IAN TALBOT

The 1947 Partition of India and Migration: A Comparative

Study of Punjab and Bengal
in
RICHARD BESSEL AND CLAUDIA B. HAAKE (eds.), Removing Peoples.
Forced Removal in the Modern World (Oxford: Oxford University Press, 2009)
pp. 321–347

ISBN: 978 0 199 56195 7

The following PDF is published under a Creative Commons CC BY-NC-ND licence. Anyone
may freely read, download, distribute, and make the work available to the public in printed
or electronic form provided that appropriate credit is given. However, no commercial use is
allowed and the work may not be altered or transformed, or serve as the basis for a
derivative work. The publication rights for this volume have formally reverted from Oxford
University Press to the German Historical Institute London. All reasonable effort has been
made to contact any further copyright holders in this volume. Any objections to this material
being published online under open access should be addressed to the German Historical
Institute London.

DOI:
 13
The 1947 Partition of India
and Migration:
A Comparative Study of Punjab
and Bengal
IAN TALBOT

The 1947 massacres and migrations were for many years little
more than footnotes in the study of the achievement of India's
and Pakistan's Independence. Since the 1980s, however, histori-
ans have increasingly focused on them. The work of such schol-
ars as Urvashi Butalia, Kamla Bhasin, and Ritu Menon enabled
a gendered dimension to be brought to the Partition experience. 1
They have addressed the sensitive issues of the large-scale abduc-
tion of women and their recovery and, in some instances, forced
repatriation. Accounts of the violence which sparked off the
greatest refugee migration of the twentieth century have begun
to see it as more than a 'temporary madness'. Anders Hansen,
Paul Brass, and Ian Talbot have shown that it was organized,
possessed a genocidal element, and was the result not merely of
the collapse, but occasionally the involvement of local systems of
civil and police administration. 2
Work on the Punjab region of India has been at the forefront
of the 'new history' of Partition. This reflects the fact that it was
at the epicentre of violence. In less than three months, over 8
million Punjabis in chaotic and often brutalizing circumstances
undertook a reverse migration across the new international
boundary which divided the region. Accounts of this vast human
1 Urvashi Butalia, 1he Other Side ef Silence: Voices from the Partition ef India (New Delhi,
1998); Ritu Menon and Kamla Bhasin, Borders and Boundaries: Women in India's Partition
(New Delhi, 1998).
2 Anders Bjorn Hansen, Partition and Genocide: Manifestation ef Vwlence in Punjab ,93;,947

(New Delhi, 2002); Paul Brass, 'The Partition of India and Retributive Genocide in the
Punjab 1946-1947: Means, Methods and Purposes',Journal efGenocide Research, 5/i (2003),
71-w1; Ian Talbot, 'The 1947 Partition of the Punjab', in id. (ed.), 7he Deadly Embrace:
Religion, Violence and Politics in India and Pakistmt (Karachi, 2006).
322 IAN TALBOT

tragedy have drawn on a variety of documentary accounts

produced in the aftermath of Partition, on oral testimonies, and
fictional representations. These together with the dramatic inten-
sity of the migration experience have resulted in a Punjab-
centred model of Partition emerging. The iconic images of the
vast refugee columns, trains packed to the rooftop, and of bloody
communal massacres have all been drawn from the Punjab. Such
a standardized account ignores not only the variety of Partition
experiences elsewhere in north India, but also homogenizes what
was a highly differentiated pattern of violence, migration, and
resettlement within the Punjab itsel(
The paucity of comparative studies of Partition is a striking
element in the evolving literature. The vastness of such a subject
matter, and language and visa difficulties for Indian and Pakistani
scholars have all been cited by way of explanation. Community
and nation-building considerations have also obstructed such an
enterprise. These tend to privilege the suffering, victimhood, and
the ability to 'bounce back' of particular groups. Official histories
trumpet the role of the state in dealing with the 'unprecedented
refugee problem'. These types of historical discourse would not be
served by a comparative approach.
This essay aims to examine migration in the two regions of the
subcontinent, namely the Punjab and Bengal, that were most
affected by Partition. The new international boundaries bisected
both these Muslim-majority provinces of British India. The
reasons why they were divided and the demarcation of their
boundaries lie beyond the scope of this work, although they are
addressed in an earlier comparative volume, Regi,on and Partition:
Bengal, Punjab and the Partition qf the Subcontinent, which I edited in
1999 with Gurharpal Singh. 3 This essay possesses three main aims:
first, to delineate the migratory processes in the Punjab and Bengal
regions; secondly, to seek to explain them and to reveal how they
affected the contrasting histories of resettlement and rehabilitation
in the regions; and thirdly, to examine the different histories of
migration constructed by Punjabi and Bengali migrants. Before
turning to these themes, however, it is necessary to remind
ourselves of the wider picture of Partition-related migration flows
and to consider the variety of migration experience.
3 Ian Talbot and Gurharpal Singh (eds.), Region and Partition: Bengal, Pwyab and the

Partition qf the Subcontinent (Karachi, 1999).

 The 1947 Partition oflndia

The 1947 Partition efIndia and Migration

British officials and Indian politicians were alike unprepared for

the mass migrations that accompanied the division of the subcon-
tinent. There had been warning signs from the Great Calcutta
Killings of August 1946 that a new phase of communal violence
had emerged which involved elements of ethnic cleansing. The
outbreak of violence in the Punjab itself in March 1947, which
started in Lahore and Amritsar but became known as the
Rawalpindi Massacres after the region worst affected, rendered
around 40,000 people homeless. Nevertheless, a transfer of popu-
lation was unexpected in August 1947.
Politicians reassured minority populations that they would be
safe and encouraged them to stay in their ancestral towns and
villages. Within days of Independence, however, it became clear
that a chaotic two-way flight was under way across the plains of
the Punjab. This stemmed from outbreaks of violence that the
impotent Punjab Boundary Force was unable to stem in the
twelve districts in which it was deployed. The level of casualties
remains controversial. Figures vary from the low estimate of
200,000 by the British civil servant Penderel Moon,4 to that of 2
million. The Indian judge G. D. Khosla put the figure at 500,000
with an equal number of Muslim and non-Muslim casualties. 5
The MQM in its publications for reasons of community assertion
maintains that as many as 2 million mohajirs (Muslim refugees
from India) died. Such writers as Patrick French have adopted a
median figure of a million casualties. 6 In the absence of verifiable
figures, Gyanendra Pandey has correctly pointed out that the
historical discourse on the killings 'continues to bear the stamp of
rumour'. 7
The tragedy unfolding in the Punjab dominated national and
international headlines, but other regions were also affected.
Muslims fled from UP, Delhi, Bihar, and Bombay to Pakistan. In
Delhi, according to some unofficial estimates, violence in
'1 Penderel Moon, Divide and Qyit (new edn., New Delhi, 1998), 293.
5 Gopal Das Khosla, Stem Reckoning: A Survl!)' of Events Leading up to and Following the
Partition of India (2nd edn., New Delhi, 1999), 299.
6 Patrick French, Liberty or Death: India's Journey to Independence and Division (London,

1998), 349·
7 Gyanendra Pandey, Remembering Partition: Violence, Nationalism and History in India
(Cambridge, 2001), 91.
 IAN TALBOT

September claimed about ro,ooo Muslim lives. 8 Muslims were

driven to take sanctuary in refugee camps atJama Masjid, Purana
Qla, and Humayan's Tomb. The latter camp was still bulging
with over 30,000 refugees in December 1947. 9 Shahid Ahmad, the
publisher and progressive writer, has provided a harrowing
account of its desperate conditions in his autobiographical work
Di!Jzi ki Bipta. 10 Some 300,000 Muslims, two-thirds of the commu-
nity's total population, eventually abandoned India's capital. A
comparison of the 1941 and 1951 census reveals the dramatic
demographic transformation. Muslims comprised 40.5 per cent of
the population in 1941 with Hindus in a majority of 53.2 per cent.
A decade later, Hindus made up 82.1 per cent of the population
and Muslims a mere 6.6 per cent. Delhi's population increased by
r.1 million in the period 1941-51. This unprecedented growth of
106 per cent largely resulted from the influx of Partition
migrants. 11 These were members of the Hindu Khatri and Arora
commercial castes of the West Punjab. Most were drawn by the
economic opportunities afforded by India's new capital. Some
refugees already possessed professional, commercial, and kinship
ties in the city which had a growing Punjabi community from the
late nineteenth century.
Muslims also migrated from UP, Bihar, West Bengal, and
Assam to the eastern wing of their new homeland. East
Bengal/East Pakistan received around a million and a half
refugees. Hindus and Sikhs left not only West Punjab for India,
but also the West Pakistan province of Sindh and the North West
Frontier Province. There was also, as we shall see later, an
ongoing Hindu migration from East Bengal that continued for
decades after the movement of population in Puajab had ceased.
The figures for migration are only slightly less haphazard and
controversial than those for Partition-related deaths. They vary
from 14 to 18 million. In either case, this eclipses violence-related
migrations as a result of two world wars and the end of empire else-
where in the twentieth century. There was no accurate accounting
8 Ibid. 199.
9 Report of AS. Bhatnagar, Secretary to the Chief Commissioner Delhi, 4 Dec. 1947.
MB1/D276, Mountbatten Papers, University of Southampton.
10 Shahid Ahmad, 'Dilhi ki Bipta', in M. Shirin (ed.), <,ulmat-e-Neem Ro;:,e (Karachi,

1990), 145,.
I I Ravinder Kaur, 'Narratives of Resettlement: Past, Present and Politics among the
1947 Punjabi Migrants in Delhi' (Ph.D. thesis, Roskilde University, 2004), 34-5.
 The 1947 Partition oflndia

in the early period of flight in the Punjab. In West Bengal there

was no enumeration of refugees between March 1958, when
the government officially wound up its rehabilitation work, and
1January 1964. After the latter date, refugees were termed 'new
migrants' to differentiate them from Partition-related migrants.
The most accurate figures are those provided by the Pakistan and
Indian military organizations. They were established in response to
the spiralling violence in the Punjab and oversaw what was a
virtual exchange of population in the region. 12 According to the
figures of the Pakistan Military Evacuation Organization, 4,715,919
Muslims were transported to West Punjab from East Punjab
between 23 August 1947 and May 1948. Its Indian counterpart
produced figures of 3,672,851 making the reverse journey. In both
instances most refugees had moved by December 1947. This
concentrated period of an organized exchange of population
contrasted with the situation elsewhere in India.
Sarah Ansari's work on Sindh, for example, reveals a much
longer timescale. 13 The province possessed a large Hindu and
Sikh minority (a quarter of the population). Nevertheless, it
remained calm in the immediate aftermath of Partition. The first
serious outbreak of communal violence did not occur in Karachi
until January 1948. This claimed around 200 lives and was
accompanied by widespread looting in the city centre. 14
Following the disturbances, 10,000 Hindus crowded into refugee
camps in the city, before their evacuation to India. The violence
was linked with the flood of Muslim refugees into the city. The
Sindh Prime Minister, Muhammad Ayub Khuhro, had been
reluctant to accept refugees who could not be absorbed in West
Punjab. This 'unco-operative stance angered the Centre and was
to be a factor in his eventual downfall in April 1948'. 15 In August
of that year, the Centre declared a state of emergency and reset-
tled an additional 200,000 refugees in Sindh. The figure was even-
tually to increase following the closure of the Punjab camps in
October 1948. 16 There was a further influx of refugees from India
12 For the work of the Indian Military Organization see Brigadier Rajendra Singh, The
Military Evacuation Organisation 1941 48 (New Delhi, 1962).
13 See Sarah Ansari, Life efter Partition: Migration, Communiry and Strife in Sindh 194 1 1962
(Karachi, 2005).
14 Ibid. 56.
15 See Ian Talbot, Pakistan: A Modern History (rev. edn., London, 2006).
16 Ansari, Life efter Partition, 66 and ff.
 IAN TALBOT

following communal disturbances in Uttar Pradesh in 1950. By

April, refugees were arriving at the rate of 3,000 to 4,000 a day. 17
Migration from India was to continue well into the 1950s. The
inflow of 6,683 refugees was so high in July 19y2 that the Pakistani
authorities considered sealing the main border crossing into Sindh
at Khokropar. 18 Eventually around 60 per cent of the refugees
known as mohajirs from Uttar Pradesh resettled in Karachi,
Hyderabad, and some of the smaller towns in the interior. By the
time of the 1951 census, mohqfirsnumbered 616,906 and accounted
for 58 per cent of Karachi's population. 19 The transplantation of
an Urdu-speaking enclave into the sands of Sindh has possessed
profound consequences for Pakistan's politics. 20
The timing of refugee movements thus varied considerably
between regions. There were also differences in experiences of
migration. This has been obscured by the standardized portrayal
drawn from an overly homogenized Punjab model. Recent schol-
arship has revealed a gendered dimension to Partition migration.
Women were not only vulnerable to assault as symbolic upholders
of community honour (izzat); they had to cope in unaccustomed
roles as household heads following the slaughter of their menfolk
and to adapt to close proximity to strangers on trains, in refugee
camps, in queuing for rations, and in making claims for compen-
sation. Work on Hindu fem ale refugees from East Bengal has
shown them not only as victims, but as undergoing processes of
radicalization in the struggle for survival in Calcutta's post-
Independence squatter communities. 21
The outlines of a class-based differentiation in refugee experi-
ences are also emerging from current research. The higher rank
of government servants who had opted for service in the neigh-
bouring country were guaranteed transportation, often by air,
and had accommodation provided for them in their new post-
ings. The British Overseas Airways Corporation transported
28,000 people from Pakistan and 18,000 from India in the period

17 Ibid. 128. 18 Ibid. 131. 19 Ibid. IIO.

20 It should be noted that Karachi received not only Urdu-speaking refugees, but also
Gujarati migrants from Bombay. They came largely from the Khoja and Memon trading
communities.
21 See Jhuma Sanyal, 1vfaking ef a New Space: Refugees in West Bengal (Calcutta, 2003),
183-7; Nilanjana Chatterjee, 'The East Bengal Refugees: A Lesson in Survival', in
Sukanta Chaudhuri (ed.), Calcutta: 17ze Living City, 2 vols. (New Delhi, 1999), ii: 77ze Present
and the Future, 70-8.
 The 194 7 Partition oflndia

15 September to 7 December 1947. This was in addition to the

twice daily service from Lahore to Arnritsar and the daily service
from Delhi to Rawalpindi run by Indian National Airways. 22
Such passengers could look down on the burning villages and
antlike refugee columns traversing the Punjab's killing fields. On
the rare occasions that the elite travellers were inconvenienced it
could reach even Cabinet-level discussion. Nehru noted with
displeasure, for example, an incident early in October when a
flight direct from Peshawar to Delhi had to set down at Lahore
because of slight engine trouble and its 'occupants had been
stripped of all their belongings'. 2 3
The business and political elites were also privileged migrants.
Many had shifted female family members to hill stations and
businesses to 'safe' areas well in advance of Partition. Huge sums
of money flowed out of the future Pakistan areas of West Punjab
and Sindh in the months leading up to Independence. The
leading Muslim newspaper Dawn claimed that in the last days of
June 1947 alone Rs. 6 crores (60 million) had been withdrawn
from West Punjab. 24 New property-dealing businesses sprang up
to exchange Muslim and Hindu properties. English-language
newspapers advertised residences for sale or exchange in 'safe
areas'. Similarly in East Bengal, it was the wealthiest individuals
who migrated first. The bhadralok class was not only politically
aware, but its members frequently had properties and social
connections in Calcutta. Artisans such as drummers and idol-
makers followed their wealthy religious patrons. Landless labour-
ers were the last to leave and did so because of insecurity. They
had no experience of migration, or connections to ease their
plight on arrival in West Bengal.
Wealthy anticipatory migrants did not share in the dangers of
the crossing to a new homeland that faced 'acute' migrants who
were literally fleeing for their lives. Refugees also faced very
different experiences on their arrival. On the whole, the ability to
recover from the trauma of displacement was markedly
improved if an individual possessed previous social and economic
capital.
22 Kaur, 'Narratives of Resettlement', 111 and IT.
23 Extract from Emergency Committee 20th Meeting, 3 Oct. 1947 MB1/D275,
University of Southampton.
24 Raghuvendra Tanwar, Reporting the Partition qf the Punjab 1947: Press, Public and Other

Opinions (New Delhi, 2006), 227.

 IAN TALBOT

From relief camps through to the construction of housing

colonies, the Indian and Pakistan states differentiated between
different classes of refugees. The ability to afford their own food
rations, for example, determined whether refugees in Delhi
would be directed to a life under canvas in the Edward and
Outram Lines of the Kingsway camp, or be accommodated in
concrete barracks at the Hudson and Reeds Lines. 25 Both Indian
and Pakistani satellite towns and refugee housing colonies had
varieties of house-plot sizes, streets, and availability of services to
suit the different classes of refugee. 'The class differences visible
during the population movement', Ravinder Kaur has appositely
declared, 'became further entrenched when permanent housing
projects were undertaken on such a basis. This ensured that
refugees were reinvented in their old class of social stratifica-
tion.'26 Perhaps the clearest example ofthis was the Indian state's
provision of separate colonies and camp accommodation for
Untouchables.
The post-colonial state also upheld traditional gender roles in
its treatment of refugees. Young female orphans were housed
and trained in Lahore, for example, so that they would not
become an economic burden. The state's guardianship role
extended to their arranged early marriage. Like a family patri-
arch, it saw its role as to establish control over female sexuality.
The social stigma attached to widowhood was reflected in the
establishment in Delhi, for example, of a separate refugee colony
for young widows. In Calcutta there was a similar Women's
Camp. 27 But given the gender imbalance amongst East Bengal
refugees, the conditions prevailing in the squatter colonies in
Calcutta, and the dramatic changes in both caste-based and
gender-based economic roles, traditional social mores and family
roles were not as rigidly re-established.

17ze Puryab and Bengal Experiences efMigration

There are significant differences in both the intensity and the

timescale of Partition-related migration between the Bengal and
25 Kaur, 'Narratives of Resettlement', 143.
26 Ibid. 166.
27 Sanyal, Making qf a New Space, , g 1.
 The 1947 Partition oflndia

Punjab regions. Migration in Punjab was highly concentrated in

the period immediately after the British departure. While more
than 8 million Punjabis were uprooted between August and
December 1947, the total number of refugees in Bengal at this
time numbered only around half a million. The greatest migra-
tion in the Bengal region was in 1950, rather than at the time of
Partition. According to official Pakistani sources, by the end of
that March there were 400,000 refugees from West Bengal and
Assam. 28 Within six weeks, the figure had leapt to r.r million. 29
The Hindu influx into West Bengal in 1950 peaked at r.5 million.
Migration in the Bengal region was to continue thereafter when-
ever there were periods of tension in Indo-Pakistan relations, or
communal riots in the region or elsewhere in the subcontinent.
The displacement of population continued after the West Bengal
government officially wound up its rehabilitation work in
March 1958. By 1981 the West Bengal Refugee Rehabilitation
Committee put the number of refugees at around eight million,
or one-sixth of the total population. 30 This was still not the level
of concentration of refugees in Pakistan Punjab, which received
5.3 million refugees accounting for over 25 per cent of the popu-
lation. 31 Nevertheless, because of its existing population density
and weak regional economy West Bengal was much less able to
cope with this burden. Partition had left it as the smallest (34,000
square miles) and most densely populated state in India. Most of
the refugees from East Bengal/Pakistan settled in the Calcutta,
Nadia, and twenty-four Parganas districts of West Bengal.
Calcutta was favoured because of its hoped-for job opportunities.
Migrants also clustered around the border areas of the Nadia
district where there had been some exchange of population.
These three areas were eventually to contain two-thirds of all the
refugees from East Pakistan.
By July 1952 there were over 400,000 refugees in the
Nadia district, where they constituted about 40 per cent of the

28 Civil and Military Gazette (Lahore), 27 Mar. 1950.

29 Ibid. 26 May 1950.
30 Gyanesh Kudaisya, 'Divided Landscapes, Fragmented Identities: East Bengal

Refugees and their Rehabilitation in India 19471 9', in D. A. Low and Howard Brasted
(eds.), Freedom, Trauma, Continuities: Northern /ndui and Independence (New Delhi, 1998), 111.
31 Mohammad Waseem, 'Partition, Migration and Assimilation: A Comparative

Study on Pakistani Punjab', in Ian Talbot and Gurharpal Singh (eds.), Regi,on and Partition:
Bengal, Punjab and the Partition of the Subcontinent (Karachi, 2000), 211.
 IAN TALBOT

population. 32 The 1951 census highlighted Calcutta's demo-

graphic transformation as a result of refugee migration. The
city's population was nearly 20 per cent higher than it had been
just five years earlier. By 1973, the number of refugees was just
under 2 million and represented around two-thirds of West
Bengal's urban refugee population and one-third of the state's
total refugee numbers. 33 One in four of Calcutta's inhabitants
was a refugee. Seven out of every ten migrants from East
Pakistan had found their way there. The absence of arrange-
ments over property exchange and of a population balance
between Muslim evacuees and incomers meant that for all but
the wealthy, accommodation was difficult to obtain in an
already overcrowded city. The household deities in the posses-
sion of almost all East Bengal Hindu families had also become
'refugees', as there was no space for them. Some were anony-
mously abandoned in temples, others found their way to a
'camp' in Upper Chitpur Road Calcutta, where Swami
Satyanand Tirth was deputed to perform puja (worship). 34
In both the Punjab and Bengal, migration occurred in the
context of outbreaks of violence. This differed from 'traditional'
Hindu-Muslim conflict in that it contained dimensions of 'ethnic
cleansing'. Communal riots previously had been about the 'rene-
gotiation of local hierarchies of power'. 35 This is, in essence, what
the common disputes over the routes of religious processions were
about. The resulting clashes have been termed by some scholars
'consensual' in character. 36 The signs of 'ethnic cleansing' are first
evident in the Great Calcutta Killings of 16-20 August 1946. Over
100,000 people were made homeless. They were also present
in the wave of violence which rippled out from Calcutta to Bihar,
where there were high Muslim casualty figures, 37 and to
32 Statesman Week!), (Calcutta), 12July 1952.
33 Pranati Chaudhuri, 'Refugees in \,Vest Bengal: A Study of the Growth and
Distribution of Refugee Settlements within the CMS', Occasional Papers 55, Centre for
Studies in Social Sciences Calcutta (Mar. 1983), table 1, 38.
3 4 Civil and Military Gazette (Lahore), 4June 1950.
35 See Veena Das and Ashis Nandy, 'Violence, Victimhood and the Language of
Silence', in Veena Das (ed.), The Word and the World: Fantasy, ~mbol and Record(New Delhi,
1983), 177-90.
36 Shail Mayaram, 'Speech, Silence and the Making of Partition Violence in Mewat',
in Shahid Amin and Dipesh Chakrabarty (eds.), Subaltern Studies IX: Writings on South Asian
History and Society (New Delhi, 1996).
37 See Report on Disturbances in Bihar and UP (Muslim Information Centre, 1946). PIT

3363, IOL.
 The 194 7 Partition oflndia 33 1

Noakhali, deep in the Ganges-Brahmaputra delta of East Bengal.

With respect to the Noakhali riots, one British observer spoke of a
'determined and organised' Muslim effort to drive out all the
Hindus, who accounted for around a fifth of the total popula-
tion. 38 The Punjab counterpart to this transition in violence was
the Rawalpindi Massacres of March 1947. The massacres paved
the way for the later August violence in that they both created a
Sikh desire for revenge and revealed the ease with which minority
communities could be expelled in the absence of effective law
enforcement. About 40,000 Sikhs had been left homeless. The
August 1947 Punjab violence repeated this uprooting on a huge
scale. The high levels of violence can be variously attributed to the
militarization of the region's population which formed the sword-
arm of India; the cycles of retributive genocide which began with
the March 1947 Rawalpindi Massacres; the collapse or unreliabil-
ity of the police and local administration; the existence of a Sikh
Plan to 'ethnically cleanse' Muslims from East Punjab in order to
carve out a Sikh state; and the involvement of the armed forces of
such Sikh states as Patiala in attacks on minority populations. The
princes themselves denied that they had contacts with the Sikh
war bands (jathas), but high-ranking court, state, and military offi-
cers such as Bir Davinder Singh and Colonel Bhagwan Singh of
Patiala along with the Chief Minister ofJind were widely believed
to have connived at their activities. 39 The fact cannot be denied
that troops from the princely states not only attacked their
Muslim inhabitants and passing refugee trains, but joined in
assaults on neighbouring districts of the former British-adminis-
tered Punjab.
Bengal had witnessed horrific communal violence in 1946 in
Calcutta and Noakhali. The figure of 4,000 deaths was officially
quoted at the time of the Great Calcutta Killings. An English
official maintained that this was 'a new order in communal
rioting'. He described Calcutta as a 'cross between the worst of
London air raids and the Great Plague'. 40 Around ro,ooo people
shifted out of the city. Seven weeks after the Great Calcutta
Killings, violence spread to this south-eastern district of Bengal.
38 Suranjan Das, Communal Riots in Bengal 1905-1947 (Delhi, 1991), 199.
39 Copland in fact cites evidence that the rulers ofKapurthala and Faridkot had direct
contact with the jathas. Ian Copland, 'The Master and the Maharajas: The Sikh Princes
and the East Punjab Massacres of 1947', Modern Asian Studies, 36/ 3 (2002), 639.
40 Das, Communal Riots, 171.
33 2 IAN TALBOT

From Noakhali it fanned out to the Tippera district. In all, 350

villages were affected in the two districts. The disorders were
only finally quelled by the deployment of around 2,500 troops
and police. The minority Hindu populations living in inaccessi-
ble villages were the victims. Fifty thousand people took refuge in
relief camps.
The cycle of violence had largely spent itself in Bengal by the
time of Partition. Crucially, the region did not have the level of
preparation for communal attacks that existed in the Punjab.
Here weapons had been stockpiled and training given to para-
military groups by ex-soldiers. Calcutta's relative quiescence has
been attributed to Gandhi's moral influence. He stayed in a poor
Muslim neighbourhood where he prayed and fasted for peace.
Mountbatten dubbed him a 'one man boundary force'.
Official responses to the 1947 violence impacted on the
patterns of migration in the two regions. As we have already
noted, the violence in Punjab resulted in the two governments
determining on a virtual exchange of population under the
control of the Pakistan and Indian Military Evacuation
Organizations. 41 Despite misgivings about this in both countries,
fears for the security of minorities along with the need to utilize
properties and land evacuated by them to accommodate refugees
determined this policy. There were some hopes that Muslims
who numbered 13.17 lakhs (roo,ooo) could stay in the south-
eastern Ambala division of the Indian Punjab. Eventually all the
Muslims departed for Pakistan.
The Military Evacuation Organizations set up headquarters
on both sides of the new international boundary which now
bisected the Punjab. A joint civilian machinery was also estab-
lished consisting of Liaison Officers. The two Chief Liaison
Officers held the status of Deputy High Commissioners. There
were also district officials who were provided with funds, escorts,
and scarce supplies of petrol to facilitate their work on behalf of
refugees. The Joint Evacuation Plan agreed on 20 October
between the two Military Evacuation Organizations set a
December target for the evacuation of ro million refugees from
both sides of the Punjab. 42 This elaborate machinery brought
some order to the migration process, although not to the extent
41 Singh, The Military Evacuation Organisation.
42 Kaur, 'Narratives of Resettlement', 97.
 The I 94 7 Partition oflndia 333

maintained in official histories. 43 Minority populations had no

choice but to leave their ancestral homes, even when they were
not in imminent threat of attack. The result was the denuding of
Hindus and Sikhs from Pakistan Punjab who had numbered
33.99 lakhs before Partition and 53.85 lakhs Muslims from the
Indian Punjab and surrounding Indian states. Further govern-
ment control over the migration process was exerted by settling
rural refugee populations together in assigned districts. 44 Both
populations and land were, in effect, being exchanged.
The lower levels of violence in Bengal in 1947 encouraged a
reverse policy. When migration did occur, it was seen as a
temporary process. This meant that evacuee property was not
freed up to house refugees. It led the right-wing Hindu leader,
Shyama Prasad Mukherjee, to call for an exchange of popula-
tions or, alternatively, to demand that a third of the territory of
East Pakistan should go to India to be utilized for refugee reset-
tlement. 45
Inter-dominion conferences were held to assure the minorities
in East and West Bengal of their security. The rehabilitation
ministers of India and Pakistan jointly declared in Calcutta in
April 1948 that they were determined 'to take every possible step
to discourage such exodus and to create such conditions as would
check mass exodus in either direction'. 46 Partition left around 4
million Muslims in West Bengal (approximately 17 per cent of the
population) and around 1 r.5 million (approximately 42 per cent)
Hindus in East Pakistan. In contrast with the Punjab's dramatic
demographic transformation, migration in the Bengal region
occurred in a series of waves, rather than one tidal force.
The governments of both East and West Bengal sought to
reassure minority populations. Upper-caste Hindus living in the
eastern wing of Pakistan were, however, disturbed by status
reversal, and were also sensitive to any threats to the honour
(maan) of their female family members. In 1948 twice as many
refugees left East Bengal (around 800,000) as at the time of
Partition. The Muslim populations of West Bengal and Assam

43 Ibid. I01-3.
44 For details of this process with respect to Indian Punjab, see Kirpal Singh, The
Partition efthe Punjab (Patiala, 1989), 181-3.
45 Sanyal, Making ef a New Space, 133.
4 6 Cited in Kudaisya, 'Divided Landscapes', w9.
334 IAN TALBOT

felt less insecure. The serious disturbances early in 1950,

however, resulted in a peak of migration in both Bengals far
greater than had occurred in 1947.
I visited Muladi ... where I found skeletons of dead bodies at some
places. I found dogs and vultures eating corpses on the riverside. I got
the information there that after the wholesale killing of all adult males,
all the young girls were distributed among the ringleaders of the
miscreants. 47
This was not Punjab in August 1947, but East Bengal in February
1950. The violence had started in Dacca, but spread within a few
days to the Tippera, Noakhali, Syhlet, and Barisal districts. The
coastal district of Barisal witnessed severe disturbances. Four
villages were completely burned down. Muladi was one of its
important riverine ports. Most of its Hindu victims died in the
compound of the police station where they had taken shelter.
The officer in charge was later found in possession of large
amounts of looted property. 48 In the Syhlet district, over 200
villages were devastated and 800 Hindu temples desecrated. 49 In
large areas, the repertoire familiar since 1946 was repeated:
forced conversions, dishonouring of women, and attacks on
trains. 50 The latter lasted from II to 14 February and claimed
many victims on the Chittagong Mail. The bodies of at least a
hundred Hindus were buried by the side of the railway line. 51
Similar scenes of violence occurred in vVest Bengal. There
were widespread disturbances in such Muslim localities of
Calcutta as Bagmari, Beliaghatu, and Goolpara. Houses were
looted and burned along with mosques. According to the Civil
and Military Gazette newspaper, nearly 10,000 Muslims from
Chinsurah, Paikpara, Goolpara, and T elnipara had to leave their
homes and take shelter in open fields opposite the Victoria Jute
Mills. Muslims were also attacked in Jalpaiguri town. Two
47 This report was made by the Scheduled Caste former Minister for Law and Labour

in the Pakistan government,Jogendra Nath Manda!. He resigned in the wake of the East
Bengal killings. A. J. Kamra, The Prolonged Partition and its Pogroms: Testimonies on Violence
against Hindus in East Bengal 1946-64 (New Delhi, 2000), 173-4.
48 Ibid. 63.
49 Ibid. 76.

so The trains packed with up to 4,000 refugees had been easy targets in August 1947,
despite the presence of armed escorts. Blood-splattered trains arrived in both India and
Pakistan with whole compartments of butchered corpses.
51 Kamra, The Prolonged Partition, 89.
 The 1947 Partition of India 335

hundred shops were looted in the Muslim-controlled bazaar of

Karimgunj in Assam. 52 Muslim refugees from Karimgunj and
Hailakandi claimed that the police led the looters. 53 Certainly
more resolute law enforcement would have cut short the distur-
bances. Their continuation sparked the largest wave of migration
in the eastern Indian region since 1947.
The Indian and Pakistan Prime Ministers tried to restore some
stability in the celebrated Nehru-Liaquat Pact in April 1950. It
promised equality of citizenship for minority communities and
stated that refugees who returned home by 31 December 1950
would be entitled to the restoration of their houses and land. This
stemmed migration for a time and even led to around 12 lakhs of
refugees returning to East Bengal. 54 Most were able to recover the
houses they had abandoned, but the provincial government set
aside a further Rs. 7 lakhs to cover the cost of their rehabilita-
tion.55 The returning Hindus together with Muslim refugees from
West Bengal meant that the population of East Bengal increased
by nearly 950,000 by the middle of 1951.56 Thereafter, migration
continued to fluctuate and depended on prevailing communal
and Inda-Pakistan relations. It was West Bengal, however, that
received the majority of migrants. The announcement at the
beginning of October 1952 that a passport system would be intro-
duced for cross-border travel encouraged another influx of
Hindus amid fears that future migration would be more diffi-
cult.57 Within a week, upwards of 2,000 people were daily arriv-
ing by train at the Indian border outposts of Bongaon and
Ranaghat. The trains were 'dangerously overcrowded with
passengers riding on footboards and hanging on to iron beams
and rods beneath the carriages'. 58 The authorities opened an
interception camp at Ambagaon less than half a mile from
Bongaon railway station. Many of the 12,000 or so refugees
preferred to sleep on the platform or under railway wagons.
Eventually special trains had to be called in to dispatch them to
Sealdah Station Calcutta which itself soon presented 'a scene of
52 Civil and Military Gazette (Lahore), 22 Mar. 1950.
53 Ibid.
5 ·1 Chaudhuri, 'Refugees in West Bengal', 38.
55 Dawn (Karachi), 18June 1951.
56 Ibid.
57 India introduced passports on 15 Oct. 1952; Pakistan followed suit two days later.
58 Statesman Week[y (Calcutta), 11 Oct. 1952.
 IAN TALBOT

indescribable confusion'. 'A fog of blue smoke' from countless

cooking fires, a correspondent for the Statesman newspaper
recorded, 'hangs over the listless grey brown mass ofhumanity.' 59
Many of the refugees were from poorer, lower-caste Namasudra,
Mahisya, and Sadgop communities in the East Bengal hinterland.
They were eventually dispersed to transit camps on the outskirts
of the city or fended for themselves in the crowded and squalid
squatter colonies which had sprung up in Calcutta. Their exis-
tence attested to refugee self-settlement in the absence of an
adequate government response.
The squatter colonies literally sprang up overnight, when
thatched huts were constructed under the cover of darkness on
vacant land. Their hogla leaves became a refugee symbol. One of
the earliest refugee self-settlements was at Bijoygarh, the site of
American barracks during the Second World War. The pre-1950
developments were eventually to be regularized through the
government's payment of compensation to landowners and the
gifting of deeds to the occupiers. Even before their regulariza-
tion, many colonies in such south-eastern areas as Tollygunj,
Behala, and Javedpur were well managed by committees which
raised subscriptions and labour for the construction of drains,
roads, and water supplies. This was not the case, however, in the
'illegal' colonies that sprang up after 1950 on the west bank of the
Hooghly river between Magra and Uluberia. Even worse off
were the large numbers of agriculturalist refugees consigned to
prolonged residence in refugee camps in Calcutta and elsewhere
in West Bengal. In 1958 the camp population stood at 800,000.
One-third of their inhabitants had spent anything from six to ten
years living in these squalid conditions. 60
Refugee camps in Punjab had been closed as early as 1948. In
both Indian and Pakistan Punjab evacuee property had been
supplemented by building refugee colonies in existing towns and
creating new satellite towns such as Faridabad and Rajpura. The
latter development on the GT Road, fifteen miles west of Ambala,
was built at the cost of Rs. 20 million. It was termed 'one of the
biggest experiments of the Government of India in building a well
planned and simple yet dignified home for refugees'. 61
59 Ibid. 18 Oct. 1952.
6 ° Chatterjee, 'The East Bengal Refugees', 74.
61 Statesman (Calcutta), 28 May 1949.
 The 1947 Partition oflndia 337

The sense that refugees from East Bengal were the main victims
of Partition because of their neglect in contrast to their Punjabi
counterparts is a central element in the Bengali historical
discourse. It was an important factor in the support refugees gave
to the Communist Party of India (CPI), especially in the wake of
the threatened legislation in 1951 to evict 'Persons in Unauthorized
Possession of Land'. The CPI portrayed the Congress administra-
tion of Bidhan Chandra Roy as being more concerned with the
rights of landlords and property speculators than with the distress
of the refugees. Before examining the problems surrounding the
West Bengal government's rehabilitation programmes, we shall
first consider the ways in which the characteristics of migration
impacted on rehabilitation measures.

Refugee Rehabilitation in Puryab and Bengal

The circumstances of the Partition-related migration in East and

West Punjab made it clear from the outset that violence had initi-
ated a permanent demographic transformation. This was not the
case in Bengal. Moreover, the scale of the Punjab crisis facing
both the Indian and Pakistani governments inevitably prioritized
rehabilitation efforts in the region. A number of Bengali writers
have gone further and argued that the Punjab's greater proxim-
ity to the .seats of power in India and Pakistan and the lobbying
of powerful Punjabi bureaucratic and political elites resulted in a
concentration of scarce resources in the region. This discourse
will be examined in the final section of the essay. Suffice it to say
here that oral testimonies from Punjabi refugees on both sides of
the border, along with written complaints to newspapers and
politicians, reveal a far less positive assessment of government
rehabilitation efforts than some allegations of pro-Punjab bias
would allow. Indeed, Raghuvendra Tanwar has declared that
'Attractive statements supported by huge statistics indicating the
dimensions of the resettlement effort were routinely issued, sadly
these statements concealed a whole body of corrupt decisions of
injustice and unfairness. This trend increased as days and months
passed for as long as the resettlement measures continued.' 62

62 Tanwar, Reporting the Partition ef the Pwyab 1947, 473.

 IAN TALBOT

Those without wealth or connections were at the mercy of

corrupt and incompetent officials.
It is important to recognize here that there is no echo of
Indian Bengali allegations of a pro-Punjabi bias in rehabilitation
efforts amongst Muslim refugees who migrated to East Bengal.
This in part reflects the fact that the refugee situation in East
Bengal has been under-researched, for it cannot be deemed
insignificant. Contemporary press reports in March 1950, for
example, record that Dacca was 'overflowing with refugees', 63 It
may also be the result of a more efficient government response
which was aided by the relatively greater availability of evacuee
property.
The East Bengal government early in 1950 established refugee
camps on the outskirts of Dacca and Chittagong to relieve
congestion. Around 1 ,ooo people were accommodated, for
example, in specially constructed barracks at Samair in the
Kurmitola area of Dacca. 64 About 8,000 refugees were rehabili-
tated in the Bogra district of East Bengal, where they were allot-
ted between 3 and 6 acres ofland. 65 Such short-term measures
were accompanied by longer-term rehabilitation efforts. These
were coordinated by a new East Bengal Relief Commissioner, N.
M. Khan. He reported early injune 1950 that Rs. 50 lakhs had
already been spent on such measures as the provision of stalls for
shopkeepers, and the distribution of sewing machines and
looms. 66 His Employment Bureau had placed 45,000 persons in
various jobs by September 1950. 67 A month earlier, the Pakistan
government had announced that it was advancing 12 million
rupees to the East Bengal authorities for rehabilitation purposes.
This enabled ambitious urban and rural rehabilitation schemes
to be planned. The former involved the construction of five satel-
lite townships near Dacca, Chittagong, Syhlet, Jessore, and
Rangpur at the cost of Rs. 1 million. They were designed to
house over roo,ooo refugees. Three million rupees were set aside
for a rural rehabilitation scheme. Families were to receive 5 acres
of land along with a maintenance allowance of Rs. 50 per family
until their first harvest. 68 InJune 1951 the government finalized a
scheme for a ro,ooo-acre refugee colony at Aflong in Sylhet. 69 It
63 Civil and Military Ga;,_ette (Lahore), 18 Mar. 1950. 64 Ibid. 5 Apr. 1950.
65 Ibid. 20 Apr. 1950. 66 Ibid. 7June 1950. 67 Ibid. 22 Sept. 1950.
68 Ibid. 6 Aug. 1950. 69 Ibid. 6June 1951.
 The 1947 Partition oflndia 339

simultaneously gave 250 houses free of charge to refugee families

in the Mirpur colony, Dacca. 70
The early rehabilitation effort of the Pakistan authorities in
East Bengal appears both more planned and urgent than that of
the West Bengal government. This was borne out by the latter's
own Statistical Department figures published in February 1952.
They revealed that 12 per cent of the 2.14 million refugees were
living on land on which they had trespassed and 72 per cent were
unemployed. 71 In addition to relying on grants from the Centre
for rehabilitation purposes, the East Bengal provincial govern-
ment from July 1952 levied a 'refugee tax' on the licences
required for the export of raw jute. In contrast, the \,Vest Bengal
state government only moved from ad hoe responses to the
formulation of a comprehensive rehabilitation policy after the
June 1953 report of a fact-finding committee. 72 Impressive as the
East Bengal effort appears, it did not match that in West Punjab
during the same period. In 1951, for example, the West Punjab
government earmarked Rs. 17 million for different rehabilitation
schemes in the latter province. They included the establishment
of ten satellite towns, three of which were to girdle Lyllapur. 73
Why, then, is there no anti-Punjab discourse amongst Muslim
Bengali refugees, unlike their Hindu counterparts?
Until there is further research, a definitive response cannot be
given. It appears, however, that the funding was adequate, given
the fact that the overall situation was more favourable for the
Muslim refugees in East Bengal than their Hindu counterparts in
West Bengal. This meant their lot was less intolerable than that
of the highly articulate Hindu refugees who settled in Calcutta.
There was a relative abundance of evacuee houses in the East
Pakistan towns. 74 The Hindu population of Dacca, for example,
declined from 58 per cent of the population on the eve of
Independence to just 4.6 per cent. 75 According to a survey of

70 Dawn (Karachi), 18June 1951.

71 Statesman (Calcutta), 16 Feb. 1952.
72 Samir Kumar Das, 'Refugee Crisis: Responses of the Government of West Bengal',

in Prudip Kumar Bose (ed.), Refugees in West Bengal: Institutional Practices and Contested
Identities (Calcutta, 2000), 13.
73 Civil and Military Gazette (Lahore), 23 Sept. 1951.
74 Tai Yong Tan and Gyanesh Kudaisya, 1ne Afiermath ef Partition in South Asia

(London, 2000), 169.

75 Ibid.
340 IAN TALBOT

Hindu neighbourhoods in December 1950, Muslims now

controlled 6,255 out of 7,175 properties owned by Hindus in
1947. 76 Unlike in Calcutta the terms of trade in properties very
much favoured the refugees. Although Dacca's population mush-
roomed by 53 per cent in the decade 1941-51, the housing
demands of its migrants could be met. The East Pakistan coun-
tryside was also much better able to accommodate refugees than
West Bengal's. The latter region was more overcrowded and its
agriculture was both qualitatively and quantitatively inferior. 77
Muslim refugees in West Punjab were also in more favoured
circumstances than their counterparts in the Indian Punjab.
Hindu and Sikh refugees vacated 9.6 million acres of land in
Pakistan, while Muslims left behind 5.5 million acres of land in
India. 78 While refugee Sikh farmers had to make do with unirri-
gated tracts of land and smaller holdings under the system of
'graded cuts', Muslim cultivators from the East Punjab took over
the fertile tracts ofland in the Canal Colony areas abandoned by
Sikh farmers. Rural resettlement was nevertheless not all plain
sailing in West Punjab. This was in part because Muslim owners
of tenanted land abandoned by non-Muslim farmers still sought
their share of the crop (bataz) from refugees. There were also
delays in making permanent allotments of land to refugees
because of the time it took to exchange land settlement records.
Some refugees were even ejected from land they had been semi-
permanently allotted, in May 1952. Such organizations as the
Muhajir League and the Jinnah Awami League lent their politi-
cal support to the hunger strikes and public protests of the
affected refugees. 79
The East Punjab government attempted to ease the resettle-
ment of West Punjab cultivators by establishing rural housing
schemes. Model villages were constructed on the sites of evacuee
villages which had been demolished during the violence.
According to M. S. Randhawa's figures, there were some 1,800
East Punjab villages where go per cent of the houses were demol-
ished. 80 Before refugees were permanently allotted land they
76 Ibid. 168. 77 Ibid. 144.
78 SeeJ. B. Schechtman, 'Evacuee Property in India and Pakistan', Pacific Affairs, 24
(1951), 411-12.
79 Statesmen Week!), (Calcutta), 31 May 1952.
80 M. S. Randhawa, Out qfthe Ashes: An Account qfthe Rehabilitation qfRefagees.from Pakistan
in Rural Areas qf East Punjab (Bombay, 1954), 153.
 The 1947 Partition of India 341

were disbursed loans for food and for fodder. Large sums of
money were set aside for the purchase of bullocks, seed, and the
reconstruction of houses and wells. When land was permanently
allotted in 1950, loans were provided for agricultural moderniza-
tion such as water pumps, tractors, and tube-wells. Tractor loans,
for example, amounted to Rs. 32 lakhs. 81 This laid the precondi-
tions for the Green Revolution success story of the East Punjab.
During the period September 1947 to March 1951, Rs. 4 and a
half crores were disbursed to displaced cultivators. 82
The state government in West Bengal, in contrast, did little
more than provide basic immediate relief. It lacked sufficient
funding to build large numbers of townships and houses. Its
eventual solution to the refugee accommodation crisis was to try
to disperse refugees to neighbouring states. This became politi-
cally controversial not only within the state, but in such places as
Assam, which was reluctant to receive refugees. A large dispersal
centre was established at Bettiah in the Champaran district of
Bihar. The most ambitious and controversial settlement scheme
involved the moving of over 25,000 families to the 270,000
cleared acres of forest at Dandakaranya in Orissa and Madhya
Pradesh. 83 Many refugees saw this as deportation rather than
rehabilitation. They were exiles in the 'dark forest' like Lord
Rama in the Ramayana. By 1978 over 11,000 families had deserted
the settlement.
Untouchable refugees from East Bengal were also resettled in
the remote Sunderbans region of West Bengal at the beginning
of the 1960s. The scheme was better thought through than that at
Dandakaranya, in that the cultivators who were sent there origi-
nated from the Khulna and Barisal districts, so had some experi-
ence of roughly similar agricultural conditions. They were
provided with 3 acres of land each and loans to build houses and
purchase agricultural equipment. Nevertheless, a survey of the
settlement a decade later found the bulk of the inhabitants mired
in poverty. At least 25 per cent of the cultivators were living in
distress and barely 58 per cent at subsistence level. 84 The govern-
ment cash doles had ceased, while agricultural productivity

81 Ibid. 167. 82 Ibid. 162.

83 See Kudaisya, 'Divided Landscapes', 115 and ff.
84 S. L. De and A. K. Bhattacharjee, 7he Refugee Settlement in the Sunderbans, West Bengal:

A Socio-economic Study (Calcutta, 1972), 47.

342 IAN TALBOT

remained low. The poor communications hindered economic

activity. The nearest market was 18 miles away, while a journey
by country boat of 35 miles was required to reach the nearest
railway station. As the investigators concluded in their findings,
much poverty was the result of the 'lack of far-sightedness of the
Government department responsible for planning the project'. 85
The West Bengal government's limited response to the refugee
'problem' was in part the result of disputes with the Centre over
funding, and in part reflected the sense of the state's being over-
burdened because of its weak economy and high population
density. The refugee plight was further worsened by the decision
of the government of India not to extend evacuee legislation to
either West Bengal or Assam. This reflected the view that migra-
tion was not permanent and that further transfers of population
should be discouraged. Indeed, the latter waves of refugees were
seen as economic migrants rather than as victims of Partition. In
the Punjab region, the governments of East and West Punjab
agreed on a response to the problem of abandoned property as
early as September 1947. Arrangements were made for the
exchange of property and compensation for abandoned prop-
erty. This, together with the official refusal to recognize the
illegal seizure of property, eased the rehabilitation of refugees. At
the same time, the East and West Purtjab governments through
refugee taxes and disbursements from the Centre set aside large
sums for resettlement.

Puryabi and Bengali Refugee Discourses

The Indian government's account of refugee resettlement

focused around the Punjab 'success story'. This was reproduced
in official works such as Millions on the Move and in semi-official
studies such as M. S. Randhawa's carefully named study Out qf the
Ashes. Such accounts reproduced the statistics of government
assistance interspersed with photographs of smiling refugees in
their new homes in East Punjab. There were even lapses into
Orientalist stereotypes of the 'sturdy' Punjabi Jat peasants with
their capacity for hard work and enterprise. These stereotypes
were later internalized. V. N. Datta's pioneering study of Punjabi
85 Ibid. 50.
 The 194 7 Partition oflndia 343

refugees in Delhi extols their willingness to turn their hand to

anything in order to make their way. 86
First-hand accounts go even further and attribute the success-
ful rebuilding of lives after the upheaval of Partition to self-help.
Interviews in such cities as Lahore and Amritsar leave little room
for government support for resettlement. We have seen above,
however, that the governments of both East and West Punjab
diverted large sums of money to 'the refugee problem'. Some
refugees, for psychological reasons, may choose to forget the role
government assistance played in their resettlement. They refer if
at all to outside assistance coming from community and religious
organizations. This may reflect the persisting influence of such
organizations in their lives. It also is rooted in the fact that, as
Raghuvendra Tanwar has noted, much government aid was
mired in delays and corruption.
Official accounts often contrast Bengali dependency with
Punjabi flexibility and entrepreneurial spirit. Aside from dealing
in colonialist stereotypes, such understanding displaces blame
from the government, for the more problematic resettlement
process, to the refugees themselves. As we have seen, however, the
emergence of the squatter communities in Calcutta displays just
as much enterprise as that attributed to the Punjabi migrants.
Hindu Purbo Bongiyo (East Bengal) refugees argued that they
were the principal victims of Partition because of the Indian
government's half-hearted approach to their rehabilitation. 87
Salil Sen's 1950s play summed up this sense of victimhood in its
title, Natun Yehudi (The New Jews). Such writers as Prafulla
Chakrabarti have been at pains to highlight the greater govern-
ment assistance to Punjabi than Bengali refugees. 88
There was undoubtedly some truth in the claim made by the
Rehabilitation Minister, A. P.Jain, in New Delhi on r8June 19y2
that 'it was unfair to compare official efforts to rehabilitate
refugees (from East Pakistan with West Pakistan) for attention to

86 V. N. Datta, 'Panjabi Refugees and the Urban Development of Greater Delhi', in

R. E. Frykenberg (ed.), Delhi through the Ages: Essays on Urban History, Culture and Society
(Delhi, 1986), 451, 453-5.
87 For a discussion of these claims, see Joya Chatterji, 'Right or Charity? The Debate
over Relief and Rehabilitation in West Bengal, 1947-50', in Suvir Kaul (ed.), The Partitions
efMemory: The Aflerlife ef the Division efIndia (Delhi, 2001), 74-1 ro.
88 Prafulla Chakrabarti, The Marginal Men: The Refugees and the Lefl Political Syndrome in

West Bengal (Calcutta, 1999), 250 and ff.

344 IAN TALBOT

evacuees from West Pakistan had been paid over a period of 5

years whereas the problem of refugees from East Pakistan had
become a serious one over the past two years'. 89 Moreover,
government assistance in West Bengal, although belated, was
considerable. In response to the 1952 refugee influx, New Delhi
allotted the West Bengal government Rs. 55 lakhs for relief and a
further Rs. 25 lakhs for rehabilitation. 90 The following year, the
West Bengal government received a sum of Rd. 2.7 crores for the
task of refugee rehabilitation in the Burswan, Nadia, and 24
Parganas districts. 91 By the end of 1957, there were 83 govern-
ment-sponsored refugee colonies with 21,500 families. Up to the
previousjune, Rs. 24.28 lakhs had been sanctioned to fifty-nine
cooperative societies. 92 Loans were also advanced through the
Refugee Businessmen's Rehabilitation Board and separately to
lawyers and medical practitioners. Grants were also sanctioned
for the opening of schools in refugee colonies. 93 Nevertheless,
there was a clear difference in the amount and type of assistance
that was on offer by the Indian authorities in the Punjab and
Bengal regions. Moreover, delays in the flow of funds from the
Centre to the West Bengal government hampered its rehabilita-
tion programmes. 94 As much as a crore of rupees earmarked for
relief in 1953-4 remained unspent. 95
One similarity between the refugee experiences is that of
nostalgia for the land left behind. Punjabi Hindus in Delhi look
back to Lahore. This is reflected in the titles of such accounts as
Lahore: Portrait ef a Lost Ciry; Lahore: A Sentimental Journey; Lahore:
Loved, Lost and 7hereafter. 96 East Bengal refugees look not to a city,
but the villages they abandoned in 'Golden Bengal'. Accounts
focus on memories of public holidays at the time of the major
religious festivals, boat races, the abundance and beauty of the
countryside, harmonious social relations, and the 'respect' for
elders and women. 97 One of the Bengali words for refugee,
89 Statesman Week!), (Calcutta), 21June 1952. 90 Ibid. 25 Oct. 1952.
91 Ibid. 17 Jan. 1953. 92 Das, 'Refugee Crisis', 20 and ff. 93 Ibid. 25.
94 See Statesman Week!), (Calcutta), 20 Dec. 1952. 95 Ibid. 20 Feb. 1954.
96 See Som Anand, Lahore: Portrait of a Last Ciry (Lahore, 1988); Pran Nevile, Lahore: A
Sentimental]oumry (New Delhi, 1993); Sahdev Vohra, Lahore: Loved, Lost and Thereefler (Delhi,
2004).
97 See Sanyal, Making of a New Space, 163 and ff.; Dipesh Chakrabarty, 'Remembered
Villages: Representations of Hindu-Bengali Memories in the Aftermath of Partition', in
D. A. Low and Howard Brasted (eds.), Freedom, Trauma, Continuities: Northern India and
Independence (New Delhi, 1998), 133-53.
 The 1947 Partition of India 345

udvastu (outside of home), attested to this painful separation from

ancestral roots. There is an elegiac quality about their memory of
the idyllic villages of 'Golden Bengal', contained in such collec-
tions as Chhere asha gram (The Abandoned Village), just as mohajirs
remember the small towns of eastern UP they have left for
Sindh. 98 Punjabi and Bengali Hindu identification of Partition
with loss, of course, sits more easily with the national historical
discourse than does the similar sentiment of the migrant Uttar
Pradesh elite within Pakistan.

Conclusion

The 'new history' of Partition reveals the extent to which it was a

highly differentiated experience. This essay represents a modest
contribution to this understanding. The waves of migration in
Bengal contrast v.rith the single flood of refugees in the immediate
post-Partition period in Punjab. These differences reflected the
varieties in the patterns of violence. The year 1950 was a far
more significant date for East and West Bengal than 1947. Even
when accounting for the propagandist element, it is clear that
government responses were less focused and effective in the
Bengal than the Punjab regions. A litany of reasons why rehabil-
itation was a more difficult task in West Bengal than East Punjab
was provided by both state and Union-level ministers. 99 There
was, however, differentiation in government responses within
regions, as well as between them. The East Bengal government,
for example, appears to have diverted greater energy to the reha-
bilitation task than its Indian counterpart. It allocated large sums
of money from the Centre for both rural rehabilitation and the
construction of satellite refugee towns and colonies. The East
Pakistan authorities also raised their own funding by means of a
'refugee tax'.

98 The latter is the setting for the famous Urdu writer lntizar Husain's short story Akhri
Mom Batti, which explores the theme of Partition loss. See Ian Talbot, Freedom's Cry: The
Popular Dimension in the Pakistan Movement and Partition Experience in North-West India (Karachi,
1996), 142~4. Nostalgia for East Bengal is seen in such poems as Taslima Nasreen's
'Broken Bengal', which has been translated into English by Subhoranjan Dasgupta, from
the selection Behula eko. bhasiyechilo bhela.
99 See e.g. the press conference in Calcutta on 24 Dec. 1954 given by Mehr Chand
Khanna, the Union Rehabilitation Minister. Statesman Weekfy (Calcutta), 25 Dec. 1954.
 IAN TALBOT

As for the refugee experience, this was highly diverse. Even

within the city of Calcutta, which underwent profound changes,
differentiation in refugee experiences is visible. Despite their
human suffering, the first wave of upper-caste refugees, whose
memories and complaints fill the historical narrative, were better
able to rebuild their lives than the later migrants from poorer
agricultural backgrounds. Within Punjab, it was the poorer and
less politically acute communities that were caught unawares and
had to cross the killing fields in August 1947. Sections especially
of the West Punjab Hindu elite had moved their money and
families to places of safety well in advance of Independence.
Wealth and connection impinged equally on refugee success in
resettlement on both sides of the border.
At the same time, as a more detailed understanding of the
Partition process is emerging it is clear that major gaps remain.
Little, for example, has been written about the Untouchable
community's experience of violence, migration, and resettle-
ment.Just as the Punjab experience has been privileged in the
general historical discourse of Partition, so the experience of the
bhadralok refugees who settled in Calcutta has dominated writings
on the population movement in eastern India. This excludes the
movement to West Bengal of non-elite groups such as Santhals.
There are also the untold stories of the migration of Bengalis to
Tripura and of Muslim migrants from West Bengal, Assam,
Tripura, Bihar, and Uttar Pradesh to what is now Bangladesh.
Finally, work is required on the experiences of Muslims in West
Bengal who did not migrate to what is now Bangladesh.
Anecdotal evidence points to the fact that they faced displace-
ment as Hindu refugees illegally occupied properties. There
appears to have been a process in which former mixed localities
in Calcutta were increasingly replaced by what were, in fact,
Muslim ghettos. Rural Muslim populations also appear to have
shown a tendency to migrate from south-west Bengal to concen-
trated Muslim areas in north Bengal, as well as clustering near
the border in Nadia district, for example, over which they could
flee in times of communal strife. Much more research is required
on Muslim Partition-related migration within West Bengal.
Actual violence or its threat was a crucial factor in the migra-
tion processes in both Punjab and Bengal. Individuals were
reluctant to leave their ancestral homes, not only because of the
 The 194 7 Partition oflndia 347

material loss this involved, but because their identity was vested
in these localities. Uprootedness was to be avoided at all costs.
The violence which demographically transformed vast swathes of
north India was not spontaneous. While it possessed elements of
retribution and opportunism, it was in many instances carefully
planned and executed. Communal organizations sought forcibly
to remove minority communities. The state afforded scant
protection to all its citizens in situations of extreme polarization
along community lines. In both Bengal and Punjab, local officials
and policemen not only acquiesced, but participated, in commu-
nal violence. The dislocation was so severe in the Punjab that
spontaneous mass migration gave way within a fortnight of
Independence to a virtual exchange of population. Despite the
continued violence and suffering which accompanied this
process, it eased the Pakistan and Indian states' task of refugee
resettlement in the region. In Bengal, the pattern of waves of
migration occasioned by violence from non-state and sometimes
state actors continued for years after Independence. The demo-
graphic transition was never as complete as in Punjab. The
absence of evacuee property intensified the housing shortage,
especially in Calcutta. The result was that refugees from East
Pakistan termed themselves the 'New Jews' and principal victims
of Partition. This discourse was created by the most privileged
section of the refugee community. Their lower-caste counter-
parts were reduced to a miserable existence of pavement
dwelling which etched Calcutta in both the national and interna-
tional consciousness as the 'city of dreadful night'.