This article was downloaded by:[CDL Journals Account][CDL Journals Account]

On: 23 May 2007Access Details: [subscription number 770849126]Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Social IdentitiesJournal for the Study of Race, Nation andCulturePublication details, including instructions for authors and subscription information:http://www.informaworld.com/smpp/title~content=t713445719

Specificities: Official Narratives, Rumour, and the SocialProduction of HateVeena Das

To cite this Article: Das, Veena , 'Specificities: Official Narratives, Rumour, and theSocial Production of Hate', Social Identities, 4:1, 109 - 130To link to this article: DOI: 10.1080/13504639851915URL: http://dx.doi.org/10.1080/13504639851915

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

This article maybe used for research, teaching and private study purposes. Any substantial or systematic reproduction,re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expresslyforbidden.

The publisher does not give any warranty express or implied or make any representation that the contents will becomplete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should beindependently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings,demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with orarising out of the use of this material.

© Taylor and Francis 2007

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

Social Identities, Volum e 4, N um ber 1, 1998

Speci® cities:

O f ® cia l N arratives, Rum our,

and the Socia l Production of H ate

VEEN A DAS1

U niversity of D elhi

This paper is not about discourses of genocide as the term is understood in its

current avatara in international instruments and covenants. It is about the social

production of hate: it examines the movement of images between emergen t

discourses of militancy and the diffused understandings of events in rumours

circulated during crises. I argue that such movements create the conditions

under which social groups become pitted against each other in fear and mutual

hatred, constructing images of self and other from which the subjectivity of

experience has been evacuated. In this social production and circulation of

hate, the images of perpetrator and victim are frequently reversed , depending

upon the perspective from which the memories of traumatic events and of

everyday violence are seen and re-lived. This exchange of images does not

mean that we cannot recognise those who suffered during traumatic events. It

does suggest, however, that we need to bear in mind that in the phenomenol-

ogy of panic, aggressors can experience themselves as if they were victims.

The events I examine here occurred in India during the 1980s, when the

militant movement in the Punjab and the related counter-insurgency opera-

tions of the state generated considerable animosity between the Hindus and the

Sikhs. The result was violent confrontations that obliterated the solidarities of

everyday life. I focus on an especially grievous event Ð the brutal violence

against Sikhs in the resettlement colonies in Delhi in 1984, following the

assassination of then-Prime Minister Indira Gandhi. The discursive formations

and the enunciative modalities that emerged were premised on a particular

feature of language, viz, its perlocutionary force, to wrest open the un® nished

stories of the past and to give shape to the present.

The events that anchor my argument are, by any de® nition, extreme.

However, the discourses generated during this period bear a relation to

stereotypes in everyday life such as the fanaticism of the Sikhs or the effeminate

character of the Hindus. How and under what conditions these stereotypes

came to acquire the lethal quality that I describe here is the matter of enquiry.

It is important to understand both the processes through which everyday life

gets transformed and the processes through which trust (at least a minimal

trust) is re-established among communities at one time torn apart by hate.

I begin with a description of the militant discourse generated by the

1350-4630/98/010109-22 $7.00 Ó 1998 Carfax Publishing Ltd

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

110 V eena Das

movement for a separate homeland (Khalistan) for the Sikhs. Then, I trace how

the images from that discourse travelled to other sites and crystallised into

stereotypes that fed the rumours that led to violence against the Sikhs in 1984.

This ordering of events is not meant to af® rm a causal connection betw een the

earlier events and the later ones; nor is it intended to cast the Sikh militant

movement as responsible in any way for the violence against the Sikhs after the

assassination of Mrs Gandhi. In telling a story, the order of events never

corresponds exactly to their unfolding in pure durational time. No matter at

what point one chooses to begin a story, there is always a prior story waiting

to be activated, one that has been lying inert or circulating within only limited

zones. Thus, the way different social groups recount their memories of the

events discussed here is strongly in¯ uenced by their social location, but it is not

completely determ ined by it. This play of forces is what I hope to capture in

this article.

The Discourse of M ilitancy

Kapur (1987) provides an excellent, comprehensive account of the emergence of

a militant movement among the Sikhs in India and among emigrant Sikhs.

Between 1981 and the end of 1984, Sikh leaders led a series of mass civil

disobedience campaigns against the Indian government in hopes of pressuring

the state to ful® ll several demands such as a separate personal law for the

Sikhs, and greater ® nancial autonomy for Punjab. At the same time, these

leaders also encouraged the use of violence to achieve the movement’ s ends.

The process of formulating these demands had unanticipated consequences: it

contributed to the development of new stereotypical images of Hindus and

Sikhs in the published literature of the militants, on posters, and in oral

discourses. The simultaneous growth of an intolerant militancy among Hindu

fundamentalist groups produced many of the same kinds of stereotypes,

especially with regard to the `weak’ and `emasculated’ Hindu (see Das, 1995,

1996b; Kakar, 1995; Nandy, 1994; Pandey, 1993, 1994).

In the Sikh militant discourse, the image of the self was that of the martyr

whose past sacri® ces had sustained and energised the community. H indus, in

contrast, were represented as weak and effeminate, or as cunning and sly. In

either case, though they were represented as historically dependent on the

protection offered by the Sikhs, they were also seen as not averse to betraying

their erstwhile protectors.

In the written and oral discourses of Sikh militancy, the Hindu `character’

was envisaged in terms of the dangers it posed to the masculinity of the Sikhs.

There was a move to establish that the history of the Sikhs Ð inscribed on the

body of the martyr Ð was a re¯ ection of the masculine Sikh character, while

the feminine Hindu character is what has been inscribed on the history of the

Indian nation. Thus, masculinity became the de® ning feature of the Sikh

community; and the Hindu community was understood to be characterised by

an emasculated feminin ity. Overlapping imagery blurred the boundaries be-

tween the Hindu character and the character of the nation as a whole, so that

over time, the character of each came to stand for the other.

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

O f® cial N arratives, Rumour, and the Social Production of H ate 111

C ommunity, K in and Masculinity

Anderson (1983, p. 131) points out that the style of thought in which large

collectivities (such as the nation) are invested with the status of a `natural’

entity often incorporates the use of kinship terms as metaphors for de® ning the

relationship between individual and collectivity. The story is incomplete,

however, if we look only at metaphors of `true kinship’, for simultaneously

there is a repudiation of `false kinship’ , through which previous intimacies

between communities are disavowed. This complex interplay between themes

of true kinship and false kinship offers some caveats to Anderson’ s assumption

that kinship categories bestow the nation with the status of a `natural’ entity.

Both for purposes of inclusion and exclusion, the metaphors of kinsh ip and

intimacy provide frameworks for thought and hence create intense feelings of

love and hate, ® delity and betrayal.

The oral discourse of the militants relied extensively on metaphors of male

relatedness to create a sense of community among the Sikhs. Kinship ties were

used in two different senses. One was the true tie of father-son relations to be

acknowledged and celebrated; the other was created from a negative skewing

of an earlier ® lial relationship between Hindus and Sikhs, that of a parent

religion (Hinduism) and its descendant (Sikhism). In a novel articulation this

relationship was imaged as a metaphor that implied illeg itimacy. Clear exam-

ples of these images are present in the following excerpt from a speech

Bhindranw ale made when addressing a congregation.2

Khalsa ji [you, who are the pure ones], the Sikhs are the son of the true

king Guru Gobind Singh ji. Now you know that a son must resemble his

father. If the son does not resem ble his father, then you know the term

used for him [i.e., bastard]. If a son does not behave like his father, then

people begin to view him with suspicion. They [the Hindus] say the

Sikhs are the descendants of Hindus. Are they pointing a ® nger at our

pure ancestry Ð how can a Sikh bear to be called any one else’ s son?

This concern with establishing `pure ancestry’ and the doubts about illeg it-

imacy and true paternity are obviously male doubts: they point to the extent to

which the imagined homeland of the Sikhs was conceived as a masculine

nation. This view of the group as being worthy of having a nation, of being

able to lay claims to a homeland, moves on the axis of being the deserving sons

of a valiant father. This claim, in turn, is articulated through the motif of

martyrdom Ð being capable of making sacri® ces, bearing pain and hardship.

To grasp the full meaning of this imagery in the Sikh discourse, it is

important to know that in the Hindu imagery of the nation, the homeland was

a motherland ; and though the nation was conceived as masculine, it was made

up of the sons of a mother. The imagery in the nationalist discourse during the

struggle against British colonialism represented the nation as a mother who

had been shackled by the ties of a foreign rule and had to be rescued by her

valiant sons. Thus, although a concern with masculinity marks both the Hindu

and Sikh militant discourses, the differential genealogy (sons of a father, or

sons of a mother?) shaped the notions of self and other in diverse ways.

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

112 V eena Das

Further complicating matters is the fact that Gandhi, in his non-violent

movement, transformed the notions of masculinity and femininity. He advo-

cated strategies of passive resistance, such as fasting , or submitting one’s body

to the violence perpetrated by the British colonial regime, rather than the more

masculine strategies of violent resistance.3

In the written and oral discourse of Sikh militancy, it was repeatedly stated

that the Sikhs could not belong to a nation that claimed a feminine ® gure such

as Mahatma Gandhi for its `father’ (bapu Ð the affectionate title given to

Gandhi by the genera l populace). The de® ning principle of the Indian national-

ist struggle, non-violence, was said to be `passive’ and `womanly’. Anxiety

about such femininity became palpable in the oral discourse, for it seemed to

threaten the inheritance of a manly way of confronting evil, an inheritance

which was further characterised as the natural birthright of the Sikh. Gandhi’ s

non-violent movement, it was said, was appropriate only for the feminised

Hindus. In one of his speeches, Bhindranwale propounded the idea that it was

an insult for the Sikhs to be included in a nation that considered Mahatma

Gandhi to be its father, for his techniques of ® ghting were quintessentially

feminine. He (Gandhi) was symbolised by a charkha, the spinning wheel, which

was a symbol of women. Bhindranwale asked his listeners,

Can those who are the sons of the valiant guru , whose symbol is the

sword, ever accept a woman like Mahatma as their father? Those are the

techniques of the weak, not of a race that has never bowed its head

before any injustice Ð a race whose history is written in the blood of

martyrs.

Thus the construction of the past in terms of a genealogy of father/son

relations was also a construction of the self and the other. To be able to claim

true descent from the proud Gurus4

(the ten acknowledged founders of the Sikh

religion ), it was argued, all corruption that had seeped into the Sikh character

because of the closeness to the Hindus must be exorcised. Through the

narrative web of Sikh history as a history of martyrdom, a Sikh character was

sought whose negative counterpart was the Hindu character. A `Hindu’ history

represented a dual danger: it denied Sikhs their rightful place in history; and,

as many authors stated, it was a conspiracy to make the martial Sikhs into a

weak race.

The Sikhs have been softened and conditioned during the last ® fty years

to bear and put up with insults to their religion and all forms of other

oppress ion, patiently and without demur, under the sinister preaching

and spell of the narcotic cult of non-violence, much against the clear

directive of their G urus , their Prophets, not to turn the other cheek

before a tyrant, not to take lying down any insult to their relig ion, their

self-respect, and their human dignity.5

As seen in this discourse, the danger issued not from a heroic confrontation

with a masculine other, but from a feminine other who would completely

dissolve the masculine self of the Sikh. `With such an enemy’, one warning

predicted, `even your story will be wiped out from the face of the earth’ .

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

O f® cial N arratives, Rumour, and the Social Production of H ate 113

Given this articulation of community as meaning a community of men (and

further, as a community of sons descended from the father in whose ancestry

all signs of the mother have been obliterated), it is not surprising that the oral

discourse contains direct exhortations to the Sikhs to shed their femininity.

Since the most visible sign of the masculinity among Sikhs is the sword many

speeches simply exhort `Shastradharihowo ’ , that is, that the listener become the

bearer of weapons. In most of his speeches, Bhindranwale asked Sikh house-

holds to collect weapons Ð especially Kalashnikov guns Ð so that they could

protect the honour of the community when the time came.

In addition to the external and historical sign of Sikh masculinity embodied

in the sword, a Sikh ’s masculinity is also said to be visible in his beard.

Bhindrawale chastised the Sikhs,

If you do not want beards then you should urge the women to become

men and you should become women. Or else, ask nature that it should

stop this growth on your faces. Then there will be no need to exhort you

to wear long beards. Then there will be no need for me to preach

[prachar karna], no need to break my head on this [matha khapai karna].

Another leader, a functionary of the Akali Dal,6

ins isted that the ¯ owing beard

of the Sikh man was a direct challenge to the authority of the state. When a

threat to the Sikh community was articulated, it was often stated that `they’

have their eyes on `your’ sword, on `your’ beard [`ona di nazar twadi kirpan te

hai Ð ona di nazar twadi dadi te hai’].

Thus, we see the importance of the theme of the feminine other destroying

the community by robbing it of its masculinity and bestowing a feminine

character on it. The production of communal hate in this case was not based

upon a long history of hostility between Hindus and Sikhs. Instead, hate

seemed to be the shears destined to rent a shared history and a social ecology

of connectedness.

A master narrative of Sikh history, cast as a series of systematic dualisms

separating the Sikh self from the Hindu other, could not have been woven

without a systematic `forgetting’ of the close relations between Hindus and

Sikhs in everyday life, especially the bonds of language, common mythology,

shared worship, and the community created through exchanges. In fact, even

the participation of the Sikhs in the communal riots against the Muslims,

which, in addition to being documented in history textbooks, could still be

veri® ed by living people, was no longer acknowledged. Projecting all the

darker aspects of the past onto the Hindus purged the Sikhs of memories they

could no longer afford to own.7

As one example of this kind of `forgetting’ ,

incidents of communal tensions from the 1920s were discussed primarily in

terms of Hindu-Muslim con¯ icts, as if the Sikhs had not ® gured in these

con¯ icts at all. Under the subheading of `It Happened Before’ , the SGPC’s

White Paper entitled They Massacre Sikhs stated the following:

This phenom enon in which Sikh religious sensibility is calculatedly

outraged and their human dignity cruelly injured , has its historical

antecedents in this part of the world. It was in the late twenties of this

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

114 V eena Das

century that a cultural ancestor of the present anti-Sikh Hindu urban

crust wrote and published a small book, purporting to be a research

paper in history under the title of Rangila Rasul: Moham mad, the Pleasure

Loving Prophet ¼ The entire Muslim world of India writhed in anguish

at this gross insult to and attack on the Muslim community but they

were laughed at and chided by the citi® ed Hindu press of Lahore ¼ But

the process of events that led to bloody communal riots in various parts

of India til the creation of India and Pakistan and the partition of the

country itself , with tragic losses in men, money and property, is directly

and rightly traceable to a section of the majority community exempli® ed

in the matter of Rangila Rasul ¼ (pp. 29± 30)

The ® rst act of wilful forgetting, then, involved purging the community of

evil by tracing all wrongdoing to a single source, `the citi® ed Hindu majority ’ .

The second act of forgetting was to construe all acts of violence, both those

within the Sikh community (in institutionally sanctioned practices such as

feuding) and those directed outward, as the violence of martyrdom. Finally,

there was the assumption that the Indian state was an external institution, in fact

a Hindu institution, that had been imposed upon the Sikh community rather

than one created through practices prevalent in the Punjab region itself . This

view allowed the Sikh community both to absolve itself of all blame for corrupt

institutional practices on the part of the state by shifting the burden to the

Hindu community, and to create a discourse of betrayal.

The Betrayed Lover

One of the metaphors that occurred repeatedly in the militant discourse was

that of the Sikhs as the betrayed lovers of the Indian state. The militant

literature stated that it was their sacri® ces that had brought India freedom, but

the Sikhs were denied their rightful place in the new con® guration of nations.

Bhindranw ale made this analogy during a speech in which he asserted that

while Muslims got Pakistan, and Hindus became the de facto rulers of India, the

nishan sahib (i.e., the ¯ ag that would be the sign of the Sikh nation) was not

allowed to ¯ y over the country. On another occasion, he stated that an

agreem ent had been reached that the saffron colour that is the symbol of Sikh

martyrdom would ¯ y over and above the other two colours Ð green for the

Muslims and white for the Hindus because it had always been a Sikh who led

the procession of satyagrahis (Gandhi’ s term for the non-violent protesters,

literally signifying the adherence to truth), as Hindus were too cowardly to do

so. Obviously, this is not the of ® cial interpretation of the colour scheme of the

national ¯ ag. But the story Bhindranwale told anchored a ¯ oating truth with

the stamp of authenticity.

In other examples of metaphor in militant discourse, the character of the

Hindu was compared with the character of the snake. One poster declared, `A

Hindu never kills a snake. He asks the Muslim to kill the snake. If the snake

dies the Hindu is happy; if the Muslim dies the Hindu is happy’ . The poster

concluded with this warning: `In confrontation with such a community even

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

O f® cial N arratives, Rumour, and the Social Production of H ate 115

your name will be wiped out from the annals of history’. The theme of the

untrustworthiness of the Hindus was further elaborated through reference to

the feminised character of its rulers . Thus, the leader Indira Gandhi was said

to be a widow, one born in the household of Pandits, the Brahmin caste which

was always subservien t to the ruler classes. The implication was that only a

Hindu could accept being ruled by a woman.

The superimposition of the images of femininity and masculinity over the

images of Hindu and Sikh assumed the state as an overarching presence, and

one that provided a speci® c context for the contests between the two groups.

One could go further and say that it is this overarching presence that tore the

Hindus’ and Sikhs’ mutual creation of images from its anchors in everyday life.

Each group turned to public spaces, and each vied with the other for control

over those spaces. The repertoire of the negative and hateful images of self and

other slowly seeped into the understandings of many people, forming the

unconscious grammar through which the grievous events of Operation Blue

Star8

(discussed later in this article), the assassination of Indira Gandhi, and the

collective violence against the Sikhs in Delhi in 1984 were partially produced

and interpreted.

The Phenomenology of Rumour

I turn now to tracing the ways in which the diffused understandings of Sikh

and Hindu character found expression in the phenomenology of rumour. To do

so, I examine the speci® c context of the societal crisis surrounding the assassin-

ation of Indira Gandhi. The key characteristics of this crisis were as follows: a

mounting panic which signalled the breakdown of social communication; the

animation of a societal memory composed of incomplete or interrupted social

stories; and the appearance of panic-laden rumour,9

in the form of a voice

which was unattributed, unassigned and yet anchored to the images of self and

other that had been circulating in the discourses of militancy. The withdrawal

of trust from normally functioning words constituted a special vulnerability to

the signi® er which in turn led to ways of acting that seemed both out of control

and uncontrollable. It was in these moments that images generated in the

speeches of a `Sant’ Bhindranwale or a `Sadhavi’ Rithambra found a place in

the collective repertoire of social groups, displacing the subjectivity of every-

day life with a subjectivity more appropriate to a form of death rather than a

form of life.

My emphasis on rumour’s displacement of the subjectivity of everyday life

places the functioning of rumour in a somewhat different, even negative, light

compared to many other formulations. A brief detour to sketch the contours of

this difference may be useful, especially to indicate the social contexts in which

rumour may perform a critical function, as well as those in which it creates

conditions for the circulation of hate. At the very least, my view of the function

of rumour suggests that a satisfactory theory of rumour independent of the

forms of life (or forms of death) within which it is embedded is not possible.

One well-known analysis of the power of rumour is George Rude ’s (1959,

1964) classic study of rumour’s role in mobilising crowds during the French

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

116 V eena Das

revolution.10

As I have indicated elsew here (Das, 1990a), many scholars have

viewed this mobilising power of rumour positively, seeing crowds as agents of

collective action, concerned with redressal of moral wrongs. E. P. Thompson,

for example, observed , `The food riot in the eighteenth century was a highly

complex form of direct popular action, disciplined and with clear objectives’

(1971, p. 78).

In Indian historiography, Ranajit Guha (1983) has secured an analytical

place for rumour as important trigger and mobiliser, `a necessary instrument of

rebel transmission ’ in popular peasant uprisings (p. 256). Guha identi® es

several other characteris tics of rumour as important elements upon which to

construct a theory: the anonymity of the source of rumour, its capacity to build

solidarity, and the overwhelming urge it prompts in listeners to pass it on to

others . He draws repeated attention to rumour as an important means of

mobilising the peasantry. He sees this function as one which was `speci® c to a

pre-literate culture’ (pp. 226, 251), re¯ ecting `a code of political thinking which

was in conformity with the semi-feudal conditions of the peasant’s existence’

(pp. 264± 65). Of® cials viewed the peasant insurgencies fuelled by rumour as

instances of peasant irrationality. For the peasant insurgents , these were means

of spreading the message of revolt.

Bhabha (1995) deftly isolates two aspects of rumour from Guha’s analysis

that he (Bhabha) considers important for building a general theory of rumour.

The ® rst is the rumour’ s enunciative aspect, and the second the performative

aspect. `The indeterm inacy of rumour’ , he says, `constitutes its importance as

a social discourse. Its intersubjective, communal adhesiveness lies in its enun-

ciative aspect. Its performative power of circulation results in its contiguous

spreading, an almost uncontrollable impulse to pass it on to another person’

(p. 201). He concludes that psychic affect and social fantasy are potent forms of

potential identi® cation and agency for guerrilla warfare and hence rumours

play a major role in mobilisation for such warfare.

Other views of rumour, especially those derived from mass psychology,

have traced it to the emotional, capricious, temperamental, and ¯ ighty nature

of crowds. Le Bon (1952) declared that crowds are everyw here distinguished by

feminine characteristics. Some of this denigration of crowds may be easy to

understand in terms of Guha’s formulation of elite prejudice against the

subaltern forms of communication. Still, it is dif® cult to ignore the fact that the

twentieth century has also seen the spectacular politics of crowds in the Nazi

regime and nearer home in the communal riots. In these cases, too, there were

certain moral premises (in terms of the participants’ own understanding of

events) upon which the crowds acted (Das, 1990a, 1996a) but the unconscious

exchange of images drew upon a repertoire that cannot be schematised within

Guha’s framework of subaltern politics.

In his analysis of Nazi crowds, Muscovici (1985) suggests that crowds come

to be spoken of in feminine terms simply to mask the exchange of homosexual

images between an `active’ leader and a `feminine’ crowd. I have elsew here

suggested (Das, 1990a) that themes of revenge dominate in the imagery in

which a crowd is mobilised around the image of a raped woman, or a dead

child. Moreover, the imagery of a community that has been emasculated and

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

O f® cial N arratives, Rumour, and the Social Production of H ate 117

that seeks to recover its masculinity through crowd action plays on the register

of gender in various ways. What is common in these situations where rumours

are deployed is the perlocutionary force of words; their capacity to do some-

thing by saying something.11

In this way, words come to be transformed from

a medium of communication to an instrument of force.

These prelim inary remarks , I hope, authorise me to conclude that the

essential grammatical feature (in Wittgenstein ’s sense) of what we call rumour

is that it is conceived to spread. Thus while images of contagion and infection

are used to represen t rumour in elite discourse, this is not simply a matter of

their non-comprehension of subaltern forms of communication. It also speaks to

the transformation of language; namely, that instead of being a medium of

communication, language becomes communicable, infectious, causing things to

happen almost as if they had occurred in nature. This speaks to the phe-

nomenological feature of mounting panic, as I shall show, but it also naturalises

the stereotypical distinctions between social groups and hides the social origins

and production of hate.

The A ssassination of Indira Gandhi

Indira Gandhi was shot and killed by two or more of her Sikh security guards

on 31 October 1984, at about 9 am. Although there was no of® cial announce-

ment of the religious identity of the assassins until the next morning, people

somehow `knew’ that she had been murdered by her Sikh guards. There was

speculation in Delhi about the genealogy of this event and its consequences for

the safety of the city’s Sikhs.

Many people traced the origin of the assassination back to Operation Blue

Star. Launched by the Indian army in July 1984, this undertaking alleged ly was

intended to ¯ ush out the militants from the Golden Temple in Amritsar; the

militant leader Bhindranwale died during the army assault. The army’ s forcible

entry into the temple was viewed as a desecration, and many if not most Sikhs

found the event deeply hurtful. The various constructions and counter con-

structions of this event will be examined later in this article. Here, let me note

only that some Sikhs were alleged to have taken oaths in Gurdwaras (Sikh

temples, lit. the doorway to the Guru) to assassinate Mrs Gandhi before the end

of October in order to avenge the insult embodied in Operation Blue Star. Some

people said that Mrs Gandhi had been warned by police security that she

should not have Sikh bodyguards, but that she had refused to have the men

transferred. Opinion was divided over whether to regard the two body guards

as martyrs who had risked their lives to penetrate the security system of the

formidable Indian state,12

or whether they should be considered cowards who

had shot a defenceless woman Ð a woman who had trusted them against the

advice of her security personnel. It is interesting to note that the identity of the

assassins was assumed purely on the basis of rumour. This uncanny knowl-

edge, I believe, can be traced to the un® nished character of Operation Blue Star.

There was a sense at the time that the story had not ended with the storming

of the temple and the death of Bhindranwale; indeed, it was widely believed

that there was bound to be a sequel in the form of a calamitous national event.

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

118 V eena Das

The events of Operation Blue Star were themselves the subject of contention

from the beginning. The government insisted that the sacred shrine of Amritsar

Ð the D arbar Sahib (or the Golden Temple, as it is popularly known) Ð had

become a sanctuary for militants and terrorists, and that the shrine was a

storage place for large quantities of illegal weapons, whose presence endan-

gered public order and state sovereignty. The militant literature depicted the

army operation as a ¯ agrant violation of the rights of the Sikhs with regard to

their sacred shrines. The militants argued that Operation Blue Star had been a

deliberate insult to the Sikh religion and to the Sikh community and therefore

would not go unavenged.13

Many civil rights groups also maintained that

innocent pilgrim s had been shot, among whom were women and children. The

army contended, on the other hand, that the soldiers had gone into the temple

with their hands tied behind their backs (so to speak), because the terrorists14

had used innocent pilgrims as human shields. They claimed that the army’s

losses were far in excess of what would have been expected in such a

confrontation because the soldiers were obliged to protect ordinary civilians.

For each element of the story, there were allegations and counter allegations.

The narrative’ s un® nished character meant that the event lived on in different

versions in the social memory of different social groups.

The uncertainty introduced by the assassination of Mrs Gandhi in the present

seemed to many to have a link with the past in the form of the incomplete

character of the story of Operation Blue Star. Mrs Gandhi’ s death appeared to

complete one segment of the story. It was as if this particular turn of events

was part of the larger plot that had been unfolding since Operation Blue Star.

Similarly, the uncanny knowledge of the identity of the guards that was passed

along in rumours seemed part of the same seriality through which events

unfolded on their own. Yet, despite the uncertainty in which events were

enveloped during the ® rst few hours, when rumours that Mrs Gandhi had been

shot began to spread, there was no sense of panic Ð even when the events

were ® rst con® rmed by the of ® cial media. It was only later in the evening that

things began to take a different turn.

R um ours of Celebration: W as the State Collapsing?

The initial speculations and judgements about the act of assassination began,

toward the evening of 31 October, to give way to uncertainty over the context

within which to place this action. Some believed that the assassination should

be understood as the ® rst announcement of much more momentous events to

come. In many parts of the city, it was rumoured that, along with this singular

act of daring, the Sikhs had started massive violence against the Hindus in

Punjab. Some people claimed to have heard that trains loaded with dead

bodies were arriving from Punjab There were rumours that Sikh militants

planned to poison Delhi’ s water supply. Also, it was said that there had been

wide-spread defection in the ranks of the army and the police, so that the

collapse of the state was imminent. Some people maintained that these events

had already been announced in the Gurdwaras,which was why instead of being

frightened of reprisals, the Sikhs were celebrating all over the country.

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

O f® cial N arratives, Rumour, and the Social Production of H ate 119

Other people expected that the creation of Khalistan would be announced

in the wake of the utter chaos that would surely reign in India; that negotia-

tions had already been held with powerful countries. Hence, while the United

States would not rush to recognise Khalistan, rumour held that it had already

persuaded some small countries to do so. In this way, Mrs Gandhi’ s assassin-

ation was seen as the ® rst act in a massive conspiracy. The representation of the

crisis drew considerable energy from exaggerated claims about the vulner-

ability of the Indian state and the amount of support the militants were likely

to receive. For example, in one of his speeches, Bhindranwale had reported that

he had been asked by some journalists if the Sikhs would ® ght on the side of

India if India were to be attacked by the Khalistani Liberation Force located in

the USA, Canada and England (sic) and supported by the American army. He

had said that he told the journalists that not one Sikh would lift a weapon

against such a holy force.

We can see that rumour operated here in the twilight of judgement. Carlo

Ginzburg has suggested to me in a personal conversation that it would be

useful to distinguish between those events which did happen (e.g., a train full

of dead bodies arriving from across the border during the riots in 1947) and

those events which were only alleged to have happened (e.g., a poisoning of

the water supply by the terrorists). The dif® culty with drawing sharp distinc-

tions between that which happened (the brute fact) and that which was only

alleged to have happened (the imagined event) is that such distinctions can be

seen with clarity only after the event has occurred . The regions of the imaginary

to which the claims of the real become anchored vary widely. In the case at

hand, they ranged from images seen on television, to those reported to have been

seen on television, and included, as well, stories heard about other times. Contrary

to the notion that certain classes of people are protected from the mesmerising

effect of rumours (e.g., the educated), I found that many professionals Ð

bureaucrats, teachers, medical doctors Ð inhabited for a time that twilight zone

in which it was dif ® cult to know whether it was wiser to believe in rumours

or in the of® cial versions of events.

Within twenty-four hours, the diffused rumours had created the sense that

there was a conspiracy against society; that the authorities responsible for the

protection of citizens’ lives and the maintenance of public order would be

overwhelmed and unable to ful® ll their designated roles. Thus, instead of

creating the Sikhs as a group vulnerable to mass violence and in need of the

protection of law, the rumours now fed on images of the Hindus as weak and

vulnerable and the state as already having collapsed in the face of such a

massive plot against it. That this version of reality would come to be accepted

Ð with a vengeance Ð can be traced to the frequent claims made in the written

and oral discourses of militant leaders about the emasculated character of the

Indian state; the exhortation to all Sikhs to carry weapons; and the repeated

assertions that when the moment of reckoning came, every Sikh would be

ready to ® ght on behalf of Khalistan (armed with the Kalashnikov guns

carefully stored in Sikh households). The assassination of Mrs Gandhi became

a kind of proof of the power of the Sikhs and the vulnerability of the Hindus.

These ideas gained greater and greater power as they began to be evoked with

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

120 V eena Das

growing intensity by many Hindu men and women who repeated the rumours

to one another in mounting panic.

The V ulnerability of the Sikhs

While rumours about the unpreparedness of the Hindus to meet the challenge

of a Sikh attack were being freely discussed and circulated on street corners

and at paan shops,15

many Sikh households feared that they would be attacked.

On 31 October, the new spapers had reported that hoodlums and thugs had

gathered in various railway stations and that in many places Sikhs had been

dragged out from the trains and beaten or killed . The absence of the police at

these crucial places and the fact that of® cial pronouncements completely

denied any attack on Sikhs, convinced many that the anti-social elements had

the support of the police. Many Sikhs saw these alleged attacks as in continuity

with Operation Blue Star, since both were about teaching Sikhs a lesson . Being

able to interpret the rumours correctly became a matter of life and death for

many. Let me illustrate with a personal example.

On the morning of 1 November, on the deserted streets in the Civil Lines,

my husband and I met a distraught Sikh gentleman waving wildly at us to

stop. He was an employee of the National Defence Institute who had been

working through the night. He knew nothing of the events following Mrs

Gandhi’ s assassination, as he had been in his laboratory the previous night, but

he could sense the eerieness of the city. No buses were running and in place

of the usual street sounds there was only silence. Could we help him and drop

him at the nearest bus stop, he asked, so that he could get home. We explained

that there had been reports of attacks on Sikhs and that therefore it might be

best for him to avoid the streets . Since one of our friends lived nearby, we

offered to take him there, where he could then telephone his family. We also

suggested that he could stay with our friend until things quietened a bit. As we

were talking, I had opened the door of the car for him. He got in and seated

himself in the back, muttering his thanks. Within a moment of his settling

himself in the car, a group of four or ® ve men materialised , seemingly out of

nowhere. They did not look threatening, but they spoke in conspiratorial tones.

Further ahead, they said, there was a mob. The men advised us that if the

crowd saw a Sikh in our car, they would not only drag him out and beat him

or kill him, they would attack us, as well. I became visibly angry. The men

shook their heads sadly and said they were doing their best Ð what more

could they do, besides warning us of what lay ahead? They suggested that it

would be better for the man to hide under the seat and for us to drive the car

fast so that our passenger would not be visible from far. They would neither

talk directly to the man nor look at him. It was as if they were discussing a

troublesome object rather than a person. At this point, the Sikh man panicked

visibly. I assured him that our friend’ s house was only a minute’ s drive, and

that if he wanted he could simply stay there, where he would be safe. `No’, the

man objected, `Why should you risk your life for me?’. He opened the door

even as the car had begun to move and stumbled into the road. `He will not

come with us. He probably fears that we may be trying to trick him and deliver

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

O f® cial N arratives, Rumour, and the Social Production of H ate 121

him to his killers’ , my husband surmised. I shouted to the man that he should

not try to negotiate the roads, that he should just go back to Metcalfe House

and hide there for a few days. I lost sight of him then but many months later

I saw him in the vicinity of Metcalfe House. He had survived .

Although many Sikhs were persuaded to take shelter in the houses of

Hindu or Muslim friends, it was with fragmented pieces of information that the

Sikhs were making their choices. The full impact of the violence had not yet hit

home. Horrendous violence against the Sikhs in resettlement colonies in Delhi

had begun on the evening of 31 October; more than 3,000 Sikhs died. The full

horror of this attack did not become known until 1 November, however, and

it and was not of® cially acknowledged until much later (see Chakravarty and

Haksar, 1987; Das, 1990b, 1996a; Kothari and Sethi, 1985; PUDR/PUCL Report,

1984).

A New Turn

As the facts about the extent and brutality of attacks against Sikhs came to be

known through reports in newspapers , and through the work of several

voluntary agencies, the rumours began to take a new turn. Now it was

whispered that news had come in from several places that Sikhs who had been

given shelter by Hindus had actually killed their hosts and stolen their goods

or raped their women before running off in the middle of the night. Names of

several residential colonies were evoked where people said they know some-

one who had seen acts such as these, or who had heard of such acts, or who

had known the stricken family (or families ).

This rumour soon became linked to the assassination of Mrs Gandhi. If her

security guards, whom she had trusted enough to ignore the advice of her

security personnel, could betray her trust and kill her because they had been

sworn to exact vengeance, then what further evidence was needed, people

asked each other, to show that Sikhs had no loyalty higher than what they

owed to their religion? The Sikh character was compared to that of a snake,

who turns round and bites the hand that feeds it milk. This snake analogy

surfaced over and over. For instance, one man helping to run a relief camp

went to buy milk for the children from a gwala (milk vendor). `Why do you

need so much milk?’ , asked the gwala, `Do you have a wedding in the family?’ .

The man explained that he was getting milk for the children of Sikhs in the

Ludlow Castle camp. The gw ala clearly disapproved: `You want to feed the

snake’s child with milk Ð but when he grows up, he will grow up to be a

snake, not a man’. He reminded the man about the saying, `Astin ke sam p Ð

m auka pate h i das lenge (snakes nourished in your shirt sleeves Ð they will bite

you as soon as they get a chance)’ .

The second of the strands intertwined in this complex of rumours was to

attribute the acts of ¯ ight on the part of terri® ed Sikhs to the preparations they

were making for revenge. In the earlier complex of rumours that I identi® ed,

the theme of the collapse of the state and the simultaneous passage of power

into the hands of the Sikh militants was prominent. By the second day after the

assassination, it was clear that the state would not fall. The new Prime Minister

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

122 V eena Das

had been installed. Reports of sporadic attacks on the Sikhs were percolating

in to Delhi and many people were scared to go out on the roads. They feared

that the mobs, persistently described in newspapers as composed of `anti-social

elements ’ , would take the opportunity to harm not only the Sikhs, but others,

too. At the same time, however, the rumours continued to construct the Sikhs

as aggressive, angry, and waiting to strike. When a large number of Sikhs took

shelter in G urdwaras, people said that the Sikhs had amassed a vast amount of

weapons and would launch an attack from these temples. In several middle-

class areas, residents organised night-time neighbourhood watches so that they

would not be caught unawares when attacked by the Sikhs.

In one incident, a group of frightened Sikh taxi drivers, who in normal

circumstances would sleep in make-do shelters at the taxi stand itself , had, in

order to avoid identi® cation, shaved off their beards and cut their hair on the

evening of 31 October, when they ® rst heard about the attacks on Sikhs in the

city. They were hiding in the dark shadows of the towering walls of a college

for women when they were spotted. The rumour immediately went around

that the taxi drivers had assembled there to attack the college and rape the

women. After receiving telephone calls from the college, the police came and

escorted the frightened men to a relief camp that had been set up near by.

Obviously, for many, the Sikhs remained the aggressors to the very end.

This particular angle on events Ð which turned the vulnerable victims into

aggressors and simultaneously created a sense of panic among those who (if

one were granted a god’ s eye view16

) were under no special threat Ð is

extremely important. I will return to this topic later in this article.

The third and fourth strands in the rumours through which Sikh character

was constructed address the fanaticism of the Sikhs, which was sometimes

deemed to approximate `madness’ . The emphasis in the narratives sometimes

shifted, but a slippage from courage to fanaticism to madness took place as the

stories circulated. Take ® rst a relatively benign construction. During one of the

® rst visits we made to Sultanpuri, a resettlement colony where I was involved

in relief and rehabilitation work (see Das, 1990b, 1996a), we were taken to a

street where little physical damage seemed to have occurred. But a group of

men and women there vociferously claimed their status as victims. One woman

told us, `We were all attacked Ð our men were killed in large numbers, but we

say that they were not murdered Ð they were martyred ’ .17

A young Punjabi

boy who was in the team of students helping me remarked, `Sikhs have such

an urge to claim martyrdom!’ . Changing the tone of his voice to mimic a

supposed Sikh, he said, `We want to be martyrs Ð you can put any place and

date on our act (asi tan ji shahid hona hai Ð jagah te tarikh tusi pa lao)’ . He then

described stories he had heard about Sikhs refusing to be moved to the safety

of refugee camps and trying to challenge fully armed mobs with the few

weapons they had, inevitably dying in the end. I too had heard these stories

but my interpretation was not that they were seeking martyrdom, but that it

was dif ® cult in those conditions to know whether they were being taken to a

refugee camp or being entrapped into an unknown situation where death and

degradation awaited them. I have already referred to the social memory of the

partition riots, during which people were lured to their deaths in a similar

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

O f® cial N arratives, Rumour, and the Social Production of H ate 123

manner. This memory may have shaped the Sikhs’ thoughts as they tried to

decide on the best course of action.

The fourth strand, a more harmful variation on the third , construed the

Sikhs not as misguided martyrs but as people whose fanaticism shaded into

madness. In my efforts to assist in the work of relief and rehabilitation, I was

trying to organise medical help for the victims. About ® fteen days after the

riots, I spoke with a group of physicians. My primary contact in this group was

a socially conscientious physician (a member of the Arya Samaj) who had read

some accounts of the work my colleagues and I were doing. He said that he

wanted to organise his colleagues to help but found them to be so prejudiced

against the Sikhs that it was impossible to get their co-operation. `Please don’ t

think that they [the physicians] are bad people’ , he pleaded, `One of them had

worked day and night in a government hospital on a voluntary basis when a

tornado had hit the city in 1978 and hospitals were ® nding themselves very

short of staff. But somehow even he cannot be persuaded to work with the

Sikhs’ . I decided that by talking directly to this group and telling them of the

suffering of the victims in Sultanpuri, I might be able to persuade them to help.

My descriptions, were met with sullen resistance, an attitude that I con-

strued at the time as a refusal to listen. One woman doctor said that in her

opinion the Sikhs had brought all this suffering upon themselves because they

were like mad people. To substantiate her view, she told me that she had heard

that in the tyre market near Bada Hindu Rao, where many shops were owned

by Sikhs, an angry mob had put burning tyres around the necks of the owners,

locked them inside their shops, and let them burn to death. `Normal’ people,

she said, would have shouted and asked for mercy or forgiveness , but one Sikh

was seen in the windows gesturing threateningly at the mob, with his ® sts

closed. His actions had elicited hysterical laughter from the mob.

The creation of these images did not seem to have anything to do with the

physicians’ actual experience. They had not themselves gone around burning

people or looting shops. And yet, there was a voyeuristic pleasure in these

rumours of Sikh madness and extraordinary behaviour. As the discussion

gathered momentum, other kinds of evidence were offered. One person

claimed that in the Sikh tradition it was believed that one who died for the

cause of the Gurus did not feel any pain even under torture. This belief was

why, he said, Sikhs behaved like fanatics, taking questions of life and death

lightly. This belief also explained, according to him, why there had always been

so much violence in the Punjab. Another example of supposed outlandish Sikh

behaviour followed. Sikhs in the Punjab, one member of the group asserted ,

had proudly proclaimed the story that when the Indian army had rounded up

Sikh boys who had been caught in the Golden Temple during Operation Blue

Star, the soldiers told the boys to shout `Bharat m ata ki Jai (Victory to Mother

India)’ . Instead, they had shouted in unison, `Jo bole so nihal Ð bolo sri sat sriya

kal’ , the ritual proclamation of Sikh faith. The Indian army of® cer in charge

alleged ly then killed these young boys.18

How could one explain such madness ,

the physician wondered, who would not value the life of the young? The

tragedy was that he seem ed completely unaware of the irony that we were

discussing precisely the brutal killing of the Sikhs when he evoked their lack of

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

124 V eena Das

respect for life as evidence of their madness and hence an exoneration of the

violence against them.

I could give many more examples of this overlaying of the four strands to

create the Sikh as aggress ive, vengeful, incapable of loyalty, and mad Ð and,

correspondingly, of the Hindu as vulnerable, frightened, and acting out of self-

defence against a powerful enemy. But I shall conclude this part of the

description with an example which provides a sligh tly different angle on this

theme.

A Hindu priest told me that on the 31 October, a meeting had been held just

on the outskirts of Delhi in a recently built temple, known for its lavish

interiors. The temple was also known to be patronised by politicians and

powerful members of the underworld. During this meeting, the priest said,

there had been a major discussion about whether Sikhs were part of the Hindu

community. If so, the assassination of Mrs Gandhi would be properly treated

as an individual aberration/crime; if not, the whole community would be

implicated. It was agreed that for the last several years the Sikh militants and

terrorists had killed , terrorised, and looted Hindus in the Punjab.19

The sense

of the meeting was (according to the priest) that the Sikhs were now like a god

who begins to behave like a demon.20

Such gods, he said, do not learn through

reasoned conversations; they have to be kicked to rid them of evil. The phrase

he used in Hindi was `laton ke devata baton se nahin m ante (the gods who need

kicks cannot be paci® ed with words)’ . The common saying is `laton ke bhut Ð

baton se nahin m ante’. The word devata means god; bhut means demon. He had

substituted the symbolism of demons with the symbolism of gods (though the

term devata is used for lesser gods). Thus, the language of exorcism and

possession here becomes a political language through which the violence links

the aggressors and the victims on the model of the exorcist and his patient.21

These rumours on the un ® nished events of Operation Blue Star fed into a

Hindu imagination of society as in siege after the assassination of Mrs Gandhi.

Let me hasten to add that this does not mean that all Hindus believed in one

kind of construction and all Sikhs in its binary opposite. Indeed, the very

examples I have given of solidarities created across communities in these

periods of crisis is evidence that such totalisation is resisted by members of

both groups (see Kanapathipillai, 1990).

There were two important consequences of the rumours. First, they built a

structure of thought within which Sikh character was placed, and this structure

was one with characteristics similar to that of paranoia. Second, the rumours

stabilised a reality that then had to be handled by the residents of resettlement

colonies, like Sultanpuri, which bore the brunt of the violence. Thus, the

movement of images that built the stereotypes of Sikh and H indu character in

the militant discourse travelled to the Hindu constructions in giving form to

rumours which in turn made brutal violence against the Sikhs a `thinkable’

response, even for those who did not directly participate in the violence.

For the Sikhs, the situation developed in a parallel but not strictly sym-

metrical way. This was because of a deep ambivalence toward Mrs Gandhi’ s

assassins. None of the militant groups was willing to condemn the assassins in

unambiguous terms. Sometimes, public statements from Sikh relig ious organi-

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

O f® cial N arratives, Rumour, and the Social Production of H ate 125

sations condemning the assassination were claimed to have been made and

then later withdrawn. For many Sikhs who were not necessarily sympathetic to

the militant cause, Operation Blue Star was seen as an insult to the whole

religious community. Hence, they saw the aftermath of the assassination as a

further step in the politics of teaching Sikhs a lesson. Many were offended that

instead of being treated as an individual crime, the assassination was being

seen as an event that put the whole Sikh community in the dock. So, instead

of being treated as individual citizens who had nothing to do with the alleged

crime, Sikhs were being targeted as people to whom a lesson had to be taught.

It would be a grave error, however, to assume a homogeneity of opinion and

a consequent totalisation of affect among the Sikhs. Opinions varied , from

celebrating the act in the tradition of martyrs (e.g., in the militant literature) to

distancing the self from the assassins, to outright condemnation of their act.

This variation was given no recognition in the stabilisation of the attributes

of `Sikh character’ Ð although the individuality of persons who make up a

community was clearly articulated in social practices. Such acts of totalisation

seem to be a normal characteristic of times of collective violence; they are also

apparent in the processes of ethnic and religious mobilisation in the service of

violence (see Das, 1990a, 1995; Kanapathipillai, 1990). What I would like to

emphasise here is the manner in which, for many Hindus, the categories of

aggressor and victim were reversed through the application of these notions.

Different strands of rumour combined here to create (a) a sense of vulner-

ability among the Hindus through the creation of an imaginary world in which

the whole social order was seen as precarious, about to collapse as a result of

a massive conspiracy on the part of the Sikhs, even though it was the Sikhs on

whom the violence was being unleashed; and (b) an assumption that the Sikh

was not worthy of being treated as an other w ith a face, because the imagined

Sikh character was devoid of all human subjectivity , a creature of madness and

demonic possess ion.

The peculiar nature of rumour Ð its lack of signature, the impossibility of

its being tethered to an individual agent Ð gave it the signature of an

`endangered collectivity’ , one that led to the transformation of the world into

a `fantasmagoria of shadows, of ¯ eeting , improvised men’ (Lacan, 1993). It may

be said thus that many Hindus participated in the collective violence, even

though they may not themselves have engaged in any killing or looting. The

form of language Ð its force, its lack of signature, its appeals to the uncanny

Ð gave it the perlocutionary force that brought a new form, not a form of life

but (and I shall not hesitate to say it) a form of death into existence.

C onclusion

I want to conclude by suggesting that what I have described is how the world

as it is known in everyday life becomes obliterated, replaced by a world which

bears resem blance to the structure of paranoia. The social production of hate

can give birth to discourses and practices of genocide, but there are special

conditions through which such transformations become possible. My fear of the

other is transformed into the notion that the other is fearsom e. I have tried to

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

126 V eena Das

show further that such transformations are bound to the conception of import-

ant past events as `un ® nished ’ and capable of moulding the present in new and

unpredictable ways. It is not only the past then which may have an indetermi-

nate character; the presen t too may suddenly become the site in which

elements of the past that were rejected , in the sense that they were not

integrated into a stable understanding of the past, can press upon the world

with the same insistence and obstinacy with which the real creates holes in the

symbolic. It is in this manner that rumour’ s adequacy to a reality which has

abruptly become unrecognisable makes it the privileged mode of communi-

cation and constructs panic as its corresponding affect in this altered world.

Doubts and uncertainties exist in everyday life, but the worst is not what one

expects to happen every time. In contrast, the zones of emergency are marked

by diffused images of an un® nished past, efforts to void the other of all

subjectivity, and a world increasingly peopled with a fantasmagoria of shad-

ows. The perlocutionary force of rumour shows how fragile the social world

we inhabit may be. If in this article I take a considerably more pessimistic view

of the role of rumour in mobilis ing hate than many others have taken, it is not

because I completely reject the possibilities of such forms of communication to

create critical awareness but because I wish to show that rumour has a darker

side. Images of hate between social groups may take a volatile form when the

social order is threatened by a critical event and so transform the world that the

worst becomes not only possible but also probable.

V eena Das m ay be contacted at D epartment of Anthropology, Graduate School, N ew

School for Social Research, 66 West 12th Street, N ew York, NY 10011, USA, until

July 1998.

Notes

1. I am grateful to Charles Briggs for his acute comments on an earlier draft

of the paper. While I have tried to provide details on many of the events

that may not be widely known to readers outside the subcontinent, it is not

possible for me to write by displacing the Indian audience. I offer my

apologies if this makes the text less sociable or hospitable for other

audiences.

2. Recorded audio-cassettes of Bhindranwale’s speeches were widely circu-

lated in the Punjab and elsew here in the 1980. These do not give the dates

on which a particular speech was made.

3. The question of the gender of the nation is too complex to address here.

Almost every conceivable image, from seeing the nation as an all consum-

ing mother, a courtesan, a goddess, a beloved, to a sodomising father, has

made an appearance in the social imagery of different groups. (See Chan-

dra, 1992; Chatterjee, 1993; Cohen, 1995; Das, 1996b; Kaviraj, 1995.)

4. The term G uru literally means teacher, while the term Sikh is a derivative

of shishya , meaning student.

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

O f® cial N arratives, Rumour, and the Social Production of H ate 127

5. See They Massacre Sikhs. A White Paper by Sikh Religious Parliament (Shiro-

mani Gurdw ara Prabandhak Committee. Amritsar, no date).

6. The Akali Dal is a political party based primarily in the Punjab among the

Sikhs. Not all Sikhs owe allegiance to the Akali Dal, however. The Con-

gress Party has been its main rival in the State, and the latter has also had

a long history of Sikh leadership and Sikh support. The interface between

religious organisations and political parties in the Punjab is a complicated

question; it should be borne in mind, though, that political allegiances for

both Sikhs and Hindus cut across political parties.

7. The relation between memory and forgetting in constituting the com-

munity has been noted in many contexts in recent years. In a very

interesting paper, Gross (1986) shows the importance of memory in the

resistance to totalitarianism, and of simultaneous forgetting for the con-

struction of community as purged of its past evil in the case of Polish-Jew-

ish relations during the second world war. He comments powerfully on the

Polish conviction that `a half way victory over totalitarianism’s attempts to

destroy social solidarity would still be won if the community’s history

were rescued from the regime’ s ambition to determine not only the

country’s future but also its past’ . Yet the same Polish people developed

elaborate myths to conceal from themselves the nature of Polish-Jewish

relations and the anti-Semitism in Polish society which led to both covert

and overt support being given to the fascist ideology of scapegoating the

Jew.

8. Operation Blue Star refers to the army offensive allegedly to ¯ ush out the

militants from the Golden Temple undertaken in July 1984 in which

Bhindranwale died. There is a voluminous literature on this, but see Kapur

(1987).

9. Although panic seems to be an accompanying effect of rumours in times of

trouble, I hesitate to assume that all rumours are accompanied by panic. I

do not have the space here to deal with rumour in everyday life, but a

general theory of rumour needs to come to terms with the different

discursive forms within which rumours function.

10. See also Lefebvre (1973, p. 74), who wrote, `Indeed, what was the Great

Fear if not one gigantic rumour?’ .

11. In Austin’ s (1975) classic formulation, illocutionary force is distinguished

from perlocutionary force in that in the former case one does something in

saying something while in the latter case one does something by saying

something. The presence of the ® rst person indicative marks out utterances

that have illocutionary force. In the case of perlocutionary force the

situation is much more complicated but at least in the case of rumour we

can say that its force would be lost if it was tethered to the words of the

speaking agent; or for that matter, if one were to frame a rumour by saying

that `I am spreading the rumour that ¼ ’ .

12. The idea of the two guards as martyrs was to crystallise much later in the

militant literature of the Sikhs. The assassins came to be seen as incarna-

tions of two heroic ® gures, Sukha Singh and Mehtab Singh who had killed

a minor Muslim chieftain, Massaranga, in order to avenge the dishonour he

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

128 V eena Das

had done to Harmandar Sahib in Amritsar in 1752. See Das (1995). To my

knowledge, this interpretation was not evoked early on. Initially , the men’s

action and the risk to their own lives was compared to the suicide squads

of different militant groups in the Middle East

13. For a further description of the rendering of this event in militant literature,

see Das (1995).

14. In the army’ s view, the followers of Bhindranwale were `terrorists’ ; in their

own self-understanding, they were `martyrs’ , and in popular usage in the

media or in conversations, there was frequent slippage between different

kinds of terms. As Kosseleck (1985) has repeatedly pointed out, most social

scienti ® c concepts are marked by a political plenitude. In this case what is

clear is that ordinary people in the Punjab and elsew here had to bear the

burden of much violence due to both the insurgency operations of the

militants and the anti-insurgency operations of the police and the army.

Thus it would be a mistake to assume that the distribution of terms was

neatly distributed among a populace ® ghting for freedom or justice on the

one hand and a repressive state apparatus on the other.

15. Paan shops, i.e., small road-side shops where betel leaves and betel nuts are

sold, are typical gathering places for exchange of news, gossip, and

information. These are strongly gendered spaces: women would not be

found standing and gossiping around these shops.

16. I am tempted to say that the `objective’ conditions did not warrant this fear

of a plot against the whole of society being hatched by Sikhs. But the

problem in this essay is precisely to see a crisis by placing oneself within

it and to explain how categories of people who are themselves vulnerable

come to be attributed with such evil powers.

17. I have shown in my earlier work that in the streets where the violence

occurred , people simply did not use the category of martyrdom, nor did

any other ready-made categories seem to come very easily to them (Das,

1990b).

18. This story was much in circulation after the Operation Blue Star and was

cited in a letter by a senior police of® cial in Punjab in his resignation letter

to the then President of India, Sardar Zail Singh (see Das, 1995).

19. There was alw ays a `forgetting’ at these points that, statistically, more Sikhs

than Hindus had died in terrorist attacks.

20. Hindu mythology is replete with such examples and the expression `devata

chaddha hai (a god has possessed one)’ and `bhut chaddhaa hai (a ghost has

possessed one)’ can both be used in seeking to exorcise a troublesome

spirit.

21. See Kepferer (1990) for some analogies with the situation in Sri Lanka,

where the themes of possession and exorcism made an appearance in

political cartoons in relation to the Tamil militants .

R eferences

Anderson, B. (1983) Imagined Com munities. Re¯ ections on the Origin and Spread of

Nationalism . London: Verso.

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

O f® cial N arratives, Rumour, and the Social Production of H ate 129

Anonymous (n.d.) They Massacre Sikhs. A White Paper by Sikh Religious Parlia-

ment, Amritsar: Shiromani Grdwara Prabandhak Committee.

Austin, J.L. (1975) H ow to D o Things with W ords, J.O. Urmson and M. Sbisa (eds)

Cambridge, MA: Harvard University Press.

Bhabha, H.K. (1995) `By Bread Alone: Signs of Violence in the Mid-nineteenth

Century’ , in Location of Culture, London: Routledge.

Chakravarty, U. and N. Haksar (1987) The D elhi Riots: Three D ays in the Life of

a N ation, New Delhi: Lancer International.

Chandra, S. (1992) The O ppressive Present: Literature and Social Consciousness in

Colonial India, Delhi: Oxford Univers ity Press.

Chatterjee, P. (1993) N ationalist Thought and the Colonial World: a D erivative

Discourse, Minneapolis: University of Minnesota Press.

Cohen, L. (1995) `Holi in Banaras and the Mahaland of Modernity’ , GLQ , 2:

399± 424.

Connerton, P (1989) H ow Societies Rem ember, Cambridge: Cambridge Univers ity

Press.

Das, V. (1990a) `Introduction: Communities, Riots, and Survivors’ , in V. Das

(ed.) Mirrors of V iolence. Com munities, Riots, and Survivors in South Asia,

Delhi: Oxford Univers ity Press.

± (1990b) `Our Work to Cry: Your Work to Listen’ , in V. Das (ed.) Mirrors of

Violence. Com munities, Riots, and Survivors in South Asia , Delhi: Oxford

University Press.

± (1995) Critical Events: An Anthropological Perspective on Contemporary India,

Delhi: Oxford Univers ity Press.

± (1996a) `The Spatialisation of Violence: Case Study of a `Communal Riotº ,

in K. Basu and S. Subrahmanyam (eds) U nravelling the Nation. Sectarian

Con¯ ict and India’s Secular Identity, Delhi: Penguin Books.

± (1996b) `Language and Body: Transactions in the Construction of Pain’ ,

Daedalus, 125 (1): 67± 91.

Gross, J.T.T. (1986) `Polish-Jewish Relations During the War: an Interpretation’ ,

Arch ives Europeens de sociologie, XXVIII: 199± 214.

Guha, R. (1983) Elementary Aspects of Peasant Insurgency in Colonial India, Delhi:

Oxford Univers ity Press.

Kakar, S. (1995) The Colours of Violence. Delhi: Viking.

Kapferer, B. (1983) A Celebration of D em ons. Exorcism and Aesthetics of H ealing in

Sri Lanka , Washington DC: Smithsonian Institution Press.

Kapur, R.A. (1987) Sikh Separatism : the Politics of Faith , Delhi: Vikas Publications.

Kanapathipillai, V. (1990) `July 1983: the Survivor’ s Experience’ , in V. Das (ed.)

Mirrors of Violence, Delhi: Oxford University Press.

Kaviraj, S. (1995) The Unhappy Consciousness: Bankimchandra Chattopadhyay

and the Form ation of N ationalist D iscourse in India, Delhi: Oxford Univers ity

Press.

Kosselleck, R. (1985) Futures Past: O n the Sem antics of H istorical Time, Cambridge

MA: MIT Press.

Kothari, S. and H. Sethi (1985) Voices from a Scarred City : the Delhi Carnage in

Perspective, Delhi: Lokayan.

Lacan, J. (1993) `The Other and Psychosis’ , in J.-A. Miller (ed) The Sem inar of

Dow

nloa

ded

By: [

CD

L Jo

urna

ls A

ccou

nt] A

t: 01

:35

23 M

ay 2

007

130 V eena Das

Jacques Lacan. Book III The Psychosis, 1955 ± 1956, trans. by R. Grigg , New

York: Norton.

Le Bon (1952) The Crow d , London: T.F. Unwin.

Lefebvre, G. (1973) The Great Fear of 1789: Rural Panic in Revolutionary France ,

London: Armand Colin.

Muscovici, S. (1985) The A ge of the Crowd: a H istorical Treatise of Mass Psychology ,

Cambridge: Cambridge University Press .

Nandy, A. (1994) The Illegitimacy of Nationalism , Delhi: Oxford University Press.

Pandey, G. (ed.) (1993) H indus and O thers, Delhi: Penguin Books.

± (1994) `The Prose of Otherness’ , in D. Arnold and D. Hardiman (eds)

Subaltern Studies, Vol. 8, Delhi: Oxford University Press .

PUDR/PUCL (1984) W ho are the Guilty? Report of a Joint Inquiry into the Causes

and Impact of the Riots in D elhi from 31 O ctober to 10 N ovem ber, D elhi.

Rude , G. (1959) The Crow d in the French Revolution, Oxford: Clarendon Press.

± (1964) The Crowd in H istory , 1730± 1848, New York: Wiley.

Thompson, E.P. (1971) `The Moral Economy of the English Crowd in the

Eighteenth Century’ , Past and Present, 50: 76± 136.