was a
prize
because of its position
on the
trade route,
and it was
important
to
leaders
in
two
different regions, thus directly
and
indirectly
a
source
of
conflict.
Similar marketing networks existed throughout China
and
undoubtedly
gave rise
to
warfare. Winston Hsieh
has
begun
to
open
up
this subject.
22
The more
or
less formal wars between major cliques attracted most
at-
tention because they determined
who
controlled
the
national govern-
ment
in
Peking,
the
symbol
of
legitimacy. When
one
clique gave promise
of becoming
so
powerful
as to
restrain
the
other militarists,
and
create
20 Lary,
Region
and
nation.
21 Nathan, Peking politics, discusses factions and his appendix provides profiles of seven
important ones. For the leading officers in the military cliques, with their units see Wen
Kung-chih, Tsui-chin
san-shih-nien Chung-kuo chiin-shih
shih (Military history
of
China
for the
past thirty years), passim,
but
especially
vol. 1, pt 2.
22 Winston
W.
Hsieh,
"The
economics
of
warlordism',
Chinese
Republican Studies Newsletter
T.I
(Oct. 1975)
15-21.
Cambridge Histories Online © Cambridge University Press, 2008
THE VIEW FROM THE PROVINCES 297
a genuine centralized control,
the
other leading warlords temporarily
combined their strength
to
bring
it
down. Thus,
in
1920 the Chihli and
Fengtien Cliques cooperated to eliminate the power of the Anhwei Clique
in the Peking government, and transferred most
of
its provinces
to the
victors.
In
1922 Fengtien allied with the remnants
of
Anhwei power,
as
well as with forces
in
South China,
to
try
to
overthrow Chihli.
It
failed
but was saved from destruction because it had a strong, rich base that was
removed from the cockpit
of
struggle
in
North China; Chihli was not
prepared
to
invade Manchuria,
so
the Fengtien forces could
go
home,
reorganize and prepare
to
try again.
In
1924 Fengtien, again
in
alliance
with Anhwei adherents in North China, and with southern forces as well,
fought the Chihli Clique
a
second time, and successfully engineered the
defection
of
one
of
the leading Chihli generals, Feng Yii-hsiang. Chihli
militarists continued
to
control provinces
in
Central China, and
in
1926
they allied with Chang Tso-lin
to
fight Feng Yii-hsiang, finally pushing
his forces far to the west.
2
' That left Chang Tso-lin as the major power in
Peking when
a
new kind
of
armed force,
the
National Revolutionary
Army, began its march northward to eliminate the warlords once and for
all.
The accompanying maps show the general changes in clique distribu-
tion of power as
a
result of these wars.
24
Scholars have analysed these wars
in
terms
of
the balance
of
power,
sometimes using
a
model from international relations.
2
' Certainly
the
recurring coalitions
of
warlords against potential unifiers exemplifies the
central process
of
the balance
of
power concept.
In
China, however,
it
was
a
highly unstable system. Warlord cliques did not seek balance as an
end
in
itself;
each sought hegemony until
its
imminent success moved
the others
to
opposition. Moreover,
it
was
a
terminal system, and every-
body knew
it;
each warlord took
it for
granted that warlordism would
one day cease and the country be reunited. Each seems to have accepted
reunification
as
inevitable, even desirable,
but
wanted
it to
take place
without limiting his personal power. That was a contradiction which each
major warlord hoped
to
resolve
by
effecting
the
reunification
himself.
Warlords who could not reasonably hope
to
head
a
national unification
simply wanted
to
postpone
it
and make hay
in
the meantime. Warlords
tended to think in very short term; they rarely thought about what things
might be like
in
five
or
10 years, but, rather, worked
to
reap next year's
taxes this year.
23 See Wen,
Tsui-chin san-shih-nien Chung-kuo chiin-shih
shih, vol. 2 for
a
straightforward account
of the military aspects
of
these wars.
24 See maps pages
298-301.
25 Ch'i, Warlord politics in China, 201-39.
Cambridge Histories Online © Cambridge University Press, 2008
298
THE
WARLORD
ERA,
I916-28
MONGOLIA
I
A M
Fengtien Clique
Anhwei Clique
Chihli Clique
•
• .
500 km
1
1
ioOmiles
MAP
9.
Distribution
of
warlord power
on the eve of the
Anhwei-
Chihli
War,
1920
USSR
:::: HEILUNGKIANG
:
PfllM;
- -
I I I I I I I I
-H-M
1 1
CHAHAR
|
\-
::::::
'li§i:::::l::
if JEHOL
.XTTtTlttTltTy.W> ,
V , J ^~
rfffFENGTIEN
: /
. SU.YUAN
(dL 1
/f^llr^
I'
f
^•
s
^
en
^^rf|
H
'
H
||
(|,}
|
^' -4.ru-?
SHANSI
jgjKr
SHANTUNG
/
u
V:SHENSI
.•$MBBB&mk^mm^
"v^^L-
^gMi^^lKv
ANHWEI
i:
>,,llllMHIIIlll|fc>-> : •
-^Nanking
JAPANl^.
r
^^^^^^^^^f^^j^
FUKIEN
FRENCH
INDO-CHINA
Cambridge Histories Online © Cambridge University Press, 2008
THE VIEW FROM
THE
PROVINCES
MAP
IO.
Distribution
of
warlord power
on
the eve
of
the
First
Chihli-
Fengtien
War, 1922
US
S R J,
111111111111111
n
1111 ::::::::
HEjLUNGKIANG^
::::::::
M
0 N G 0 L I A
n
aa
*
BBt
*f^v^J'«*i*">a
•~'
iaa||
""
aB
<
|BH
•••r^H'.'
.;•
MiiiiiaJ
•>
Hiiitiiiiit >ai naaaaaaaaaaaaaa- •••••••••iiiiw
«BBH
• •••••• •••••••••• a'
.lilk
•••••••••••••"llllill'tiiiit*\
aaa
|
BE
• ••••"'"•••••••••••'•*<"l'*
ll
"
l
"
lla
'
illmiiiillilllL^B
M
••••••••••••••
,•••••••»••••••••in«•••••••••••§ffH m
WinHMillMinimmmm
•••••••••'
II
•••••••••lllH
il
,f,
fITn7T'TrV««i««'
••••••••»
•••••••••••k*
aaaaaaaaaa
l
lniHumimi^irnnmnniiin
.
•*-• lalll
CIB^^H
••••••••••••••r.aiaaiVT7*7alBBi
BBaBBI,BBIBBaBBBBBBBBk
• VA.IiTflB
k.j
iii.
»••••••
niiilfNYMliiii'
•iiiiiiiiiiiiiiiiiifOKi^K<'
••••••••
iiniair
•
•••••••- .JIIIIIIIIIIMHIIIIIV^HH
• • '
IIIII»I->
IIIIII
ilium
imiUiuAaaTaaal
•MIMIIIMIMIIIM
II
•••UK
j
'
TT'TW/lCm^fcBBl
•
••••a- niiaia aaiiiiiiaaiii^aaaaaaadUl'UtrLillir
Afagfaf^^^V^H
uai>
iiiiaiiii^imiiiiHii
•••••••iiiiiiiiipf^H
^Kfl
Tllli
•••••••in
••••••••••••JiiiaiiMiaaiaauuM
ijlliia
i•••••••
4BBf^aBaaiBiflBjaf*^^H
^^^aflH
flaaiaiPtiiaiiai •aiuMiai •it^
B
ai
;aaaaaarfJM
•FlM
Mr
j i - rr
•
•.^'•TBTBT
• • • • •
••M
!«•
tfjjyy * •
w*r S
Vr
'^
aB
a
BfAW
r
•
•Vic ********ii«
JAPAN
0
^
A
|YUNNAN
/A
«r—
'W^F
v
.^J
^t^^L^f^
X
St//i
Yat-sen
and
u
I.
FRENCH
Ch'en
Chiung-ming
<-• INDO-CHINA
f \j
S
I A M ^
II11111 Fengtien Clique
liiiiiiiiil
Anhwei Clique
HUH Chihli Clique
O 50Okm
"
i| I
I
—J
6 ' ' 300 miles
Cambridge Histories Online © Cambridge University Press, 2008
300
THE WARLORD ERA, 1916-28
^H^^^^i^^l
«f FRENCH
' "), INDO-CHINA
MAP
11. Distribution of warlord power on the eve of the Second
Chihli-Fengtien War, 1924
Cambridge Histories Online © Cambridge University Press, 2008
THE
VIEW
FROM THE PROVINCES
301
Feng Yii-hsiang (Kuominchiin Clique)
ChangTso-lin (FengtienClique)
Sun Ch'uan-fang
0 500 km
300 miles
MAP
12.
Distribution
of
warlord
power
at the
beginning
of 1926
U o a " •»
-HEILUNGKIANG
:
- 1111111111111111111111 -
MONGOLIA
'^Z^^CM
^o^v\\II11111
[Illl'iiPS.
• ••• KIRIN •::
SZECHWAM
1
lyJHPBl'^liO^^
0 (fragmented)
•
^<lii^jHBa£5iBB *>888y
4
YUNNAN &
KWANQ8J
^ C^^S^
y.
FRENCH
»J ^jjT
Kuommtang
)^
INDO-CHINA
/ \£J
° \ \ m
s i
A M
i
JAPAN^
Cambridge Histories Online © Cambridge University Press, 2008
302 THE
WARLORD
ERA, I 9 I 6-2 8
NOTE FOR MAPS 9 TO I 2
These maps are based on biographies, government documents and announce-
ments, chronologies, reports of foreign observers, and warlord studies. Ch'i,
Warlord
politics
in
China,
210, 212, presents similar maps of conditions before the
Anhwei-Chihli War and before the First Chihli-Fengtien War.
These maps try to show the fragmented state of China in the early 1920s,
and how the pattern of fragmentation changed, but they give an utterly false
sense of precision and certainty. They are inaccurate in several ways: (1) Clique
affiliations largely follow the affiliation of the provincial tuchun, ignoring the
existence of minor warlords, who often controlled important areas. (2) The
maps do not show areas under contention, or where authority was non-existent
or unclear. For
example,
maps
9
and 10 show Fukien as under the Anhwei Clique.
During those years, Li Hou-chi was tuchun of Fukien, and he had strong
ties with Tuan Ch'i-jui. Yet the southern part of the province was sometimes
under Cantonese militarists, sometimes under rival northern commanders,
and Li Hou-chi's authority was limited at the best of
times.
Shensi is shown in
the Chihli camp on the eve of the Second Chihli-Fengtien War, yet in fact so
many petty warlords contended for power in that province that it might just
as well be labelled 'fragmented'. (3) The maps do not distinguish between strong
clique affiliation and weak, questionable, or changing affiliation. For example,
map 9 shows Honan in the Chihli camp. Yet Chao T'i, tuchun of Honan from
the beginning of the warlord era, only moved towards some affiliation with
the Chihli warlords when he thought Tuan Ch'i-jui was planning to replace
him. Map 11 shows Shantung under Anhwei influence, although Chihli power
in North China was then at its height. The reason is that T'ien Chung-yii, a
Chihli adherent who had been tuchun since 1919, was relieved of his position
in 1923, and succeeded by Cheng Shih-ch'i, who, all sources agree, leaned
towards the Anhwei Clique; in 1923, that was not tantamount to being hostile
to Wu P'ei-fu and Ts'ao K'un. (4) There is no correlation between the size of
a clique's territory, as shown on the map, and its real power. For example,
control of Suiyuan, Chahar, and Jehol makes for an impressive expanse of
territory, but it was not particularly significant militarily, because those areas
were poor, sparsely populated, and far from the main lines of communication.
(5) Changes in the successive maps were not all due to the defeat or victory of
this or that clique in the major wars to which the maps relate. For example,
Li Hou-chi was ousted from Fukien in the late summer of 1922, and the map
sequence inevitably implies that this was a consequence of the First Chihli-
Fengtien War. Yet in fact he was ousted by Kuomintang troops from Kwang-
tung; that fact is not even suggested by the maps, because within a few months
the Kuomintang forces had departed and Sun Ch'uan-fang was at the head of
the province.
Cambridge Histories Online © Cambridge University Press, 2008