Genocidal violence leaves but a broken historical trace. Not surprisingly, therefore, the surviving records of the subcontinent's (1) Partition are marked by their fragmentariness. They move, in fits and starts, through jerks and breaks and silences--incoherent, stuttering, even incomprehensible--between the poles of testimony and rumor.
Testimony, Langer notes, is "a form of remembering." Rumor, by contrast, is a form of doing--of making happen--by telling. (2) The record of Partition clearly bears the mark of both. The importance of first-person testimony (for the judge, as for the historian) requires no underlining. "I was there"; "I saw"; "I can name"; "I recognize"; and (more than occasionally for the journalist, as well as for the historian, though less commonly, we are told, for the judge) "I learned from the most reliable witnesses." Testimony's method is that of particularizing and individualizing, specifying sites and bodies that carry the marks of particular events, making "real" in everyday, physical, nameable terms. Its difficulty in the "limit case" is that it needs to articulate an unparalleled, "unthinkable" history struggling to find voice. How does the witness share "the particularity, the unshareability, and the incommunicability of pain in torture"? (3) How can we speak for the dead, who are no longer present? How can we t estify on behalf of the dead, if we are not dead? How can anyone who is not a Muselmann know what it is to be a Muselmann, as historians of the Holocaust have repeatedly said? (4)
The importance of rumor in the record of violence is also established, though perhaps more in the matter of its making than in that of its evaluation or reconstruction. Rumor moves in a direction almost contrary to that of testimony: generalizing, exalting to extraordinary (even miraculous) status, and employing the sweeping terms of deluge and just desserts (actual or impending). In rumor, language is transformed from a mode of (possible) communication to a particular kind of imperative condition, communicable, infectious, possible (and almost necessary) to pass on. The impact of this anonymous, mercurial, fleeting figure is well attested in accounts of the history of violent uprisings--from Lefebvre's and Rude's writings on the French Revolution to Guha's analysis of peasant insurrections in colonial India and Veena Das's account of the 1984 massacre of the Sikhs. (5) That rumor is no stranger also to the written records and oral accounts of 1947 is hardly surprising. How seriously has all this affected our assessment of that moment?
It is the purpose of this article to examine the extent to which historical discourse on Partition, from 1947 to today, takes the form of testimony or that of rumor--or hovers between the two. For this purpose, I focus on the twin questions of violence against women and the number of casualties, both of which loom large in the annals of the event. "The figure of the abducted woman became symbolic of crossing borders, of violating social, cultural and political boundaries," Menon and Bhasin write. By the time that the rape, looting, and migrations were finished, "about eight to ten million people had crossed over from Punjab and Bengal ... and about 500,000-1,000,000 had perished." (6) "Estimates of the dead vary from 200,000 (the contemporary British figure) to two million (a later Indian estimate) but that somewhere around a million people died is now widely accepted," writes Butalia. She goes on to note the statistical evidence of "widespread sexual savagery": "about 75,000 women are thought to have been ab ducted and raped by men of religions different from their own (and indeed sometimes by men of their own religion)." (7)
Observers described the violence that erupted so fiercely between Hindus/Sikhs and Muslims in 1946 and 1947 as "a war on each other's women" and as a war waged "especially" on women and children. "Unless Hindus and Muslims stop this war on each other's women folk," F. V. Wylie, the governor of Uttar Pradesh, wrote immediately after the Garhmukhteshwar massacre of November 1946, "the whole country may go mad." (8)
As a result of a number of important recent writings, the extent and brutality of this war is now fairly well known. (9) That the concentrated and extended violence against women was glorified all around is also established. The celebrations emphasized the pride and valor of the community. They included recitations of the "unparalleled sacrifices" of "our" women, as well as boastful tales of the capture, disfiguring, rape, and humiliation of the other's women--even, one suspects, in cases where the women concerned were later made into "our" own, adopted, married, and settled into the "victorious" local community.
The discourses surrounding this "war on women"--on the issue of both rape and sacrifice--require some reflection since they had much to do with the progress of the war itself and they crucially conditioned the fate of its women victims. New tales of bravery and new accounts of tradition were retailed. These were developed in large part to restore pride and self-respect in the midst of humiliating circumstances. If this entailed the gathering of testimony, the remembering of sacrifices to preserve and strengthen the community, it also entailed a fight against rumor--the hushed insinuations and the extravagant claims that became the primary mode in which the "shaming" questions of women's honor and community losses were discussed in 1946-1948. Yet the disciplining of rumor and the (re)imposition of self-respect and order was accomplished, paradoxically, by the reproduction of the conditions in which rumor flourished and by the incorporation of rumor into the "historical" accounts of the period, as I show in the following pages.
Primary Discourse: The Signature of Rumor
In his writings on peasant insurgency, Ranajit Guha distinguishes between three levels of "historical" discourse: a "primary" level, referring to reports from the front, as it were; a "secondary" level--commentaries and memoirs that aspire to the status of history; and a "tertiary" level, which might be described as history proper, with the full paraphernalia of referencing and footnotes, objective distance and "scientific" language. Of primary discourse, Guha notes that it is marked both by its immediacy and its official character. (10) This is certainly the case with the first reports we have regarding many individual instances of looting and killing. The "First Information Reports" found in police records, for example, belong to a form of reporting crime in the police posts of the Raj whose very condition of existence was a bureaucratic/judicial context, language, and imperative.
Yet statements of this kind are not always the first that we come across regarding the occurrence of "riots" or "disturbances," especially in the twentieth century. The first reports we have of major incidents of violence, faithfully (or not so faithfully) reproduced in newspaper reports and official documents, are often rumors: "Something has happened there"; "Don't go further, it is not safe just now"; "A disturbance has broken out"; "People are fleeing"; "The shutters are down"; "Curfew has been imposed"; "Two [or ten, or a hundred or more] people have been killed." (11) The state usually makes its appearance only in the wake of these rumors and other fleeting reports. And many of the state's inaugural reports--the first that historians often have to hand--are marked by the signature of rumor, jostling now with a vocabulary of "civilization" and "counterinsurgency." Rumor is marked characteristically not only by indeterminacy, anonymity, and contagion, but also by a tendency to excess and "certainty" (12)- -a certainty confirmed when the report moves from a verbal to a graphic or filmic mode. As Lefebvre has it on revolutionary France,
It goes without saying that (this mode of transmission) favoured the spread of false reports, the distortion and exaggeration of fact, the growth of legends. ... In the empty silence of the provinces, every word had the most extraordinary resonance and was taken as gospel. In due course, the rumor would reach the ear of a journalist who would imbue it with new strength by putting it into print. (13)
Consider one of the earliest reports in a Delhi newspaper on the massacre of November 1946 in and around the Garhmukhteshwar fair, not fifty miles from the capital. Under the headlines "Pilgrim Train Attacked Near Meerut: 50 Butchered, Over 100 Injured: Another Forty-five Perish in Neighbouring Village," the account reported that "about 200" had been killed at Garhmukhteshwar on November 7. (14) The rounded-off figures are one indication of the persistence of rumor: precise and yet extravagant, they are suggestive of so much more than the numbers themselves. And such numbers are easily multiplied for reasons of journalistic or political advantage, as observers have noted in the case of more recent "riots" in India. (15)
Or consider the following report from the commissioner of Meerut, F. W. W. Baynes, the most senior Indian Civil Service (ICS) official in the area, writing after a personal inspection of the site of the massacre. Baynes described some of the worst horrors:
3 women and 9 babies all dead in one well, a child of three or four with its face kicked in and then charred with fire, a woman beaten to death while in the act of producing a baby, girls of twelve or thirteen raped, and then killed by thrusting spears up them and ripping them apart. (16)
The description perhaps outdoes the Hindustan Times in its precision and its extravagance. It appears as an eyewitness account; what is notable, however, is that every part of this sentence, except for the clause relating to the number of bodies found in the well, depends on hearsay: hearsay transformed into truth--the "truth" of the riot, and the "truth" about India in the last days...