only had the police come and camped and stopped the fight, they had fired on the mob and killed people. Suddenly the stakes had risen dramatically. Mr. Ghosh's point of view continued to be woven into his historic memory. The only named official was not the high-ranking S.D.O., but a Hindu policeman who played the active role. Mr. Ghosh called him Raghubabu, a familiar and respectful form of address to a Hindu bhadralok, thus suggesting that they were acquainted, that this man of power was on his side. The Hindus, in Mr. Ghosh's account, continued to behave lawfully. They were stronger, more courageous; they advanced, but when the police told them to retreat they did so. The Muslims, on the other hand, disobeyed, moving forward, and so more of them were killed when the police fired. Mr. Ghosh managed to convey the notion that there was justice in their disproportionate slaughter. At first he seemed unsure just how many had died, but when pressed he gave us exact figures and an amazing interpretation:
Two Muslims were killed, and one Hindu. Since two men were killed on the Muslim side, they had to kill one more on the Hindu side. Otherwise it would look as though the police were partial. So the police later ran and killed a man on the Hindu side, to make it even.
Mr. Ghosh was clearly a reasonable and respect-worthy individual. Yet his construction of events seemed improbable. Biased accounts, memories at variance depending on the allegiances of the storyteller, were only to be expected. Who among us has ever engaged in contentious activity, from a fight with one's mate to a union battle or a campaign for a favored political candidate, without skewing reality toward our own perspective? But bias is more than false representation. If we take seriously the accounts of those involved, what emerges is a history of differing realities. Not only did people remember differently, or report differently; they actually lived the experience differently.6
To understand an event, not to mention a phenomenon, we must encompass the varying experiences of that event, even though we must understand that what we understand is only an approximation. That the approximation departs from the experience of the actor is not a problem; it is the point. How experience is experienced is one topic of importance. But how that experience is formulated, remembered, and retold tells the hearer something beyond “what happened,” which we cannot in any case know and which did not in any case happen, since what happened happened to many different people differently. To add to the fun, what happened also changed as it happened and went on changing later. Memory is formed by past knowledge and experience, but it is also altered by the future. Each new experience, including each telling, changes the tale. The telling itself is part of the experience, for what we live combines with what we think to construct what we do, and all three in constant interplay define experience. When the experience is a social one, when the telling is done by members of a group about a group event, the history of memory sheds light on the history of events. Collective memory is social action in its own right, and it is part and parcel of every historical act.7
The Muslims
Riveted by Mr. Ghosh's honest and forthright one-sidedness, I didn't immediately register the arrival of a newcomer. He was hard to miss, though, an imposing presence in the crowded room. He was the elected chairman of Panipur Union, Altaf-uddin, the very man to whom Mr. Ghosh had referred a little earlier. He wore a long white beard, the long white shirt and lungi, or ankle-length skirt, favored by Muslim men, and an Islamic cap. Our guide to this village had scheduled us to interview him first, at his home. I had, it seemed, upset social protocol by dropping in before that to see Basantibala. Word spreads quickly in a village, and when he had heard we were there, talking in the Majumdar house, Altaf-uddin had determined to right the situation by bringing himself to us.
He also brought a marked change to the tone of the discussion. Once Altaf was settled, I turned back to Mr. Ghosh and asked him what his reactions had been to the battle. There had been a decided shift in Mr. Ghosh's sails:
We had no adverse reaction.
Why not?
[With a sideways glance at Altaf: ] Because there were influential elders in this area. So whenever there were tensions among us, they would intervene. So we lived as brothers.
Altaf, as influential elder, promptly agreed:
There were some small incidents [before the riot], but they were smoothed over quickly…. Once an incident began, all the Hindus would take the side of the Hindus and all the Muslims would take the side of the Muslims. The influential people, both Muslims and Hindus, would come forward to solve the problem.
I already knew that Hindus talked differently in the presence of Muslims, and the tension in this case was tangible. So I eased Mr. Ghosh off the hook, turning to Altaf and asking for his version of the story.
I got the news at night that there was a little conflict between Hindus and Muslims, and that the Hindus were already out organizing a riot for the next day. The reason was that a cow ate the lentils in one field. A Muslim's cow ate the kheshari in a field of a Hindu. It was a very petty thing. For that, they had some chase and counter-chase in the late afternoon. A little fighting, too. Now the Hindus were out with their horses to inform the other Hindus to come next day, to riot.
Where before Muslims plotted in the bazar, now Hindus rode the countryside mobilizing warriors. If Mr. Ghosh spoke for caste Hindus, I thought, Altaf-uddin was about to give me the version of the Muslims.
Muslims in Bengal
Islam became a factor in the life of the subcontinent very early: Arab traders journeyed there within a few years of the death of Muhammad in 632. However, the Muslim community in India traces its roots to Moghul conquests much later, in the twelfth century. The first sultanate in India was established in Bengal, at Gaur in the district of Malda, in the early thirteenth century, and Moghul rule was consolidated elsewhere in the subcontinent only four hundred years later.8
The earliest Bengali Muslims were immigrants from central Asia, Afghanistan, Persia, Arabia, and northern India, 9 but only a tiny minority of the subcontinent's Muslim population today can trace their ancestry to immigration. Most are converts. Until the British government's first census of the area in 1872, Bengal was considered to be a Hindu domain. What that census revealed shocked both rulers and indigenous elites: Muslims constituted very close to half the population, and in some areas they were an imposing majority.10 Their distribution was uneven: some western districts, including those that had housed the earliest Muslim administrations, showed them in the minority. But in the east, around Dhaka, the capital of Moghul Bengal from 1612 on, Muslims constituted 60 percent of the population.11
Within that community there were vast distinctions. If the gulf separating high—from low-caste Hindus was enormous, that between upper- and lower-class Muslims was in some respects even greater. Class tended to coincide with origins. Upper-class Muslims traced their heritage to immigrants and claimed membership in a group called the ashraf. Converts, or the atrap, were drawn from the most oppressed among the population. While they were theoretically united by common worship and a theology lacking the sorts of rigid distinctions Hindus suffered through caste, ashraf and atrap Muslims were nonetheless severely alienated by culture and language. The ashraf prided themselves on their knowledge and use of Urdu, the language of the Moghul court at Delhi, which is spoken widely by Muslims in the northwestern parts of the subcontinent. To the ashraf in Bengal, Bengali was a crude and unworthy tongue. They especially disdained the dialects most common among the rural masses. Upper-class names, such as Syed and Shaikh, were similar to those current in Arabia and Persia. In fact, not all people with these names could claim direct descent from Arabian or Persian immigrants; high-caste Hindus who converted tended to be awarded these honorifics. People with Bengalified names, such as Mandal, Pramanik, Sarkar, were common among the peasantry and were held in contempt.12 So, too, were those Muslims who practiced despised occupations-weavers, shoemakers, barbers, and the like-all of whom, had they been Hindus, would have been Untouchables.
Between those who made clear claim to the distinction of ashraf and those who were atrap was a very small rural gentry. Although they occupied somewhat the same economic position as the Hindu bhadralok, and although they contributed some superb and beloved