Tuesday, May 14, 2013

When the Iron Bird Flies: A summary of findings

 When the Iron Bird Flies: 
The 1956-1962 Secret War on Tibetan Plateau

A summary of findings

Jianglin Li’s study of the suppression of the Tibetan resistance to Chinese occupation 1956-62 and imposition of Maoist “Democratic Reform” has recently been published in Chinese language. It makes a significant contribution to the still little-known history of the conflict, particularly through analysis of statistics gleaned from official publications. The following is an English language summary of the key findings, in advance of a full translation.

Satisfactory confirmation of detail for this period of Tibet’s history (and indeed China’s history) is notoriously difficult, due to official secrecy and the virtual non-existence of reliable non-official documentation. The figures assessed here, though incomplete, thus provide crucial indicators of the scale of the PLA’s engagement in Tibet at that time, quite sufficient to justify the author’s characterization of it as a war on a largely unarmed population.

Although global estimates remain elusive, the study shows from official figures that something in the order of 10% of the total Tibetan population was involved  - killed, wounded, captured - in military operations during these years. In Qinghai province, where some of the more detailed statistics were discovered, 30% of the total population was involved in conflict over an 18 month period in 1958-59. The sources speak of “indiscriminate killings and executions”, and confirm that, in addition, an extraordinarily high proportion of the civilian population, up to 20%, was arrested to prevent the spread of popular resistance, and that many tens of thousands died as a result, in areas for which clear data is available.
Most of the sources, and some of the background to their interpretation, are also touched on in the following summary. -----Matthew Akester

1.    Time period covered:  
From February 1956 (Sethar, Ganzi) to October 1962 when Qinghai Military Command announced the end of their Third Stage battles.[1]

Success of “suppressing the rebellion” was marked by realization of “Three Completes”, namely, “rebels were completely annihilated, weapon confiscation completed and confiscation of counter-revolutionary certificates [i.e., letters of appointment issued by the GMD] completed”. In addition to these, no small groups of resistance forces consisting of ten or more persons were left in battle areas.  By this standard, the length of the war differed. Based on today’s administrative regions:

·         Sichuan Province: February 1956 – December 1961.
·         Qinghai Province: April 1958 (Xunhua Massacre) – October 1962 (completion of the Third Stage).
·         Central Tibet: July 1956 (Jiangda, Mangkhang etc.) – March 1962 (Tibet Military Command announced the realization of “three completes”).
·         Yunnan Province (Diqing Tibetan Autonomous Prefecture): April 1964 – March 1960.
·         Gansu Province  (Gannan Tibetan Autonomous Prefecture): Two waves of battles:
A.   June 1956
B.   March – November 1958.

2.    Actual number of battles (military actions):
Number of battles was calculated in different ways.  Sichuan Military Gazetteer counts each military engagement and gives the total number as “over 10,000 big and small battles”.
 In Central Tibet, PLA launched 12 military campaigns from March 1959 to November 1961, each consisting of many battles. For instance, the campaign referred to as The Second Stage Campaign in Chamdo in August to November 1959 consisted of 840 battles.  No total number of battles can be found in Central Tibet, but it would be no less than a couple of thousand.
3639 battles were fought in Qinghai Province.
 The 11th Infantry were responsible for military actions against nomads in Gannan and participated in a number of military campaigns in Central Tibet. They reported a total number of 996 battles fought in Gannan and Central Tibet.
The incomplete statistics shows that during the 6.5-year war, no less than 15,000 battles were fought.

3.    Tibetan population directly involved in battles
People “involved in battles” refers to the number of Tibetans killed, wounded, captured and surrendered in battles. It also includes women, children, elderly etc. caught up in the battles.  Chinese sources refer to women/children rounded up in battles as “rescued masses carried off by rebels”.

The following set of statistics was compiled from various Chinese sources.[2]  The numbers are incomplete and can only be used as reference.

A.   Qinghai:  “Over 127,000 rebels were annihilated”.
B.   Sichuan:  Over 90,800 persons.[3]
C.   Central Tibet: Over 93,000 persons.[4]
D.   Gannan Tibetan Autonomous Prefecture: over 22,400 persons.[5]
E.   Diqing Tibetan Autonomous Prefecture (Yunnan): over 13,700 persons.[6]

Based on the above, the total number of Tibetans directly involved on the battleground was around 347,000 persons. Two points should be noticed: (1) This number does not include those who were caught up in the battles but successfully escaped to neighboring countries before 1962; (2) this number does not include the male population put into jail for the purpose of “preventing rebellion”.

According to official government statistics, Tibetan population in 1953 was  2,775,622, and in 1964 it was 2,504,628.[7] Central Tibet did not participate in the 1953 census.  The population of Central Tibet in 1953 was estimated as c.1 million.  Using  2,770,000 as the population base, the number of people directly engaged in the battles was about 12% of the total population in 1953. Again, this is only a rough indicator.

Local statistics may be more accurate. For example, a 1959 Qinghai report gave the following statistics:

“Based on information compiled by Qinghai Military Command, by the end of December last year (1958), 623 battles were fought, 60,864 rebels were annihilated (among them 10,415 were killed, 2,648 wounded, 21,958 captured, 25,843 surrendered)…by the end of June this year (1959), 850 battles had been fought, 18,189 rebels were annihilated (among them 2,209 were killed, 939 wounded, 7,806 captured, 7,235 surrendered)...26,810 women and children liberated.”[8] This means from April 1959 to the end of June 1959, 105,862 people (including women and children) were directly engaged in battles.

In another report sent by CCP Qinghai Provincial Committee to CCP Central Committee and Central Military Committee dated October 15, 1959, gave the following numbers:

“…from April last year to September this year, a total of 1,969 battles were fought, 121,752 rebels were annihilated (including those captured more than once and those who fled from neighboring provinces), 34,281 rifles and pistols, and 24,247 muskets were captured.[9] In addition, 41,865 counter-revolutionaries and evil doers of all kinds were arrested and 76,258 women and children were liberated…”[10]

 If we just consider the numbers given by Qinghai Military Command, adding the 1958 figure (60,864) and 1959 figure (18,189), plus the “liberated” women and children (76,258), we have a total of 155,311 people directly involved in the battles in the period of one year and half. According to government population statistics, Tibetan population of Qinghai was 513,415 in 1957. This means, from April 1958 to September 1959,  at least  30.3% of the total Tibetan population in Qinghai was involved in the war.

4.    Majority of the “rebels” “annihilated” in battles were unarmed.

Just a few examples:

A.   In Central Tibet, 93,093 “rebels” were “annihilated” from March 1959 to March 1962, and 35,523 guns of all types were captured in battles.  This means only 38% of them had firearms.  In other words, 62% of the “rebels” were likely to have been women, children, elderly and monks.
B.   In the Battle of Mitika (Aug. 27 – Sept. 15, 1959), 63% of the people involved were unarmed, apparently women and children.
C.   In the Battle of Ngari (June – July, 1960), 59% of people involved were women and children.
D.   From July to December 1958,  Yushu Command for Suppressing Rebellion fought 148 battles in Yushu area (Qinghai), “annihilated” 90,557 enemies and captured 20,387 firearms. Only 22.5% of those “annihilated” had guns of any kind.

Some Chinese sources admit to “indiscriminate killing of escaping masses and executing prisoners”.[11]

5.    Mass imprisonment of males between 18 and 60 years of age.

Sources show that in Qinghai and Gannan, all the men captured and surrendered in the battles were put into jail. In order to “prevent rebellion”, men between 18 and 60 were mass arrested and put into jail. This practice started in Qinghai and Gannan, and was later adopted in other provinces as well. Quotas were given to local party committees and administrative bodies, but exceeding quotas was normal. In Yushu Tibetan Autonomous Prefecture, the quota was 2,000 to 2,500 arrests, but more than 22,780, or 14.3% of the 1957 total population of Yushu, people were arrested. 

The following statistics are based on sources I was able to find.[12]

A.   Qinghai: 59,183.[13] Figures by prefectures:
Yushu: 22,780
Hainan: 16,272
Guoluo: 9,262
Huangnan: 8,506
Haibei: 2,363

B.   Gannan and Tianzhu Tibetan Autonomous County: 14,183. In Gannan Tibetan Autonomous Prefecture, the total number of people arrested between March 1958 to June 1961 was up to 8.6% of total population in 1958.
C.   Ngaba, Ganzi and Muli Tibetan Autonomous County: c. 14,600, between 1956 and 1958.[14]
D.   Central Tibet: No actual number given. According to one source, Zhang Jingwu gave an order to local cadres that those arrested should not exceed 2% of total population. There were three figures given for the total population of Central Tibet in 1959:  1 million, 1.12 million, 1.18 million. Even by the lowest figure, the quota would be 20,000.
E.   Yunnan: Unknown.

I believe the total number of people arrested from 1956 to 1962 was at least 125,000.

Again, local figures are more accurate. For example:

Jiuzhi County, Guoluo Prefecture: 1,249 people, or 13% of total population, were arrested in 1958.

Maqin County, Guoluo Prefecture: 1,844 people, or 18.47% of total nomad population, were arrested in 1958.

Qumalai County, Yushu Prefecture: 21% of the total population was arrested.

            Death in jail:
 Only figures for Qinghai are available: “ …44,556 people were wrongly arrested and sentenced, up to 84% of total arrested and sentenced. 23,260 people died due to wrongful arrest and wrongful sentencing or in collective training.” [15]

In Document  55 [1982] of  CCP Qinghai Provincial Committee, the figure was higher: “In 1958, 85,285 people were handled (dealt with?), among them 64,347 were from pastoral areas, up to 8% of the total population. 899 were killed by mistake. 17,277 people died in detention and collective training.”

6.    PLA Military power used in the war

I use two measures for the term military power: (1) actual number of soldiers and officers involved; (2) cumulative figures, as the same units fought in different areas.

According to various Chinese military sources, the following army units were involved in the war: [16]

A.   Infantry: 8 divisions, about 100,000 people.[17]
B.   Air force: 3 divisions, 2 independent regiments.[18]
C.   Calvary: at least three divisions.[19]
D. Special units: armour, chemical warfare, motorcycle, demolition, signals, etc.
E. Logistic units: 4 truck transportation regiments, engineer corps, field hospitals, army stations, supply stations, animal hospitals, gas stations, etc.

Although actual numbers of people in different branches of the armed forces is not the same, from the above list I estimate that the number of combat troops was no less than 150,000.

On the local level:
A.   Sichuan (March 1956 – Dec. 1961): altogether over 80,000.
B.   Gansu (March – Dec.1958): over 25,000.
C.   Qinghai (April 1959 – October 1962): over 70,000.
D.   Central Tibet (March 1959 – 1961): over 60,000.
Total: Over 235,000.[20]

Besides PLA units, a large number of local militia participated, with many armed units directly involved in actual combat. The numbers of militia I was able to find in Sichuan, Gansu, Yunnan and Qinghai add up to over 71,000 people.

There were also a considerable number of civilian laborers drafted for various war-related purposes, for instance, to transport supplies, carry wounded soldiers, handle pack animals and so on. During the 6 years of war, no less than 143,000 civilian laborers were drafted. 

It is safe to say that during the six year war on the Tibetan plateau, the number of PLA combat troops, logistical units, militia and civilian laborers involved was no less than 300,000.

7.    Casualties on both sides:
It is very difficult to find the total number of Tibetans killed in combat. So far the only number known for sure is for Qinghai, where 16,600 were killed, according to Qinghai Military Gazetteer. 

PLA casualties: 10,934 (4,748 died, 5,223 wounded).[21]

8.    Tibetan population loss between 1956-1964
 Local statistics:
A.    Gansu:[22]
1957: 255,947
1959: 188,050
1961: 174,581
From 1958 – 1961, population loss was 81,366 or 31.8% of the 1957 population.

B.     Sichuan: [23]
1958: 686,234
1964: 605,537
Population loss was 80,697, or 11.8% of the 1958 population.

C.     Qinghai:[24]
1957: 513,415
1964: 422,662
Population loss was 90,753, or 17.7% of the 1957 population.

D.    Yunnan:[25]
1953: 64,611
1964: 61,827
Population loss was 2,784.

E.     Central Tibet: No reliable statistics.

The above statistics add up to a population loss of at least 255,600, without counting Central Tibet.  Those numbers have some problems: (1) Major battles in Ngaba and Ganzi happened in 1956-1957, therefore the 1958 population statistics do not count those who died in the first wave of military action; (2) Population statistics for Yunnan cover a 10 year span, but a more accurate number would be from 1955 to 1961; (3) Analysis of population statistics in Yushu and Guoluo from 1957 to 1961 shows that the population loss in these two prefectures was 118,172.
In Xinghai County of Hainan Tibetan Autonomous Prefecture, where severe military actions took place in 1958 and 1959, nomad population decreased from 14,050 in 1957 to 10,169 in 1960, showing a population loss of 3,881, or 38.2%. The population loss in Yushu, Guoluo and Xinghai Counties alone added up to 120,053, more than 90,753 (Cited from Tibetan Population in Qinghai). 

Adding the 120,053 to population loss statistics in other areas, the total number comes to over 287,000. Since this number does not include population loss in Central Tibet, and the Qinghai number came from two prefectures and one county only, it is safe to say that Tibetan population loss due to the war, famine, mass detention, etc. was no less than 300,000.

Population loss in certain areas was very high. For instance:

A.    Yushu Tibetan Autonomous Prefecture:
1957: 159,419
1963: 93,483
Population loss was 65,936 or 41.3%. It took 20 years for the population to return to the 1957 level.

B.     Guoluo Tibetan Autonomous Prefecture:
1953: 99,628[26]
1964: 56,936
Population loss was 48,753, or 48.9% of the 1953 population. In normal conditions, the population should show some increase from 1953 to 1957, but no accurate statistics for 1957 can be found.

C.     Chengduo County of Yushu Tibetan Autonomous Prefecture:
1957: 14,476
1960: 10226
Population loss was 29%.

D. Sethar County, Ganzi Tibetan Autonomous Prefecture:
1956: 30,600
1957: 24,785
1962: 17,641[27]
Population loss from 1956 to 1962 was 13,898, or 45.5% of the total 1956 population.

[1] Resistance to “democratic reform” first started in Tibetan areas of today’s Diqing Tibetan Prefecture, Yunnan Province in 1955. Small scale battles were reported in the second half of 1955. I didn’t find enough information in Chinese sources to figure out the course of events in 1955. It seems to have been confined to local areas.
[2] Chinese sources include: military gazetteers, county gazetteers, biographies and memoirs of army commanders, internal government documents, etc.
[3] Total number in Sichuan Military Gazetteer is over 140,000 persons. This number includes an unknown number of Yi people. I checked the numbers in Ganzi Military Gazetteer, Aba Prefecture Gazetteer and Muli County Gazetteer and added the “annihilation” statistics in each battle listed there, coming up with this number.
[4] Two Chinese sources come up with this number. In order to confirm this number, I analyzed detailed information about the 12 military campaigns in Central Tibet, and the “annihilation” numbers in these campaigns added up quite close to this number. However, about half of those involved in these campaigns were refugees from Amdo and Kham.
[5] Calculated from History of the 11th Infantry Regiment.  This number does not include the Hui people.
[6] Calculated from Yunnan Military Gazetteer.
[7] Nationality Statistics of China 1949 – 1990, p. 41. It is also worth noticing that according to this book, Tibetans were the only nationality that suffered a major population decrease from 1953 to 1964.
[8] Cited from “On the Current Situation Against the Enemies and Opinions on A Few Policy-related Principals”, a report sent by Xue Keming(19101965), Director of Public Security Bureau, Qinghai Province, to Qinghai Provincial Committee of CCP dated July 28, 1959. 
[9] A crudely made firearm used by farmers and nomads at that time.
[10] Cited “A Preliminary Summary on the Struggle of Suppressing Rebellion in Pastoral Area”, a report written by CCP Qinghai Provincial Committee. 
[11] History of National Defense, p. 131.
[12]  These statistics are mostly based on the rehabilitation figures from the early 1980s. Many of the sources only have the number of people rehabilitated but no figure of actual arrests, so the actual number of arrests must be higher.
[13] This number does not include Haixi Mongolian and Tibetan Autonomous Prefecture and Haidong Area. Adding the figure in Haixi, the total number of arrests would be 71,158.
[14] This number was calculated from the 1981 rehabilitation statistics.
[15] Cited from a report sent by CCP Provincial Committee of Qinghai to Central Committee and Deng Xiaoping, dated March 19, 1981.  
[16] Main force came from three military regions: Kunming (Yunnan), Lanzhou (Gansu) , Chengdu (Sichuan) and Tibet. Other military regions involved were Xinjiang, Inner Mongolia,  Shenyang and Jinan.
[17] Infantry Divisions 4, 11, 42, 61, 62, 55, 130 and 134.   Additional local military units and a few independent regiments are not included.
[18] Air Force Divisions 13, 23, 25; Independent Air Force Regiments 4 and 5. The 4th Independent Air Force Regiment was the Tupolev-4 Regiment responsible for bombing monasteries in Lithang, Bathang and Chatreng.
[19] First Cavalry Division, Independent Cavalry Regiments 1, 2, 3 under the command of Lanzhou Military Region; Cavalry Regiments 13, 14 under the command of Inner Mongolian Military Region; Infantry and Cavalry Brigade under the command of Lanzhou Military Region, and cavalry detachments of local military divisions in Yushu, Guoluo, Huangnan, Hainan, Haixi and Qaidam.
[20] This is a cumulative number as the same unit fought in different areas. For instance, The 11th Infantry Division fought both in  Gannan and in Central Tibet.
[21] Gannan Military Gazetteer does not give the number of PLA soldiers wounded, so this is an incomplete number.
[22] Cited from Tibetan Population in Gansu, pp. 34-35.
[23] Cited from Tibetan Population in Sichuan, p. 24.
[24] Cited from Tibetan Population in Qinghai, p.17.
[25] Cited from Gazetteer of Diqing Tibetan Autonomous Prefecture, p.221.
[26] Population figures for Guoluo in 1953 were changed twice. According to Population in Qinghai, a 1987 internal publication, population of Guoluo was 100,343. This is an “altered” number ordered by Qinghai Public Security Bureau and other government offices. In Gazetteer of Guoluo Tibetan Autonomous Prefecture published in 2001, the 1953 population was 56,936, nearly half the figure given in the 1987 book. The 1964 population was the same in both books. In 1952, Guoluo Working Group, together with a Cavalry Detachment, altogether 715 people, entered and were stationed in Guoluo. Subtracting 715 from 100343, the Guoluo Tibetan population in 1953 should be 99,628.
[27] One xiang was switched from Ngaba to Sethar in 1962, adding 939 people to Sethar County. Subtracting the number of people added by this change, population in Sethar in 1962 was 16702. This number may include Han cadres and soldiers. Cited from Gazetteer of Seda County, pp. 6, 94.

1 comment:

  1. I have a few comments on units involved in Tibet between 1956 and 1962:

    1. In the beginning (1956 to 1958) many police units were active. Those units had the same function the Armed Police has today and are often not mentioned in PLA records.

    2. For some reason units belonging to the Tibet Military District are not mentioned. Those are for example the 52nd, 53rd and 54th infantry division. Those were the main units involved in the war in central Tibet.

    3. The 11th infantry division stayed in central Tibet until 1979 and took part in the war with India in 1962 as well as later border conflicts and the Cultural Revolution.

    4. Up to a third of “Chinese” losses in this war were ethnic Tibetans. At least one regiment consisting of Tibetans took part in the campaign. Fully trained and armed PLA soldiers were far more difficult to kill than some local Tibetan communist government officials. Realistic reporting should include such facts.

    5. The war in central Tibet against Tibetans was declared finished in March 1962 to prepare for the war against India in October. Preparations for the new war took the entire summer and troops were in position to strike by early October.

    6. Many sources like footnote 5 on "History of the 11th Infantry Regiment" should be corrected to "11th divison" and dated. For example the "Ganzi Military Gazetteer" has an internal early edition, a censored new Edition and an additional internal volume with additional source material. If just
    "Ganzi Military Gazetteer" is mentioned it's difficult to trace your source.